Home

SofI Manual

1. Parameter signature Header Keyword Value Description Exposure Name DET EXP NAME SOFI File name prefix DET DIT DET DIT NODEFAULT Detector Integration Time individual exposure sec DET NDIT DET NDIT NODEFAULT Number of DITs averaged into an individual image Number of columns DET WIN NX 1024 Number of columns in the window Number of rows DET WIN NY 1024 Number of rows in the window First column of window DET WIN STARTX 1 First column of window First row of window DET WIN STARTY 1 First row of window Spectro Mode INS SMODE NODEFAULT Spectroscopic Mode Which Slit INS WHICHSLIT NODEFAULT Which Slit Combined offset F T SEQ COMBINED OFFSET False T guiding ON F guiding OFF Jitter Box Width arcsec SEQ JITTER WIDTH 40 Jitter box size Return to Origin T F SEQ RETURN T Returns the telescope to the original pointing if True Nod Throw arcsec SEQ NODTHROW 60 Nod throw arcsec NINT SEQ NINT 1 Number of exposures in each A or B position Number of AB or BA cycles SEQ NABCYCLES 1 Number of AB cycles Table C 35 SOFLspec obs AutoNodOnSlit Parameter signature Exposure Name DET DIT DET NDIT NSAMP NSAMPPIX Number of columns Number of rows First column of window First row of window Spectro Mode Which Slit Combined offset F T Jitter Box Width arcsec Return to Origin T F Nod Throw arcsec NINT Number of AB or BA cycles Header Keyword DET EXP NAME DET DIT DET NDIT
2. BOB will start again with the first offset It repeats simply the list of offsets you enter until it has the same number of exposures It is possible to randomly jitter around each position of the map within a box whose size is set by SEQ JITTER WIDTH The number of jittered frames at each offset array position is SEQ NJITT and the telescope moves to the next map array position only after they are completed The total number of frames acquired by this template will be the product of the number of array positions by the number of jitters around each one of them i e SEQ NEXPO x SEQ NJITT A modified version of this template is called SOFI_img obs_AutoJitterArray_1 It has identical pa 82 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Parameter signature Header Keyword Value Description Exposure Name DET EXP NAME SOFI File name prefix DET DIT DET DIT NODEFAULT Detector Integration Time individual exposure sec DET NDIT DET NDIT NODEFAULT Number of DITs averaged into an individual image Number of columns DET WIN NX 1024 Number of columns in the window Number of rows DET WIN NY 1024 Number of rows in the window First column of window DET WIN STARTX 1 First column of window First row of window DET WIN STARTY 1 First row of window Number of exposures SEQ NEXPO 1 Number of exposures in the sequence Filter wheel 1 INS FILT1 ID NODEFAULT Filter wheel 1 position Filter wheel 2 INS FILT2 ID NODEFAULT Filter wheel 2 position
3. Flat Field frames for Large Field imaging with broad band filters are taken by the observatory staff as a part of the calibration plan in the morning after the observations in the filters that were used during the night Narrow band flats are responsibility of the users It is also a responsibility of the users to verify the quality of the flats and any other day time calibrations 36 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Figure 3 3 Examples of Special Dome Flat images From left to right lamp off lamp off with mask lamp on with mask lamp on The stability of the flats was studied over a period of about a year and it appears the degradation over time is minimal as can be seen in Fig 3 4 Therefore it s not necessary to take flats every day especially if high precision photometry is not required to achieve the scientific goals of your program However it is advisable to take at least two sets of flats to check their consistency 3 5 T T T T T T T I T T T T T T T 3 L l L be ee 6 2 5 H e e 24 o L e 4 E 7 9 E 4 A 3 2r G F E A L 4 1 5 F 1 E 1 i 1 j 1 1 1 1 I 1 L 1 1 zl 0 100 200 300 Flat Field Age days Figure 3 4 Stability of the Special Dome Flat images accuracy of the photometry as a function of the flat field age 3 5 3 Illumination Corrections It is often the case that the flat field generated from the dome
4. LARGE FIELD IMAGING F 20 T Table C 8 SOFLimg_obs_AutoJitter Example 76 Parameter signature Exposure Name DET DIT DET NDIT Number of columns Number of rows First column of window First row of window Number of exposures Filter wheel 1 Filter wheel 2 Instrument Mode Combined offset F T Jitter Box Width arcsec Return to Origin T F Rotation offsets list degr on Sky SOFI User s Manual 2 3 Header Keyword DET EXP NAME DET DIT DET NDIT DET WIN NX DET WIN NY DET WIN STARTX DET WIN STARTY SEQ NEXPO INS FILT1 ID INS FILT2 ID INS IMODE SEQ COMBINED OFFSET SEQ JITTER WIDTH SEQ RETURN SEQ OFFANGLE Table C 9 SOFL img obs AutoJitterRot LSO MAN ESO 40100 0004 Value Description SOFI File name prefix NODEFAULT Detector Integration Time individual exposure sec NODEFAULT Number of DITs averaged into an individual image 1024 Number of columns in the window 1024 Number of rows in the window 1 First column of window 1 First row of window 1 Number of exposures in the sequence NODEFAULT Filter wheel 1 position NODEFAULT Filter wheel 2 position NODEFAULT Instrument Mode F T guiding ON F guiding OFF 40 Jitter box size T Returns the telescope to the original pointing if True 0 List of rotator offsets on the sky Parameter signature Value Exposure Name NGC6118 DIT 10 NDIT 6 Number of columns 1024 Number of rows 1024 First column of window 1
5. DET NSAMP DET NSAMPIX DET WIN NX DET WIN NY DET WIN STARTX DET WIN STARTY INS SMODE INS WHICHSLIT SEQ COMBINED OFFSET SEQ JITTER WIDTH SEQ RETURN SEQ NODTHROW SEQ NINT SEQ NABCYCLES Value SOFI NODEFAULT NODEFAULT 4 4 1024 1024 1 1 NODEFAULT NODEFAULT False 1 Description File name prefix Detector Integration Time individual exposure sec Number of DITs averaged into an individual image Number of Samples Sample Number per Reading Number of columns in the window Number of rows in the window First column of window First row of window Spectroscopic Mode Which Slit T guiding ON F guiding OFF Jitter box size Returns the telescope to the original pointing if True Nod throw arcsec Number of exposures in each A or B position Number of AB cycles Table C 36 SOFLspec obs AutoNodNonDestr SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 97 position along the slit In addition to nodding random typically small a few tens of arcsec or smaller offsets can be added in the middle of a cycle to locate the spectra at different positions on the array improving the bad pixel removal and the flat fielding A sequence of 4 cycles with jittering will result in the following sequence A B e B e A e2 A e2 B e3 B e3 A eq where ew are the random jittering offsets They are generated withing the interval defined by SEQ JITTER WIDTH in arcsec If SEQ JITTER W
6. DPR TECH IMAGE Exposure in imaging mode DPR TECH SPECTRUM Exposure in spectroscopic mode DPR TECH POLARIMETRY Exposure in polarimetric mode DPR TYPE LAMP Arc spectrum comparison lamp DPR TYPE DARK Dark frame DPR TYPE STD Standard Star DPR TYPE FLAT Flat Field Frame Table D 1 FITS header keywords defining the content of the images 105 Appendix E Photometric Standards The majority of users prefer to calibrate their data with the NICMOS photometric standard stars Persson et al 1998 A J 116 2475 Pre prepared OBs for observations of most of the standard stars from this list are available at the telescope For convenience these standards that are observable from the Southern hemisphere are listed in Tables E 1 and E 2 Finding charts with 2x2 arcmin field of view are shown in Figs E 1 E 2 E 3 E 4 and E 5 North is up and East is to the left OBs for all Person standards that can be observed from La Silla including the extremely red ones are available in a special queue accessible via the Observing Tool Note that under seeing conditions better than 0 6 arcsec the peak of the images for stars with 10 mag at any broad band filter approach the non linearity limit of the detector 10000 ADU even for the minimum DIT This is the case of many red standards listed in Table E 2 Some defocusing via the secondary mirror M3 may be necessary typically shifting the M2 by 0 1 mm in either direction is sufficient We caution aga
7. EUROPEAN SOUTHERN OBSERVATORY Organisation Europ enne pour des Recherches Astronomiques dans l H misph re Austral Europ ische Organisation f r astronomische Forschung in der s dlichen Hemisph re LA SILLA OBSERVATORY SOFI User s Manual Doc No LSO MAN ESO 40100 0004 Issue 2 3 10 05 2012 L 4 Prepared LA O cee ats O ER Name Date Signature Revised De MEME ence ateen wads a ase ek Name Date Signature TURN ee OP TE e ste ads Name Date Signature Name Date Signature Reviewed Vici Pe eae as set de dace Name Date Signature Released TeSa ae he ae agent 10 05 2012 es Name Date Signature SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 Change Record Issue Rev Date 02 05 98 14 08 98 12 11 98 25 02 99 16 08 00 05 11 02 26 04 06 01 03 07 23 05 11 10 05 12 Section Parag affected Reason Initiation Documents Remarks Creation New templates New Grism New IRACE and New Templates Some addition and corrections Merged with the template manual major addition to the data reduction section New Fast Photometry mode new rotating templates General updates general visitor info updated offset plots redone all URLs updated updated guiding options Appendix E web links detector non linearity 111 SOFT User s Manual 2 3 This page was intentionally left blank LSO MAN ESO 40100 0004 Contents In
8. e lt co En LO ceo ceo c co pu G3 E gt e oO on Q 3000 H E ma M J le J El E E J gt i E n 2000 i 7 o A B BS L i 4 de r J pS l E 1000 Bo x lo J lig cits A Loa Es l J hn APA AA 1 1 1 fi fi I Ll 1 1 O L L 1 4 10 6x 104 1 8 104 2x10 2 2 10 2 4x10 2 6x10 Wavelength A Figure A 2 A Xenon arc spectrum taken with the red grism The main lines are marked 63 64 2x105 1 5x10 105 Relative Intensity 5x10 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 T e e EN A op SO oco ec e e ou E a LO Nooo ro on d D LO LO i eo hay D cO n Q LO st eo fon D coco x st Oo co cO fon o co D ez Oc OO cO co co co cO co O e T T OO cO DW co co oO co co co o o co v o o o o o Doo o o gt lt gt lt gt lt aa 4 2 4 lt 4 gt lt gt lt gt lt 8000 8500 Wavelength A Figure A 3 A Xenon and Neon arc spectrum taken with the medium resolution grism at the Z atmospheric window The main lines are marked SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 65 as 4 E 8x10 o 5 co on mun M do op co cQ c dc 5 EM o gt ceo QQ co CQ Ww NO co Sou gt uc co oc co L c e s N o 3 co co D ga c J S e
9. e Signal to noise ratio of 5 computed over 21 pixels e Pixel scale 0 288 arc seconds pixel e 1 hour exposure made up of 60 one minute integrations e Seeing 0 75 arc seconds e Airmass 1 2 e Backgrounds of 16 0 16 0 14 2 and 13 0 in J Js H and Ks respectively The extended source detection limits were made for an aperture with a diameter of 3 arc seconds and for a S N of 4 The values of Table 2 5 can be re scaled to different S N ratios fluxes F and integration times t keeping in mind that for background limited performances S N xFx vt Background limited performances are reached when ADU gt gt 30 assuming read out noise 12 e in the double correlated read out mode and a gain of about 5 4 e ADU NOTA BENE The detection limits given above vary with the background level which is strongly sensitive to the humidity and and air temperature The background can easily change within 5096 2 8 Instrumental Overheads The fraction of time spent not collecting photons is defined as the instrumental overhead A good conservative rule of thumbs states that it is typically about 30 of the integration time There are several sources of telescope and instrument overhead SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 15 e acquisition overhead which depends on the instrument mode the selected acquisition template the brightness of the target the angular distance from the previous target the simple imaging preset takes on avera
10. plate On the other hand if the conditions grow even worse and make it impossible to complete the mapping the observer may be left with a shallow image over the entire area as opposed to a deep image of part of the mapped area The observer must decide which strategy is better suited to the goals of the program before deciding which template to use Consider the following example Table C 21 you will have 4 main positions defined in the RA and DEC offset lists Around each of the offset position you will have 3 images with jitter inside a box of 20 arcsec defined by the parameter Jitter Box Width In total this means 3 x 4 images 12 images Each of them is the average of 10 exposures NDIT of 6sec DIT so the total exposure time is 10 x 6x 3x4 12min without the overheads The template SOFLimg obs AutoJitterArray will take a sequence of 3 jittered images around the first offset position then move to the next offset position and take 3 more jittered images and so on until the entire mapped region is covered The template SOFIimg_obs_AutoJitterArray_1 will instead take one image at each 4 offset positions then it will repeat the entire offset pattern NJITT 3 times adding before each cycle a small jitter offset SOFI_img_obs_GenericImaging This template Table C 22 has some similarity with SOFLimg obs Jitter but it allows the user to do any sequence of telescope offsets whether they be combined offsets offsetting the telescope an
11. 3 3 Polarimetry sn rte y duke A A A A A 3A Spectroscopy orpine A e a ie ee 3 4 1 Small Objects and Uncrowded Fields 3 4 2 Extended Objects and Crowded Fields o 3 4 8 Telluric Standards and Flux Calibration 3 5 Calibration Frames x no lex LAS GP A S PG La eve 3 5 1 Darks Biases nls et deben pb RE eR be bo eed Hee bod 3 02 Plat Bields 2 2 oni oe Wee 39 uex a ia Me a E wat eo a eo 3 5 8 Illumination Corrections o oo o a e e Jib ATOS ce S T AI A see Bnd O a eg E Neck ete a HUS i sia a AL Gee ut iud 3 0 Fimer Points S S i ada n aa oa ee Que box a i a RD dR he tet a AL ed 3 0 h Choosing DIT NDITand NINT a a e 3 60 27 AOE E v 44 ndo SEI a a a aa ELE e N vi 4 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 3 6 53 SOFI Observing Modes es 3 6 4 Template Parameters Signatures and Keywords 3 6 5 File Naming and Exposure Number 3 6 6 Detector WindoW nn Phase 2 Preparation and Observing with SOFI 4 1 General Issue a A a bed sete 4 2 The VLT environment P2PP BOB OS TCS DCS ICS 4 3 Arriving at the Telescope e A3il Image Analysis 2 5 Ls A e Se AAA S s 4 3 2 A A A RR 44 The SOFI OS GUI Panel seres Ao RD sta e bd gle bape A AA A eod 4 6 The Data Flow Path iuta we Eae a o goe A r aa kea es 46 Sofl P
12. First row of window 1 Number of Exposures 6 Filter wheel 1 Kg Filter wheel 2 open Instrument Mode LARGE_FIELD IMAGING Combined offset T F F Jitter Box Width arcsec 20 Return to origin T F T Rotation offsets list degr on Sky 30 Table C 10 SOFI_img obs_AutoJitterRot Example SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 77 Parameter signature Header Keyword Value Description Exposure Name DET EXP NAME SOFI File name prefix DET DIT DET DIT NODEFAULT Detector Integration Time individual exposure sec DET NDIT DET NDIT NODEFAULT Number of DITs averaged into an individual image Number of columns DET WIN NX 1024 Number of columns in the window Number of rows DET WIN NY 1024 Number of rows in the window First column of window DET WIN STARTX 1 First column of window First row of window DET WIN STARTY 1 First row of window Number of exposures SEQ NEXPO 1 Number of exposures in the sequence Filter wheel 1 INS FILT1 ID NODEFAULT Filter wheel 1 position Filter wheel 2 INS FILT2 ID NODEFAULT Filter wheel 2 position Instrument Mode INS IMODE NODEFAULT Instrument Mode Combined offset F T SEQ COMBINED OFFSET F T guiding ON F guiding OFF Jitter Box Width arcsec SEQ JITTER WIDTH 40 Jitter box size Return to Origin T F SEQ RETURN T Returns the telescope to the original pointing if True Sky Offset Throw arcsec SEQ SKYTHROW 300 Radius of the sky offsets region arcsec Pupil Rotation SE
13. Rotate Pupil T F T Table 3 2 The parameters of the template SOFI img obs AutoJitterOffset with typical values In this template one observes alternatively the object and sky There are 36 exposures in total 18 20 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 on the object and 18 on the sky The sky positions are randomly chosen to lie on a circle that is centered on the object The diameter of this circle is set by the parameter Sky Offset Throw Each of the six sky positions is different Each of the six object positions are also different they are randomly distributed within a square box that is centered on the object The box size is set by the parameter Jitter Box Width Auto guiding is set by the parameter Combined offset If selected guiding is done when the telescope is pointing to the object field As the NT T tracks very well and as one usually spends no more than a few minutes on a single position it is not recommended At the end of the template the telescope returns to the original position making it especially conve nient for obtaining a sequence of images in different filters on the same field The parameter Rotate Pupil is an option which allows you to rotate the instrument in between the sky and object positions so that the one obtains better sky cancellation For imaging in filters with high backgrounds that is the Ks and narrow band filters with central wavelengths greater than 2 2 microns
14. arcsec 0 75 0 150 0 Table 3 7 Parameters of the standard star template with commonly used values 3 3 Polarimetry In Polarimetry a Wollaston prism and a mask wheel consisting of alternating opaque and transmitting strips are inserted into the beam The widths of the transmitting sections are about 40 arc seconds whereas the widths of the opaque sections are slightly larger Thus to cover the whole field one needs to take three separate images shifted by 30 arc seconds from each other This techniques is embodied in the template SOFI img obs Polarimetry The parameters for this template are listed in Table 3 8 Parameter signature Value Exposure Name Planetary Disk DIT 1 8 NDIT 10 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Number of exposures 3 Filter wheel 1 J Filter wheel 2 open Combined Offset T F F Return to Origin T F T X offset list arcsec 000 Y offset list arcsec 64 32 32 Rotator Offset 0 Table 3 8 Parameters of the polarimetric template with commonly used values This example takes three exposures in the J band By default the polarimetric mode uses the large field objective Offsets are relative to array rows and columns are independent of the rotator angle and are in arc seconds In this example three exposures with offsets along the Y direction are taken In this way an entire field can be covered Observers may wish to include additional exposures w
15. data reduction of the standards In the next example Table C 29 SOFI_img_cal_StandardStar is used to obtain 5 images the first on the acquisition position and the others defined by the list of offsets At the end you have an x pattern 90 SOFI User s Manual 2 3 Parameter signature Exposure Name DET DIT DET NDIT Number of columns Number of rows First column of window First row of window Number of exposures Filter wheel 1 Filter wheel 2 Instrument Mode Combined offset F T Return to Origin T F X offsets list arcsec Y offsets list arcsec Header Keyword DET EXP NAME DET DIT DET NDIT DET WIN NX DET WIN NY DET WIN STARTX DET WIN STARTY SEQ NEXPO INS FILT1 ID INS FILT2 ID INS IMODE SEQ COMBINED OFFSET SEQ RETURN SEQ OFFSETX LIST SEQ OFFSETY LIST Table C 28 SOFLimg cal StandardSatr Parameter signature Exposure Name DIT NDIT Number of columns Number of rows First column of window First row of window Number of Exposures Filter Wheel 1 Filter wheel 2 Instrument mode Combined offset F T Return to origin T F X offset arcsec Y offset arcsec Value SOFI NODEFAULT NODEFAULT 1024 1024 1 1 1 NODEFAULT NODEFAULT NODEFAULT F T NODEFAULT NODEFAULT Value 9104 1024 1024 1 1 5 J open Large field T T 0 45 90 0 90 0 45 0 90 0 LSO MAN ESO 40100 0004 Description File name prefix Detector Integration Time individual
16. lt c dc oo oo OC o Fo do o a oS a e e e N gt e e5 g sv Qi LO LO LO 1O LN co WwW Wd CO O mow O O co o do co F o o o o o o oO o o wo Dm o o o o 4 Z gt lt gt gt lt gt gt lt gt lt 6x104 ls E hos has cw si Eg al gt n i S q a E e Si c Drs al Ut i o i 2 4x104 4 gt i co e d c mm m 2 Z Z E 2x10 B ads a zm FT S E Or wo ase Ln L L L L 1 1 1 1 1 1 1 L 1 1 1 L L 1 2x101 1 25x10 1 3x10 149355 4104 1 4104 1 45104 Wavelength A Figure A 4 A Xenon and Neon arc spectrum taken with the medium resolution grism at the J atmospheric window The main lines are marked 66 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 2 2 RE a E 2 Q E I e c3 S a c o o c o 3x104 E eo aw LO c e c at is e I DO LO A oO ec Ve LO 10 O co p sont o Cv oo o oa Da v oo Z D x lt x lt x lt x lt Z gt lt Z Z 32x104 E L e o E E o gt Bn E E a 2 o L zd pa 10 L or 1 L 1 L 1 L L 1 1 1 5x10 1 6x10 isTTo 1 8x104 Wavelength Figure A 5 A Xenon and Neon arc spectrum taken with the medium resolution grism at the H atmospheric window The main lines are marked SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 67 8x 104 L eo m o um E MOTO e cam Sr g tT g LO
17. the twilight sky or the night sky itself does not represent completely the low frequency sensitivity variations of the array In other words the dome flat field contains a residual large scale variation that is usually due to the fact that the illumination of the screen can never represent accurately the illumination of the sky In addition the illumination of the screen can change with time due to the variations in the lamps used to illuminate it and due to variation of the screen and dome temperature for gt 2 3 micron This problem can be solved by observing a bright star preferably a standard in a rectangular grid SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 37 pattern of 9 16 positions across the array After sky subtraction and flat fielding with the uncorrected flat the intensities not the magnitudes are fitted with a low order polynomial Next the surface generated by this fit is normalized to one Then the flat is multiplied by this surface to create a corrected flat field Alternatively one can think of this surface as a second flat field and instead of correcting the dome flat the user can divide the data by the regular dome flat and then by this surface This surface is sometimes called an illumination correction Without this correction the low frequency variability from edge to edge of the array is between 1 and 3 depending on the filter the objective and the type of flat used After the illumination correction th
18. 51 11 947 0 008 11 605 0 008 11 560 0 013 11 540 14 23 45 5 84 09 58 11 232 0 007 10 990 0 007 10 904 0 009 10 915 14 40 58 0 00 27 47 12 045 0 008 11 701 0 005 11 622 0 005 11 633 14 56 51 9 44 49 14 11 341 0 007 10 924 0 005 10 851 0 004 10 849 15 39 03 5 00 14 54 10 914 0 008 10 701 0 008 10 649 0 010 10 659 16 26 42 7 05 52 20 12 180 0 007 11 895 0 006 11 842 0 007 11 844 17 27 22 2 00 19 25 11 132 0 005 10 835 0 005 10 739 0 006 10 744 17 48 22 6 45 25 45 12 477 0 009 12 118 0 006 12 026 0 006 12 031 18 18 46 2 80 06 58 11 039 0 007 10 778 0 007 10 693 0 009 10 711 18 28 08 9 69 26 03 12 252 0 006 11 916 0 007 11 834 0 011 11 839 19 01 55 4 04 29 12 10 966 0 007 10 658 0 008 10 566 0 014 10 575 20 31 20 4 49 38 58 12 464 0 011 12 127 0 008 12 095 0 007 12 070 20 41 05 1 05 03 43 11 479 0 005 11 142 0 005 11 082 0 010 11 085 20 52 47 3 06 40 05 12 247 0 004 11 940 0 004 11 873 0 007 11 880 22 02 05 7 01 06 02 12 021 0 005 11 662 0 004 11 586 0 012 11 585 23 18 10 0 00 32 56 11 403 0 009 11 120 0 006 11 045 0 006 11 055 23 23 34 4 15 21 07 11 857 0 003 11 596 0 003 11 538 0 009 11 542 0 005 0 004 0 005 0 005 0 006 0 005 0 002 0 006 0 005 0 005 0 008 0 010 0 005 0 004 0 008 0 009 0 004 0 010 0 006 0 014 0 011 0 010 0 008 0 012 0 010 0 005 0 010 0 004 0 007 0 008 0 008 0 008 0 005 0 004 0 009 0 006 0 005 0 006 0 008 0 007 0 008 0 007 0 005 0 005 0 005 0 006 0 003 T
19. 9 677 0 016 9 668 0 019 cskd 37 12 31 30 1 63 47 12 12 489 0 015 10 415 0 014 9 528 0 016 9 513 0 019 cske 23 12 31 56 0 63 37 43 10 624 0 008 8 762 0 006 7 917 0 008 7 889 0 008 cskf 12 12 31 30 1 63 51 03 29 671 0 008 8 858 0 007 8 561 0 009 8 561 0 008 cskf 13a 12 31 39 5 63 51 03 10 631 0 010 9 861 0 009 9 486 0 011 9 506 0 011 cskf 14a 12 31 45 9 63 49 36 10 095 0 008 8 851 0 007 8 387 0 009 8 361 0 008 T868 53850 15 00 26 4 00 39 29 11 589 0 008 10 993 0 008 10 633 0 009 10 657 0 012 T868 110639 15 10 17 0 02 41 05 12 612 0 009 11 865 0 010 11 336 0 009 11 353 0 011 L134 15 53 38 4 04 39 04 11 408 0 009 9 627 0 010 8 875 0 011 8 898 0 013 Oph N9 16 27 13 3 24 41 34 17 549 0 022 12 389 0 011 9 527 0 014 9 620 0 018 L547 18 51 15 6 04 16 02 11 872 0 010 9 831 0 011 8 888 0 014 8 870 0 016 BRI 2202 22 05 36 0 11 04 27 11 652 0 010 11 083 0 009 10 721 0 013 10 736 0 011 Table E 2 Additional red infrared photometric standard stars Persson et al 1998 AJ 116 2475 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 109 Figure E 1 Finding charts for the photometric standards of Persson et al 1998 I 110 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Figure E 2 Finding charts for the photometric standards of Persson et al 1998 II SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 111 Figure E 3 Finding charts for the photometric standards of Persson et al 1998 III 112 S
20. COMBINED OFFSET SEQ PRESET SEQ SAVE Value NODEFAULT NODEFAULT NODEFAULT NODEFAULT 0 0 0 0 0 0 0 0 0 0 False True False Table C 30 SOFLimg_acq_Polarimetry Header Keyword DET EXP NAME DET DIT DET NDIT DET WIN NX DET WIN NY DET WIN STARTX DET WIN STARTY SEQ NEXPO INS FILT1 ID INS FILT2 ID SEQ COMBINED OFFSET SEQ RETURN SEQ OFFSETX LIST SEQ OFFSETY LIST SEQ ROT OFFANGLE Value SOFI NODEFAULT NODEFAULT 1024 1024 1 1 1 NODEFAULT NODEFAULT F T NODEFAULT NODEFAULT 45 Table C 31 SOFLimg obs Polarimetry LSO MAN ESO 40100 0004 Description Detector Integration Time individual exposure sec Number of DITs averaged into an individual image Filter wheel 1 position Filter wheel 2 position Alpha Offset for sky subtraction arcsec Delta Offset for sky subtraction arcsec Additional tracking velocity in RA arcsec sec Additional tracking velocity in DEC arcsec sec Rotation offset on the sky degrees T guiding ON F guiding OFF T full preset F fine tunning of the pointing T preserve the acq image F no Description File name prefix Detector Integration Time individual exposure sec Number of DITs averaged into an individual image Number of columns in the window Number of rows in the window First column of window First row of window Number of exposures in the sequence Filter wheel 1 position Filter wheel 2 position T guidi
21. First column of window 1 First row of window 1 Number of exposures NODEFAULT Spectroscopic Mode NODEFAULT Which Slit T Returns the telescope to the original pointing if True NODEFAULT List of offsets in X arcsec NODEFAULT List of offsets in Y arcsec NODEFAULT Observation type O object S sky NODEFAULT N no guiding Value SOFI NODEFAULT NODEFAULT 4 4 1024 1024 1 1 1 NODEFAU NODEFAU T NODEFAU NODEFAU NODEFAU NODEFAU LT LT LT LT LT LT B Box To Star S Star To Box DON T USE Table C 39 SOFI_spec_obs_GenericSpectro Description File name prefix Detector Integration Time individual exposure sec Number of DITs averaged into an individual image Number of Samples Sample Number per Reading Number of columns in the window Number of rows in the window First column of window First row of window Number of exposures Spectroscopic Mode Which Slit Returns the telescope to the original pointing if True List of offsets in X arcsec List of offsets in Y arcsec Observation type O object S sky N no guiding B Box To Star S Star To Box DON T USE Table C 40 SOFLspec obs GenSpecNonDestr 100 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 Sky TEL ROT OFFANGLEPA 0 deg This parameter is equivalent to the classical Position Angle PA on the sky i e if you want to align the slit along the major axis of a galaxy with PA 32 d
22. Mode LARGE FIELD IMAGING Combined Offset T F F Return to Origin T F T RA Offset List arcsec 148 0 148 0 DEC Offset List arcsec 0 148 1 148 Table 3 4 Parameters of the SOFL img obs AutoJitterArray 1 template with commonly used values for 4 point observation of a semi extended object mentioned above In fact this example allows some modification the offsets can be reduced from 148 to 110 120 arcsec and the jitter from 20 to 10 arcsec The small jitter is acceptable because during every cycle we take four images at different array locations reducing the effect from the bad pixels the major reason for jittering in this case Note that this is usually not the case with SOFLimg obs AutoJitter and template observations of uncrowded targets where the jitter is needed to produce a sky image for the sky subtraction There is one more albeit simple complication in this observing strategy the offsets are executed before the images are taken so the acquisition must place the target at the location of the fourth image in the cycle This is the position shown in the figure with a dashed circle and marked with zero Finally the thick arrows show the apparent movement of the target on the RTD from one position to another North and East are also marked This example can easily be generalized one can observe a slightly small target at five positions including one at the center of the array In this case NEXPO 5 On the contr
23. PA 0 deg 800L X offsets 0 60 0 120 0 60 arcsec em Y offsets 0 0 60 60 60 60 arcsec o 3 t o 600 8 a A South i e t 63 i 4 S fa 7 9 L North E 400 H 4 E 3 gt 200 r zi East 1 Q asd oaca ca d aca ca d dieu pei 0 200 400 600 800 1000 X px detector coordinates Figure C 1 Positions of the offsets listed in Tables C 42 and C 41 on the detector The number of arc frames is defined by SEQ NEXPO Arc frames can be taken with different DIT and NDIT values defined as lists with SEQ DIT LIST and SEQ NDIT LIST SEQ SPECTROMODELIST defines the spectroscopic mode entered as a list with either B R Z J H K or NB 1 061 standing for the available spectroscopic modes In addition different slits can be called sequentially with SEQ SLIT LIST The allowed values are 1 2 or 0 6 for long slit 1 long slit 2 and long slit 0 6 respectively Finally the arc lamps are entered into SEQ LAMP LIST Valid values are N Xe Ne or B N stands for None this is why there is no need to take separate dark frames but make sure to obtain at least one image with no lamp at each configuration Xe stands for Xenon Ne stands for Neon and B stands for Both i e Xenon and Neon lamps simultaneously Nota Bene Taking images with no lamp the option N has to be specified explicitly in the SEQ LAMP LIST This is responsibility of the user Remember that these images are necessary to remove the bias dar
24. SOFI 0 05 100 100 463 463 NODEFAULT NODEFAULT NODEFAULT 5 T 87 Description File name prefix Detector Integration Time individual exposure sec Number of DITs averaged into an individual image Number of columns in the window Number of rows in the window First column of window First row of window T store detector reads F store reconstructed images T store data in 3 D fits cube F store data in individual fits files Filter wheel 1 position Filter wheel 2 position Instrument Mode Jitter box size Returns the telescope to the original pointing if True Number of exposures in the sequence Event date for timed observations Event time for timed observations Table C 24 SOFLimg obs FastPhotJitt Parameter signature Exposure Name DIT NDIT Number of columns Number of rows First column of window First row of window Turn on Burst mode T on F off Store DITs in a cube T on in individual files F T Filter wheel 1 Filter wheel 2 Instrument Mode Jitter Box Width arcsec Return to origin T F Number of Exposures Date of event YYMMDD set to 0 for no event Time of event HHMMSS set to 0 for no event open LARGE FIELD IMAGING 0 T 1 0 0 Table C 25 SOFLimg obs FastPhotJitt Example 88 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 C 2 3 SOFT Imaging Calibration Templates SOFLimg cal Darks This calibration template Table C 26 produces da
25. This has a number of advantages they would do their own quality control of the calibrations and they will have at their disposal appropriate calibrations if they wish to reduce the data during the night The calibration plan includes e Imaging special dome flats for all filters and imaging modes used e Spectroscopy arcs and flats for each spectroscopic mode and slit combination the flats are accompanied by images taken with lamps off e Polarimetry flats Nota bene Darks are not a part of the calibration plan 4 7 At the End of the Night and at the End of Your Run We will provide you with one set of CDs DVDs tapes as requested in the Data Backup Request Form that the visitors must fill in at the beginning of their run with all the raw data that were acquired during your run If you need more than one set you should make the copies yourself we can provide blank media and we will help you to start with the copying If you want to save reduced data you must do it yourself In that case you should save a copy of your data at the end of the night Do not do it at the end of your run It is common to generate over 2 Gb of data per night with SOFI Leaving the backup until the end of the run may mean that you could miss your plane The raw data is stored in the directories data raw Y Y Y Y MM DD on the machine wg5off At the end of your run you should fill out the End of Mission Report This can be done via any web browser Yo
26. WIN STARTY 1 First row of window Table 3 17 Detector Window Signatures and Keywords Chapter 4 Phase 2 Preparation and Observing with SOFI 4 1 General Issues Observations at the NTT are carried out in classical mode i e there is no support astronomer on the site The observers are expected to arrive a few days before their run begins and to use this time to familiarize with the instrument At all times there are day and night time telescope operators who provide technical support Astronomical support can be obtained by e mail contact with the instrument scientist names and contact details are listed in the instrument web pages well in advance before the observation started because immediate response cannot be guaranteed Limited help on short time scale may be available from the support astronomers at Paranal via phone but it is also not guaranteed Clearly the early preparation for the observations is the key for success Some further hints e The visitors will be given a visitor account to use the computers in the visitors office there are possibilities to link a laptop to the local network WiFi and Ethernet e The visitors are encouraged to call the La Silla safety number officer at 4444 in case of medical or other emergency and to warn a member of the observatory staff for any outdoor activity remember that the closest hospital is a few hours away e The visitors are expected to contact their support telescope o
27. allows the user to take a sequence of jittered images around user defined positions first all positions are imaged ones then a random jitter is added to the entire pattern and it is repeated as many times as necessary Note that the offsets are executed before the images are taken and the offsets are defined along RA and DEC in arcseconds On the figure the images in the sequence are numbered to show the order in which the target will move during the observations The zero number is the location of the target acquisition and it will be discussed further The parameters of both templates are listed in Table 3 4 Note that for 4 point observation the NEXPO parameter must be equal to four Therefore the total integration per position is controlled by the time spent on each image DIT xNDIT usually 1 3 minutes and by the number of jittered images taken at each individual position NJITT In this example the total integration time is NJITxNEXPxNDITxDIT 12x4x10x6 2880 sec 22 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 0 100 y Nn n D E s 3 lt p 200 F East J l J l NES QN l l l 200 100 0 Dec aresec Figure 3 1 Example of 4 point observation scheme of a semi extended objects without extra overhead for observation of clear sky Remember than although the target is moving on the RTD it is really the telescope that is moving In general the template allows to define manua
28. are available on the SOFI web page http www eso org sci facilities lasilla instruments sofi tools SofI Pipeline html 4 6 2 The Archive All data taken at the NTT are archived into the ESO archive For more retails see http archive eso org 4 6 3 The Calibration Plan The calibration plan aims to ensure that all data stored in the archive can be fully calibrated regardless of the specific needs of the program Indeed it is easy to imagine a situation that the 50 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 data may be used later after their proprietary period expires from the archive for the purpose of another object The calibration plan ensures that any limitations in that respect will not come from the lack of appropriate calibrations Therefore daily calibrations are taken as a matter of routine both by the observatory support staff and in Visitor mode runs by the observers wishing to calibrate their own data The standard proce dure is to run every morning after the end of the night time observations a special tool calob Built that inspects all files taken during the night and determines what calibrations are necessary Furthermore this tool prepares a calibration OB that can be loaded directly into BOB and executed However Visitor mode observers are encouraged not to rely on our calibration plan to provide them with these data but to take their own calibrations in the afternoons before their observations start
29. are meant only for quick and rough data quality assessment A proper reduction can be carrier out with the pipeline recipes which are available via Gasgano and esorex on the off line machine wg5off in interactive regimen The user is also provided with number of tools for data reduction i e MIDAS and IRAF scripts They are described in the SofI web page at http www eso org sci facilities lasilla instruments sofi tools reduction index html see in particular the Useful Scripts link The pipeline recipes carry out the following reduction steps e Imaging cross talk removal sky subtraction flat fielding alignment and combination of images specified in a list provided through Gasgano or esorex observations with large offsets at clear sky fields are not reduced e Spectroscopy sky subtraction flat fielding alignment and combination of the 2 dimensional spectra subtraction and wavelength calibration of an 1 dimensional spectrum of the brightest object Naturally this recipe works best for single bright objects the fainter the object the higher the chance of failure e Zero Points cross talk removal sky subtraction flat fielding simple aperture photometry Note that the tool is sensitive to the position of the star on the array and if it is not centered within 20 arcsec it may fail to recognize the standard The pipeline can be downloaded from http www eso org sci software pipelines The most recent instructions how to use it
30. bapor tine tii it Omer Conte DIT rec if iiit ex bi wort i y 5 Mere XY ve TD NS 171 Figure 4 1 OS of SOFI the stored images from the following ones You ll probably have to click on the button labeled Auto Cuts or edit the cut levels directly to see the image Clicking on Store Fixed Pattern again will load a new fixed pattern Naturally it is best to store a fixed pattern before the telescope is centered on your science target Acquisition templates Move to take an image away from the target at an offset position defined by the user in the template To expedite the process one can even store a fixed pattern during the movement of the telescope first make sure the DIT and NDIT have been updated to the values in the acquisition template but this may lead to dark negative lines instead of dots on the subtracted image The RTD can display either each individual frame or it can display the average of NDIT frames INT mode The option is selected from the button labeled DIT or INT respectively The RTD can also display the image zoomed or reduced by an integer factor clicking on the button to the right of label Scale will rise a pop up menu with the available options The menus on the top left hand corner of the RTD can be used to activate sub windows with various tools which the observer may find useful 48 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 SOFI Bed Tinin Piaf 2 4 Bii Fi
31. before the taking the images Guiding is only possible for the object frames if SEQ COMBINED OFFSET is set to True T must be done by the observed during the preparation of the Observing Block and guiding is started before the template starts done by the operator after warning from the observer Compare the following example Table C 19 with the example for SOFLimg obs AutoJitterOffset template the images are now taken in a user defined sequence instead of in a random jittering pattern The first offset 0 0 is executed before taking the first image the second offset 5 5 is executed before taking the second image and so on The number of exposures corresponds to the TOTAL number of exposures Each image is the average of 10 NDIT exposures of 6 sec DIT No guiding is used SOFLimg obs AutoJitterArray and SOFI img obs AutoJitterArray 1 These templates Table C 20 are very similar to SOFLimg obs Jitter They define a sequence of offsets in RA and DEC but in addition to that they allow to jitter randomly around each of the defined offset array positions They are useful to build maps of large areas of the sky SEQ OFFSETALPHA LIST and SEQ OFFSETDELTA LIST are used to define the map array on the sky The offsets between different positions are relative and they do not have to be equidistant i e irregularly shaped maps arrays are acceptable If you have less offsets than number of exposures defined by the parameter SEQ NEXPO
32. by the desired amount Finally the operator is given the possibility to redraw the arrow for refining the position of the target if necessary Once the operator is satisfied the template finishes Nota Bene There is no way of checking if the slit that is chosen in the Acquisition Template and the slit that will be used subsequently in the Observation Description Templates are the same It is therefore of utmost importance that the astronomer ensures that they are the same Since the spectroscopic modes use the large field objective by default the instrument mode is not a parameter of the template Guiding is possible defaulted to the Telescope Control System s option Boz To Star only if SEQ COMBINED OFFSET is set to T this must be done by the observed during the preparation of the Observing Block and if guiding is on before the start of the science templates this will be done by the operator after warning by the observer before starting the execution of the Observing block The experience has shown that the guiding is necessary if the observer plans to stay at one pointing more than 15 min This is rarely the case in imaging mode but can often happen during spectroscopic observations The SEQ PRESET parameter allows the observer to use the fine tunning pointing functionality of the SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 95 Parameter signature Header Keyword Value Description DIT DET DIT NODEFAULT Detector Integration T
33. consecutive A A or B B positions there is no telescope offset but one can include random offsets between consecutive A B or B A positions In other words every A B or B A cycle can be shifted with respect to the previous The size of the region along the slit in which the random offsets are placed is defined by the parameter Jitter Box Width In the example given here the 1st and 2nd A positions will differ Additionally the 1st and 3rd B positions will also differ In principle the nodding frequency should be similar to that used during direct imaging However the desire to reach background limited performance forces longer exposures in the spectroscopic modes SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 31 Parameter signature Value Exposure Name Hubble Deep Field DIT 60 NDIT 3 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Spectro Mode LONG SLIT RED Which Slit long slit 1 Combined Offset T F T Jitter Box Width arcsec 10 Return to Origin T F T Nod Throw Along Slit arcsec 60 NINT 3 Number of Cycles 4 Table 3 10 Parameters of the template SOFI_spec_obs_AutoNodOnSlit with commonly used values As a general rule one should limit the time spent at any one position to less than 15 minutes Although guiding is optional we recommend that you guide To do so set the parameter Combined offset to true The Non Destructive Read Out is recommended for f
34. exposure sec Number of DITs averaged into an individual image Number of columns in the window Number of rows in the window First column of window First row of window Number of exposures in the sequence Filter wheel 1 position Filter wheel 2 position Instrument Mode T guiding ON F guiding OFF Returns the telescope to the original pointing if True list of offsets along rows list of offsets along columns Table C 29 SOFLima obs StandardStar Example SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 91 C 3 SOFI Polarimetric Template C 3 1 SOFT Polarimetric Acquisition Template SOFT_img acq Polarimetry This acquisition template Table C 30 is very similar to the SOFLimg acq MoveToPixel template The polarimetric mask is displayed on the RTD and is superimposed on the image of the field Then the operator is prompted to define an offset by drawing an arrow on the RTD holding the left mouse button The offset fine positions the object into the transparent region of the mask Since the POLARIMETRY mode uses the large field objective the instrument mode is not a param eter of the template C 3 2 SOFT Polarimetric Science Template SOFLimg obs Polarimetry This observation template Table C 31 is used for polarimetry It is quite simple as it works as an imaging template The number of exposures must correspond to the number of the offsets specified in SEQ OFFSETX LIST and SEQ OFFSETY LIST Note that unlike m
35. line Time spent doing image analysis is not spent on your source thus image analysis should be used wisely It usually takes 10 to 20 minutes to perform an image analysis and verify the result Once done the telescope will be at the correct focus It is no longer necessary to check the focus after image analysis The image analysis is done by the TIO and the only action a visitor can take is to request an image analysis of the image if the quality is degrading How frequently one does an image analysis depends on the required image quality Here are some hints on how frequently it should be done i If image quality needs to be better than 1 0 image analysis should be done if the altitude changes by 20 degrees sometimes 30 degrees trust the advice of the telescope operator For long integrations where the source is changing altitude one should perform the image analysis at an altitude that is 20 degrees higher if the source is rising or 20 degrees lower if the source is setting The analysis should be repeated once the altitude of the object differs from the altitude where the previous image analysis was done by more than 20 degrees In this way the object is never observed more that 10 degrees from the last image analysis ii If the image quality is not critical this is usually the case for spectroscopic observations then perform the image analysis once at the beginning of the night It should be done with the telescope at an altitude wher
36. overheads this template is used without telescope guiding and it does not verify if the telescope is guiding or not If the DET BURST MODE parameter is set to True then the individual reads are stored as opposed to False when the reconstructed images i e the difference between the final and the initial detector reads are stored The DET STORE CUBE determines how to store the data if it is set to True the observing software gen erates a single data cube in which every layer is a 1 dimensional image The cube contain NDIT layers if DET BURST MODE is set to False and 2xNDIT if DET BURST MODE is set to True If DET STORE CUBE is set to False then NDIT or 2xNDIT individual fits files are generated respectively The fastest photometry and the most convenient data handling is achieved with both DET BURST MODE and DET STORE CUBE set to True The DET WIN STARTX and DET WIN STARTY are ignored because by default the usable window is centered on the detector to speed up the readout this allows to use simultaneously the four analog to digital converters Finally for the purpose of occultation observations the template can perform timed observations that will start at a moment defined by the parameters SEQ EVENT DATE and SEQ EVENT TIME in UT To forfeit this option set both of them to zero and the observations will start at the moment the template is executed by the operator Nota bene This is preliminary version of the template
37. parameter is true T the telescope at the end of the template moves back to where it was at the start of the template In general auto guiding is not required because of the frequent telescope offsets For a more detailed discussion about guiding options and the algorithm used to compute the offsets please refer to section 3 6 If you wish to enter the offsets individually use the template SOFI img obs Jitter This template is discussed fully in Appendix C 3 2 3 Large Objects or Crowded Fields For objects larger than 40 of the field or for very crowded fields it is necessary to image the sky and object separately Unfortunately it is common that the sky frames will contain other objects and it is not uncommon that one of these objects will be in the same region of the array as the science object To avoid this it is standard to obtain several sky images on different locations usually in the context of object sky pairs This technique assumes that the sky fields are sufficiently uncrowded SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 19 Parameter signature Value Exposure Name Hubble Deep Field DIT 10 NDIT 6 Number of exposures 60 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Filter wheel 1 K Filter wheel 2 open Instrument Mode LARGE FIELD IMAGING Combined Offset T F F Jitter Box Width arcsec 40 Return to Origin T F T Table 3 1 Parameters of the
38. raw images are displayed on the RTD Real Time Display The results are also sent to the archive machine and the off line workstation wg5off This will be discussed in more detail later When creating OBs it is useful to keep them as simple as possible Do not create complex OBs which switch from one mode to another OBs with multiple science templates with different filters imaging the same target are acceptable 44 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 4 3 Arriving at the Telescope The NTT is run by a section of the joint control room the former library located in the building just bellow the hotel Consult the map on the La Silla web page to find out where it is located Please be aware that you will be sharing the control room with the observers at two other telescopes and some understanding and patience may be required When you arrive at the control building you will be confronted by a vast number of terminals Most of these are of no concern to you You will be primarily interested in the terminals at the right end of the room NTT section Starting from right to left there is an empty table with a network connection that can be used to plug in a laptop The first desktop computer next to it is wg5off where you can run RAF MIDAS eclipse IDL web browsers and other useful programs This is a dual monitor machine From here you can access and examine your data Note that the system clocks of these computers are set to the
39. sky sampling is rarefied However in most of the cases these is a small price to pay for 20 30 more data The users must read carefully Section 5 2 4 where the sky SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 25 Parameter signature Value Exposure Name SOFT 6 point DIT 6 NDIT 10 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 NJITT 12 NEXPO 6 Jitter Box Width arcsec 10 Filter wheel 1 K Filter wheel 2 open Instrument Mode LARGE FIELD IMAGING Combined Offset T F F Return to Origin T F T RA Offset List arcsec DEC Offset List arcsec 0 0 120 0 0 120 600 600 210 600 600 210 Table 3 5 Parameters of the SOFL img obs AutoJitterArray 1 template with commonly used values for 4 point observation of a semi extended object subtraction is described 3 2 6 Faint Objects Around Bright Objects The typical scenario that fits this situation is when one observes a faint companion of a bright star As the luck will have it the larger the contrast between the two objects the more appealing the science case for such observations The bright star dictates a choice of strategy with minimal DIT and numerous frequent jitter offset to minimize the artifacts rising from the electronic cross talk effect that can not be corrected satisfac torily in case if the star is heavily saturated Such a strategy comes with a significant disadvantage with the minimum DIT 1 182 sec t
40. sky the list must include only the 10 closest sky images b Scale these 10 images to a common median To first approximation this is the sky image However it is possible to improve it One possibility is simply to reject the most deviating values i e rejection of the highest and the lowest 1 or 2 values 3c rejection etc We strongly encourage the user to experiment here Another option is a more sophisticated 2 iteration rejection After making a zero order sky as described above return to the list of the images you used to create this sky From each of them subtract the sky The result will be images with nearly zero background and only the stars above the background Study the distribution of pixel values with image histograms and determine the border value between the background and the stars Usually the limit is 3 times the background standard deviation Next mask out the values above this limit The easiest way to do it is just to replace the values above the chosen limit by a high or low value i e 1e5 to 1e5 so when you combine the masked images to produce a second iteration sky these pixels will remain out of the range of the good values used in the combination The good values are usually defined as parameters of the combining task Be careful before modifying the files with replacement It is better to copy them into dummy files first c Scale the sky image to have the same median as the image from which it
41. supported Your data is stored as FITS files in the directory data raw Y Y Y Y MM DD where YYYY MM DD is the date at the beginning of the observing night We suggest that you do all your reductions in this directory e At the beginning of your first observing night fill in the Data Backup Request Form available from the La Silla web page Note that you are requested to press the submit button twice first after you have filled in the form and the second time after you have verified the data Based upon the information in this form the observatory staff will prepare a data package for you at the morning after your last observing night It is important to fill in the correct postal address we may need it in case we have to contact you or mail you some data later The data package includes only the raw data Observers are responsible for saving their own reduced data CD and DVD writers are available for this purpose e This machine runs the important dataSuscriber program responsible for transferring the data from the instrument work station to the archive and to the off line machine Please make sure NOT to stop it SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 45 e There is a number of quick look data reductions tools available on wg5off for imaging and spectroscopic observations for photometric calibration The are described in the instrument web page On the terminal wg5dhs do the following e If required login to this termin
42. the SOFI Users manual The intensity of the dome lamp is controlled semi automatically the template maintains a database and once it is populated for a given configuration the lamp intensities are taken from there The parameter sets for both templates are identical SOFLimg cal StandardSatr This calibration template Table C 28 is intended for standard star observations in one filter If a standard has to be observed in a few different filters this template has to be called as many times as necessary with the appropriate filter defined in each template call The telescope can be offset to position the object at different positions on the array Note that Parameter signature Header Keyword Value Description Exposure Name DET EXP NAME SOFI File name prefix Number of columns DET WIN NX 1024 Number of columns in the window Number of rows DET WIN NY 1024 Number of rows in the window First column of window DET WIN STARTX 1 First column of window First row of window DET WIN STARTY 1 First row of window Number of exposures SEQ NEXPO 1 Number of exposures in the sequence List of DITs SEQ DIT LIST NODEFAULT Detector Integration Time individual exposure sec List if NDIT SEQ NDIT LIST NODEFAULT Number of DITs averaged into an individual image Table C 26 SOFLimg cal_Darks SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 89 Parameter signature Header Keyword Value Description Exposure Name DET EXP NAME SOFI File name pre
43. the VLT is only possible if frames are accurately classified because the images taken via the OS panel have incomplete headers The OS GUI is used only for trouble shooting 4 5 RTD The SOFI RTD Real Time Display Figure 4 2 is a versatile tool of frequent use As the array is continuously read out images are continuously displayed on the RTD During acquisition or when the instrument is idle the data is still taken with the last used DIT and NDIT but it is not written to disk A very useful feature of the RTD is the ability to store a frame into memory buffer and to subtract that frame from incoming frames In effect this feature allows a simple sky subtraction This is essential for recognizing faint objects on backgrounds with significant structure The stored frame is called a fixed pattern and to activate this feature first click on the button labeled Store Fixed Pattern on the RTD screen and the current image will be stored Next click on ON OFF to subtract SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 47 Si eer ewe Cart MBA 5 15 18 825 1 rN Cum Expo Fite Bo le 14354 TYATE OUBS TATE Coserver DARK sort esa ONLINE Usar Type ta MAGE Mew Dats rta wa p sori 6642 IDLE NSLEER SYYTEN WCS W Seq Marirg 100 ae m Aren Meam ente TCS Wade rca i mn wetwt X are fr QOJ eene u MES B TCS eem were Epose Cort Done Pae Ra CEE 00 00 00 owe TT Lamas Frase
44. the focal plane Therefore in a single exposure at least half the field will be missing So three exposures with telescope offsets in between are required to cover one field The Wollaston prism is not achromatic so the exact separation between the two beams is a function of wavelength At J the separation is 48 3 arc seconds while at Ks the separation is 47 4 arc seconds The beam separation is also a function of position To measure the Stokes parameters and hence the degree and position angle of polarization a second set of images with the Wollaston prism rotated 45 degrees with respect to the first pair are required This is achieved by rotating the entire instrument The Stokes parameters are then determined as follows I 90 4 0 i 45 i 135 Q i 0 i 90 U i 45 i 135 where i a is the intensity of the source which transmits light that is polarized at angle a We have assumed that the rotator is at a position angle of 0 degrees for the first measurement This need not be the case The degree of linear polarization and the polarization angle are given by 10 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 U Q P I U 0 0 5 x tanta Q To derive the correct value of 0 attention needs to be paid to the signs of U and Q This algorithm neglects instrumental polarization Preliminary measurements indicate that the in strument polarization is 2 As it is caused by the tertiary mirror the v
45. the rotator by a given offset The offset is relative with respect to the current position The user is strongly advised to check if the rotator has sufficient range to complete the observations and to monitor the rotator angle variation during the execution of the template The offset parameter is a list the default value is list of one zero the valid range is from 360 to 360 degr SOFI_img_obs_Jitter Offset SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 81 Parameter signature Value Exposure Name NGC6118 DIT 20 NDIT 3 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Number of Exposures 5 Filter wheel 1 J Filter wheel 2 open Instrument Mode LARGE FIELD IMAGING Combined offset F T T Return to origin T F T RA offset list arcsec 0 50 0 100 0 DEC offset list arcsec 0 50 100 0 100 Rotation offsets list degr on Sky 30 Table C 17 SOFLimg obs JitterRot Example This observation template Table C 18 is very similar to SOFL img obs AutoJitterOffset The only difference is that the user fully specifies the offsets in SEQ OFFSETALPHA LIST and SEQ O0FFSETDELTA LIST instead of executing randomly distributed offsets It is assumed that offsets are given so that the telescope alternates between object odd numbered frames and sky The object frames are identified as SCIENCE and the sky frames are identified as OTHER The telescope offsets are executed
46. to the minimum DIT of 1 182sec for 10mag stars The users may even have to consider splitting their observations into shallow and deep sequences optimized for different magnitude ranges These issues are discussed again in Sec 3 6 3 2 2 Small Objects or Uncrowded Fields If the object of interest is an uncrowded stellar field it is not necessary to take separate sky obser vations In this case one can dither within the field and use the object frames to create sky frames As a rule the offsets should be greater than 10 arc seconds and if very deep exposures are required the offset vector should not be replicated For example an offsetting scheme that is based on a rectangular grid of points will in a deep exposure show faint negative images arranged symmetrically around each real image As a last resort these can be minimized with appropriate masking during the data reduction Similar to the uncrowded fields small objects can be observed in the same manner In this case the user should make sure that the offsets are at least 2 3 times bigger than the size of the object A pseudo random offsetting scheme see sec 3 6 is used in the template SOFI img obs AutoJitter The offsets are restricted to be within a square box centered on the object The dimension of the box is defined by the parameter Jitter Box Width This and other parameters of the template are listed in Table 3 1 If the value of the Return to Origin
47. using the minimal DIT for a given window size is given in Table 3 6 dx dy minDIT Tesec px px sec sec 1024 1024 2 1033 2105 512 1024 1 2645 2104 1024 512 1 0519 1053 512 512 0 6324 633 128 512 0 3179 318 512 128 0 1584 158 128 128 0 0798 79 64 128 0 0667 66 128 64 0 0401 40 64 64 0 0336 33 32 64 0 0303 30 64 32 0 0170 17 32 32 0 0153 16 16 32 0 0145 16 32 16 0 0079 17 16 16 0 0075 33 Table 3 6 Minimal DIT and time to complete a series of NDIT 1000 for different window sizes The readout mode was RdRstRd Currently the mode is offered only in Visitor Mode because it typically require real time decisions by the user Members of the community interested in using the new mode are strongly encouraged to contact the Sofl Instrument Scientist in advance for further details Nota bene The FastPhot mode is not fully commissioned More details will be made available on the Sofl web page 3 2 8 Standard Stars The IR window between 1 and 2 5 microns contains several large absorption features that are primar ily due to water vapor and carbon dioxide in the atmosphere The edges of the atmospheric windows are highly variable Unfortunately the edges of some IR filters particularly J and K are defined by these absorption features rather than the transmission curves of the filters themselves Thus when the column density of water vapor is variable accurate photometry can be difficult to achieve On good nights generally when
48. whose values are 4c above the mean value e Noisy Pixels they are computed from the histograms of the sigma image obtained when averaging the flats and the darks of the previous 2 steps e Frame Pixels the mask also marks those pixels as bad that are masked by the edges of the focal plane field mask SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 53 NOTA BENE The masking must occur before the images are shifted and aligned It is up to the user how to exclude the masked pixels Perhaps the simplest way is to use two masks The first one will have zero values at the bad pixels and one at good ones Every raw data frame has to be multiplied by this mask ensuring that all bad pixels will have the same value The second mask will have zero values at the good pixels and a large negative value i e 1e6 at the bad values This mask will have to be added to the row frames after the multiplication It will set the bad pixels to a value outside the good pixel values for SOFI Of course the fractional offset during the image alignment may move some of this artificial negative flux into the acceptable values but the large value makes sure that this will happen only rarely The final step of the application of the bad pixel mask is to set an appropriate lower threshold during the combination of the aligned images Indeed some data reduction tools like IRAF have the capability to incorporate a bad pixel mask and the shifting of the ima
49. width by inaccurate centering of objects on the slit In general this problem can be remedied later during the data reduction by re aligning of the spectra first by using the telluric features to determine the offset in wavelength direction and second by geometric transformation of the image in spatial direction However these are complicated and time consuming steps that involve heavy modification of the data prior to any scientific measurements and it is preferable to avoid them if possible The division by the telluric standard a number artificial emission features introduces into the target spectrum because of the intrinsic absorption features of the telluric standard itself By choosing carefully the spectral type of the telluric standards one can minimize the number of such features but there are no stars with completely featureless IR spectra and therefore some such emissions will always remain To remove them after the division by the telluric standard one should multiply by the absolute spectral energy distribution of the telluric standard However this information is usually not available so one has to model the spectral energy distribution of the standard For hot stars a black body fit to the star may be appropriate but for stars of spectral type later than B a more accurate description of the energy distribution may be required Another possibility is to use a solar analog for telluric standard taking advantage o
50. will be subtracted and then carry out the subtraction This method reduces the negative traces of stellar images on the sky subtracted data The number of 10 images used to create the sky is somewhat arbitrary It can be 5 7 or even all available images ii subtracting from each image the average of the previous and the next image In the more general case the average or the median of the previous 2 3 and the following 2 3 images is subtracted Strictly speaking this is a specific case of the method described above However our experience shows that in most cases using only the two images nearest in time is sufficient The advantage of this technique is that it takes into account sky variation on time scale a few times NDIT x DIT In the specific case of SOFI it also removes very well the pupil ghosts usually observed as diagonal stripes in the South East corner of the array discussed earlier Both these advantages are pronounced better if the number of images we use to subtract the sky is kept small ideally just the preceding the the succeeding images While the reason for this is obvious for sky variations it is not the case with the pupil ghost until one remembers that the pupil rotates and the ghost moves across the field of view Therefore it is best to sky subtract using images that have the ghost nearly at the same position as on the target image The problem with this method is that some negative residuals fr
51. will not remove the shading pattern perfectly T his issue is discussed in more detail in the next section The experience has shown that the best way to remove the dark current is by taking data in exactly the same way and with exactly the same exposure time as the science observations This is routinely done for the removal of the sky emission so the removal of the dark is naturally removed when the sky is removed The only case when the removal of the dark remains an issue is the flat fields see the next section for discussion Darks appear to be stable over the period of a typical observing run Thus it is sufficient to take darks only once during the run or more often as a verification of the detector array status Nota bene Darks are not part of the calibration plan for the reasons described above If you still need darks prepare your own calibration OBs 3 5 2 Flat Fields As in the case of visible observations one must correct for differences in pixel sensitivities The method for creating flat fields is identical in both the spectroscopic and imaging modes They are created by exposing alternatively an illuminated and an unilluminated dome panel The flat field images are constructed from the difference of the two normalized to one This technique is especially important when working beyond 2 3 microns where it removes the thermal component of the signal which does not depend on the intensity of the flat field lamp Spectros
52. 41 on the detector 102 Finding charts for the photometric standards of Persson et al Finding charts for the photometric standards of Persson et al Finding charts for the photometric standards of Persson et al 1998 II 111 Finding charts for the photometric standards of Persson et al Finding charts for the photometric standards of Persson et al Chapter 1 Introduction SOFI or Son OF ISAAC is the infrared spectrograph and imaging camera on the NT T In many ways it resembles its parent ISAAC They both are focal reducing instruments capable of imaging spectroscopy and polarimetry SOFI offers the following observing modes Imaging with plate scales of 0 144 0 144 0 273 and 0 288 arc second per pixel in the following modes Small Field decommissioned Large Field Focal Elongator decommissioned Spec troscopic Field and Large Field respectively broad and narrow band filters in the wavelength range from 0 9 to 2 5 microns are available Fast imaging and burst mode with detector integration times of order of a few tens of milisec onds with significantly reduced detector array size i e of order of few tens pixels on the side via hardware windowing e Low resolution R 600 varies across the wavelength range 0 95 2 52 micron spectroscopy with fixed width slits of 0 6 1 and 2 arc seconds and slit length of 4 92 arcmin e medium resolution R 1500 varies across the wavelength ran
53. 78 S808 C 9181 234 E 9182 S813 D 9183 P576 F 9185 S889 E 9186 S893 D 9187 S677 D 05 42 32 1 00 09 04 11 426 0 009 11 148 0 009 11 077 0 014 11 058 06 22 43 7 00 36 30 11 723 0 011 11 357 0 009 11 264 0 016 11 261 06 29 29 4 59 39 31 12 114 0 006 11 838 0 005 11 765 0 009 11 781 06 42 36 5 45 09 12 12 719 0 004 11 434 0 004 11 372 07 19 38 6 84 35 06 10 885 0 007 10 598 0 006 10 514 0 013 10 522 08 01 15 4 50 19 33 10 914 0 007 10 585 0 006 10 487 0 021 10 496 08 25 36 1 39 05 59 11 949 0 006 11 669 0 005 11 608 0 004 11 609 08 27 12 5 25 08 01 11 521 0 007 11 048 0 008 10 965 0 016 10 960 08 29 25 1 05 56 08 11 881 0 007 11 624 0 005 11 575 0 005 11 596 08 36 12 5 10 13 39 12 362 0 010 12 098 0 011 12 040 08 54 21 7 54 48 08 12 489 0 008 12 214 0 008 12 138 0 018 12 142 09 15 50 5 36 32 34 11 153 0 007 10 891 0 007 10 830 0 019 10 836 09 41 35 8 00 33 12 11 354 0 006 11 041 0 006 10 981 0 015 10 982 09 45 42 8 45 49 40 11 409 0 011 11 085 0 008 11 022 09 48 56 4 10 30 32 11 081 0 008 10 775 0 008 10 715 0 035 10 718 10 33 51 8 04 49 05 12 344 0 007 12 121 0 005 12 067 0 006 12 081 10 47 24 1 44 34 05 11 642 0 009 11 335 0 008 11 263 0 018 11 280 12 01 45 2 50 03 10 12 323 0 007 11 002 0 005 10 931 0 003 10 936 12 03 30 2 69 04 56 12 111 0 007 11 803 0 007 11 722 0 013 11 724 13 17 29 6 05 32 37 11 661 0 008 11 310 0 007 11 250 0 014 11 267 14 07 33 9 12 23
54. B Box To Star S Star To Box DON T USE Table C 22 SOFLimg obs GenericImaging At the end of the template the telescope is returned to the original position if SEQ RETURN is set to True T If not the telescope is not moved at the end of the template Consider the following example Table C 23 It contains 7 offsets with the option to guide Boz To Star the only other option is No Guiding Star To Box has been disabled The object is always observed with a NDIT equal to 3 whereas the sky is always observed with NDIT equal to 2 At the end there are 4 images for the object each of them is the average of 3 NDIT expositions of 20 sec DIT And 3 images of the sky each of them is the average of 2 NDIT expositions of 20sec DIT SOFI img obs FastPhot Jitt This observation template Table C 24 takes a fast photometry sequence It has built capability to apply a random jitter within a predefined box on the sky and to repeat the fast photometry sequence It offsets the telescope between bursts according to a random pattern of offsets automatically gen erated within the template However the implication when using this template is that the detector will be heavily windowed to minimize the minimal DIT Thus usually the jitter box size will be extremely small and placing the target on a detector location clear of bad pixels may be a better strategy Therefore it is recommended to set the jitter box size to zero The offsets are d
55. IDTH is set to zero then the template will just nod between A and B It is recommended to use a jitter box width smaller or equal to the nodding length to avoid any chance of overlapping of the spectra on sequential images If SEQ COMBINED OFSET is set to T and if guiding was started before the start of the template the telescope will guide with Boz To Star Guiding is recommended for observations longer than 15 min but close to the meridian or the zenith it may be required even for shorter sequences The templates have been optimized to minimize overheads However after a telescope offset one or more DITs are skipped to make sure the telescope has stabilized after the move Otherwise when the telescope is not moved between frames i e when NINT gt 1 or at the beginning of a new cycle all DIT s are kept At the end of the templates the telescope returns to the original position if SEQ RETURN is true T If not the telescope is not moved at the end of the template SOFI spec obs AutoNodOnSlit and SOFI spec obs AutoNodNonDestr differ only in the control over the readout mode The second template allows to define NSAMP and NSAMPPIX as explained the section 3 4 1 Otherwise they are identical The only two additional parameters with respect to the SOFI_spec_obs_AutoNodOnSlit are DET NSAMP and DET NSAMPIX respectively the number of samples and the sample number per reading During the integration the signal is sampled a number of times
56. Instrument Mode INS IMODE NODEFAULT Instrument Mode Combined offset F T SEQ COMBINED OFFSET F T guiding ON F guiding OFF Return to Origin T F SEQ RETURN T Returns the telescope to the original pointing if True RA offsets list arcsec SEQ OFFSETALPHA LIST NODEFAULT list of offsets along RA Dec offsets list arcsec SEQ OFFSETDELTA LIST NODEFAULT list of offsets along Dec Pupil Rotation SEQ ROTPUPIL T T rotate N not rotate Table C 18 SOFLimg obs JitterOffset Parameter signature Value Exposure Name NGC6118 DIT 6 NDIT 10 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Number of Exposures 5 Filter wheel 1 Ks Filter wheel 2 open Instrument Mode LARGE FIELD IMAGING Combined offset T F Return to origin T F RA offset list arcsec DEC offset list arcsec Rotate Pupil 10 2 5 5 5 9 10 2 5 5 doodmu Table C 19 SOFLimg obs JitterOffset Example SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 83 Parameter signature Header Keyword Value Description Exposure Name DET EXP NAME SOFI File name prefix DET DIT DET DIT NODEFAULT Detector Integration Time individual exposure sec DET NDIT DET NDIT NODEFAULT Number of DITs averaged into an individual image Number of columns DET WIN NX 1024 Number of columns in the window Number of rows DET WIN NY 1024 Number of rows in the window First column of window DET WIN STARTX 1 First column
57. LT MAN ESO 14100 1510 OS Users Manual VLT MAN ESO 14100 1531 DCS Users Manual VLT MAN ESO 14100 1094 ICS Users Manual VLT MAN ESO 00000 000 1 1 P2PP Users Manual VLT MAN ESO 00000 000 SOFI TSF Parameters Reference Guide obsolete Abbreviations and Acronyms The abbreviations and acronyms used in this manual are described in Table 1 1 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 Acronym BOB DCR DCS DEC DIT EFOSC2 EMMI ESO FWHM ICS IR IRACE ISAAC NDIT NDR NINT NTT OB OS P2PP PSF RA RON SOFI TCS TSF USD VLT ZP Description Broker of Observing Blocks Double Correlated Read Detector Control System Declination Detector Integration Time ESO Faint Object Spectrograph and Camera 2 ESO s Multi Mode Instrument European Southern Observatory Full Width at Half Maximum Instrument Control System Infra Red InfraRed Array Control Electronics IR Spectrograph And Array Camera Number of DITs Non Destructive Read Number of NDITs New Technology Telescope Observing Blocks Observing Software Phase 2 Proposal Preparation Point Spread Function Right Ascension Read Out Noise Son Of ISAAC Telescope Control System Template Signature File User Support Department formerly User Support Group USG Very Large Telescope Zero Point Table 1 1 Abbreviations and Acronyms used in this manual Chapter 2 SOFI Son of ISAAC 2 1 Optical Layout SOFI is mounted on the Nasmyth A
58. OFI User s Manual 2 3 LSO MAN ESO 40100 0004 Figure E 4 Finding charts for the photometric standards of Persson et al 1998 IV SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 CSKD 8 9MN 7 CSKD 16 18 CSKD 20 21 CSKD 24 37 i 1g gt e aj a E a S vee a H B EM S CE x 5 t uhr NA d sE el Ms a A Betts gle uui ee 22 CSKF 12 14 CSKF 134 gt CSKF 148 D 158 R Rolla ON Se sas e 1i le 138 rt s ME pe i 12 LJ Figure E 5 Finding charts for the photometric standards of Persson et al 1998 V ___00O0___ 113
59. Q ROTPUPIL T T rotate N not rotate Table C 11 SOFLimg obs AutoJtterOffset skies are distributed on a circle surrounding the initial telescope position By default there is no telescope offset before the first exposure If SEQ RETURN is set to True T the telescope slews back to its original position at the end of the template If not the telescope is not moved It is assumed that odd numbered exposures are on the object and are consequently identified as SCIENCE frames Even numbered exposures are assumed to be sky frames and are consequently identified as OTHER If the number of exposures is even SEQ NEXP0 2 pairs of object sky frames are produced If the number of exposures is odd then an extra frame is taken and SEQ NEXPO 1 2 pairs of object sky frames are taken Users are encouraged to give the parameter SEQ NEXPO an even value Guiding is only possible for the object frames if SEQ COMBINED OFFSET is set to True T must be done by the observed during the preparation of the Observing Block and guiding is started before the template starts done by the operator after warning from the observer By default there is no guiding for the sky frames This template can also be used to observe small extended objects by setting the throw small enough so that the object is always within the field of view In such a case however SEQ ROTPUPIL should be set to False F Using this template with a jitter box size param
60. SOFIimg obs AutoJitter template with commonly used values that on any given position of the array most of the images will have sky and only a minority will have objects Clearly minimum three sky images are necessary for this technique but the experience shows that a reasonable minimum number of the sky images and respectively the object sky pairs is 5 7 to ensure a good removal of the objects from the sky frames Note that this may lead to an extra overhead because in some cases the NDIT has to be reduced artificially contrary to the optimization strategy discussed in Sec 2 8 to a number below the optimal just to split the total integration into 5 7 images adding an extra overhead for the telescope offsets This technique is encapsulated in the template called SOFI_img_obs_AutoJitterOffset The pa rameters of this template are listed in Table 3 2 While preparing such observations please keep in mind that the total duration of an imaging OB in Service Mode can not exceed 1 hour in Visitor Mode the duration of an OB is not constrained The example corresponds to the maximum duration acceptable Parameter signature Value Exposure Name NGC6118 DIT 10 NDIT 6 Number of exposures 36 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Filter wheel 1 K Filter wheel 2 open Instrument Mode LARGE FIELD IMAGING Combined Offset T F F Jitter Box Width arcsec 20 Sky Offset Throw arcsec 600
61. UT To the left of the wg5off is the terminal wh5dhs where you run the P2PP Immediately to the left of wg dhs is a sequence of three vertically placed dual screen terminals for the EFOSC and SOFI workstations in this order The bottom screens show their OSs BOBs etc while the upper screens show their RTDs For convenience we can run some applications on the screens of the instruments that are not in use at the moment To the left of the SOFI workstation are the computers of the telescope The user will have no interaction with them When you arrive at the telescope the entire system will have been started for you by the daytime operator During the entire night you will be accompanied a telescope instrument operator TIO It is important to emphasize that the telescope instrument operator will be the one controlling BOB and SOFI You as an observer will be selecting OBs for execution via P2PP and inspecting the incoming data NOTA BENE The TIO is responsible for the operation and safety of the telescope and instrument he has the authority to refuse an action that would jeopardize this safety this include closing the telescope in case of dangerous weather conditions On the terminal wg5off do the following e The machine will be started by the support staff together with all the necessary applications so there is no need to log in e In a workspace of your choice you can start either an IRAF or MIDAS Both reduction packages are
62. Xenon and Neon lamps The Xenon lamp produces an even spread of lines for both the red and blue grisms It is well suited for wavelength calibration Fig A 1 show the main Xenon lines for the blue grism There are two electronic ghosts caused by the very bright lines near one micron between 1 35 and 1 4 microns Fig A 2 shows the main Xenon lines for the red grism The continuum in the red is thermal emission from the lamp Files with the Xe and Ne line wavelengths are available from the Sofl web page http www eso org sci facilities lasilla instruments sofi tools Neon Xenon Lines html Both the Neon and Xenon lamps should be used to calibrate the medium resolution grism Figures A 3 A 4 A 5 and A 6 show the main Xenon and Neon lines 61 62 Relative Intensity SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 4000 F 2000 e e MINIMA c co co Gp co o xy c O C CO OO co pe X eo co QQ gt S st 0 YO coc Q cO Nl c QQ rel cO O ornoo Q wo CO is f e o C nd e NON QQ OO oD wb LO Kej 9516 0 TE JS 9802 4 9925 9 Ghost Op PC CM Figure 1 2104 1 4104 1 6 104 Wavelength A or ES A 1 A Xenon arc spectrum taken with the blue grism The main lines are marked SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 4000 eo Ww ceo o gt ceo LO QQ c e cx oF c LO o gt LO c c e eo o gt c b M e
63. able E 1 Infrared photometric standard stars Persson et al 1998 AJ 116 2475 107 108 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Name RA 32000 DEC J o H o H K o K Ks o Ks BRI 0021 0024 24 6 01 58 22 11 835 0 008 11 086 0 007 10 552 0 010 10 561 0 008 T832 38078 03 04 02 0 00 45 52 11 833 0 010 11 248 0 008 10 890 0 013 10 913 0 008 LHS 191 04 26 20 1 03 36 40 11 621 0 013 11 058 0 012 10 667 0 020 10 717 0 016 IRAS 537 W 05 40 10 5 07 27 38 12 974 0 009 11 032 0 009 9 981 0 013 10 013 0 011 IRAS 537 S 05 40 15 4 07 28 46 13 855 0 012 12 087 0 009 10 972 0 014 11 035 0 012 LHS 2026 08 32 30 5 01 34 37 12 066 0 006 11 497 0 005 11 129 0 007 11 156 0 008 LHS 2397a 11 21 49 2 13 13 10 11 897 0 008 11 190 0 007 10 691 0 008 10 709 0 010 cskd 8 12 31 13 3 63 40 21 11 741 0 010 9 913 0 010 9 151 0 011 9 137 0 013 cskd 9 12 31 16 7 63 40 11 11 372 0 010 9 788 0 010 9 161 0 011 9 154 0 013 cskd 12 12 31 30 0 63 42 41 11 585 0 007 9 506 0 005 8 617 0 007 8 608 0 007 cskd 15a 12 31 44 5 63 49 08 10 879 0 010 9 286 0 007 8 625 0 010 8 617 0 011 cskd 16 12 31 57 8 63 42 21 11 986 0 011 9 982 0 008 9 137 0 011 9 124 0 010 cskd 18 12 31 59 2 63 41 42 9 365 0 010 9 058 0 008 8 999 0 011 8 992 0 010 cskd 20 12 32 04 0 63 43 46 9 903 0 011 8 590 0 008 8 089 0 011 8 077 0 010 cskd 21 12 32 10 9 63 43 16 10 382 0 006 9 918 0 006 9 647 0 007 9 658 0 007 cskd 34 12 31 23 6 63 46 45 12 494 0 015 10 537 0 014
64. age produced in the previous step from the final image ideally this should remove the residuals from the horizontal lines However one should exercise caution because any imperfect flat fielding may affect the procedure leaving artificial gradients across the image As explained above in many cases the cross talk effect can not be removed completely but sometimes the field of view can be rotated so that the ghost image can be relocated to an area of the field away from the target It is enough to relocate the bright source on a different row or column than the target and on a different row than the one corresponding row in the upper lower half of the array 5 2 2 Masking the Bad Pixels The SOFT detector array suffers only minor cosmetic defects it has 0 1 or 1000 bad pixels Some bad pixels masks are available in the SOFI web page http www eso org sci facilities lasilla instruments sofi tools reduction bad pix html We consider the following types of bad pixels e Dead pixels they are computed from the histograms of flats taken during SOFI Calibration Plan the distribution of the pixels values is fitted with a Gaussian mean sigma and we select the dead pixels as the ones whose values are 4c below the mean value e Hot pixels they are computed from the histograms of darks taken during SOFI Calibration Plan the distribution of the pixels values is fitted with a Gaussian mean sigma and we select the hot pixels as the ones
65. aint objects that require long DIT This is done with the templates SOFI_spec_obs_AutoNodNonDestr and SOFI_spec_obs_GenSpecNonDestr In addition to the parameters of Table 3 10 these two templates include those showed in Table 3 11 The array is read within each DIT a number of times equal to NSAMP and for each read out the signal is sampled NSAMPIX times The minimum DIT that can be used for a given NSAMP is NSAMP x 1 64 sec Parameter signature Value NSAMP 30 NSAMPPIX 4 Table 3 11 Parameters in the templates with Non Destructive Read Out and commonly used values 3 4 2 Extended Objects and Crowded Fields For more complex nodding patterns or for observations of either extended objects or crowded fields please use the template SOFI_spec_obs_GenericSpectro This template is discussed in detail in Appendix C It allows to place the slit on a user defined sequence of positions alternating between the object sky and the sky Similar to the imaging of extended objects an effort should be made to take advantage of the large slit length of 4 92 arcmin to obtain observations with the simple nodding procedure described above because the integrating of the clear sky increases the total time spent on the object by a factor of two for the same integration time on the target NOTA BENE Frames taken with the NDR mode must be flat fielded using dome flats taken with the same mode For this reason a specific template exists SOFI spec cal DomeFlatNo
66. ake images with a scale that is a bit finer than that of the large field objective however this objective is chromatic as mentioned above This means that the pixel scale is a function of wavelength It also has an illumination pattern with a central light concentration All the aforementioned objectives suffer various amounts of image degradation over their respective fields of view We routinely try to improve the image quality during interventions Please consult the instrument web pages or the instrument scientists for the latest status of the instrument There is a focal elongator in the grism wheel which can be used in combination with the large field objective to image with a pixel scale that is identical to the pixel scale of the small field objective The image quality is usually superior than that of the small field objective but this mode is 10 to 15 less efficient because of the additional optical elements Both the Small Field and the LF Focal elongator modes are decommissioned Imaging can be done through the standard IR broadband filters J Js H Ks a Z filter which peaks at 0 9 microns and many narrow band filters The filters together with their central wavelengths and widths at half maximum are listed in Table 2 2 Listed also are the cut on wavelengths for the SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 T two order sorting filters These filters are used with the two low resolution grisms Transmission curves of
67. al as visitor The day time telescope operator will provide you with the password e In a workspace of your choice click the left mouse button on the background window free area of the screen to start a terminal and from the prompt type p2pp to start the P2PP tool Next you will need to login with your username and password They are the same as the ones you were given to login in the web letter service where you had found the notification for the approval of your program If you do not have the username and the password ask the day time telescope operator or contact the User Support Department usd help eso org You are now ready to observe and even to obtain some calibrations before the beginning of your observations You can use them to reduce the data during the night 4 3 1 Image Analysis The optics of the NTT are actively controlled and to get the best out of the NTT frequent corrections to the primary are necessary The telescope aberrations are determined by examining images of a guide star recorded through a Hartmann mask The process is called image analysis Its outcome is used to compute corrections to the optical configuration which are applied as deformations to the primary mirror and as displacements of the secondary along three axes Preferably the image analysis should be done with the telescope pointing in the direction of your next target As most exposures in the IR are relatively short the image analysis is done off
68. ameter SEQ SAVE allows to save an acquisition image at the cost of small additional overhead for file transfer and saving The interactive pop up windows are usually displayed before new images have arrived on the RTD Therefore operators are strongly advised to carefully check that a new image has arrived before clicking on these windows e g for storing a fixed pattern for changing the DIT NDIT The arriving of a new image on the RTD is marked by a flashing green dot in the middle of the upper part of the RID window C 3 6 SOFT Spectroscopic Science Templates SOFI spec obs AutoNodOnSlit and SOFI_spec_obs_AutoNodNonDestr These observation templates Table C 35 and C 36 nod the telescope between two positions A and B along the slit A cycle is a pair of two observations AB or BA Cycles are repeated in ABBA se quences E g 3 cycles corresponds to an ABBAAB sequence 4 cycles correspond to an ABBAABBA sequence etc Each observation consist of one or more exposures as defined in the template by the NINT parameter The total number of frames corresponds to SEQ NABCYCLES x SEQ NINT x2 The mean amplitude of the nod is defined by SEQ NODTHROW in arcsec The first exposure A is done after offsetting the object along the slit by SEQ NODTHROW 2 arcsec which can be either negative or positive The second exposure B is therefore SEQ NODTHROW 2 arcsec from the initial 96 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004
69. and this process is fully transparent to the user However recently the collimator has shown tendency to get stuck and currently it is set at a fixed position optimal for wide field imaging and spectroscopy The telescope focusing can be done with a focus pyramid that produces five separate images There is a specific template to do the focusing Note that the image analysis normally corrects for the telescope focus so the focus pyramid is used mostly for health check and the only change of the focus that actually has to be done manually is the correction for the temperature of the primary mirror All this is done by the TIO The role of the visitor is to monitor the quality of the incoming raw data for any degradation 4 4 The SOFI OS GUI Panel The SOFI OS GUI panel Figure 4 1 is used by the instrument operator to show the status of the instrument It can also be used to setup the instrument the detector execute exposures and even point the telescope A detailed description of this panel is given in the OS manual VLT MAN ESO 14100 1510 In addition to the OS GUI panel there are various panels related to ICS DCS and IRACE showing the status of the instrument and the detector These panels are not of direct interest to the observer We do not allow observers to use this panel to start exposures as it is far more efficient and safer to observe with OBs running on BOB Furthermore data archiving a fundamental requirement at the NTT and
70. and it is likely to evolve with time Currently it is only available the telescope Consult your support astronomer or the SofI Instrument Scientist for details Consider the following example Table C 25 the template will produce 6 data cube files Each image is located on the same place i e no jittering and it consists of 2000 reads because NDIT 1000 and DET BURST MODE is set to True each with DIT 0 1 sec The detector is windowed down to 100x100 px The sequence will begin immediately after the start of the template and the set up of the instrument as opposed as at predefined time At the end of the exposures the telescope moves back to the preset position Return to origin True SOFT User s Manual 2 3 Parameter signature Exposure Name DET DIT DET NDIT Number of columns Number of rows First column of window First row of window Turn on Burst mode T on F off Store DITs in a cube T on in individual files F Filter wheel 1 Filter wheel 2 Instrument Mode Jitter Box Width arcsec Return to Origin T F Number of exposures Date of event YYMMDD set to 0 for no event Time of event HHMMSS set to 0 for no event Header Keyword DET EXP NAME DET DIT DET NDIT DET WIN NX DET WIN NY DET WIN STARTX DET WIN STARTY DET BURST MODE DET STORE CUBE INS FILT1 ID INS FILT2 ID INS IMODE SEQ JITTER WIDTH SEQ RETURN SEQ NEXPO SEQ EVENT DATE SEQ EVENT TIME LSO MAN ESO 40100 0004 Value
71. and un vignetted frames The method is outlined below i Collapse the following image sections along detector rows to form one dimensional images columns 50 to 150 in the vignetted image columns 500 to 600 in vignetted image and columns 500 to 600 in the un vignetted image For reference call these images A B and C respectively ii The difference between the shade pattern in the vignetted and un vignetted images is C B iii The shade pattern in the un vignetted frame is then A C B and this one dimensional image 56 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 should be subtracted from the un vignetted frame iv Steps i to iii should be done for the images with the lamp on and the images with the lamp off Once the shading pattern is removed for both one subtracts the image with lamp off from the image with the lamp on to form the flat field Small IRAF and MIDAS procedures performing these steps are available at the telescope and from the SOFI web page in the Data Reduction section named special flat cl and midas specialflat prg respectively The illumination correction removes the difference between the illumination pattern of the dome flat screen and the sky It is derived from a grid of 9 16 observations of a star preferably a standard across the field of view The illumination correction is created by fitting a plane to the fluxes not magnitudes of the star after it has been flat fielded Therefore each illu
72. ary an edge on galaxy with a major axis of 3 4 arcmin a minor axis of 1 2 arcmin must be observed at only two array positions so NEXPO 2 In this case the user should define a suitable rotator offset to align the major axis with the array edge ii Observations of objects that need 2 3 fields to cover the entire target with additional images to sample the sky In principle the template SOFI img obs AutoJitterOffset can accommodate such observations if they are carried out in a sequence of one field at a time However this means 10096 overhead for the sky sampling Instead the user can reduce the overhead by alternating between one sky position and multiple target positions A typical case is observing a large 6 10 arcmin galaxy in a sequence Sky target field 1 target field 2 sky target field 1 target field 2 sky This is a 3 point pattern two of the three pointings are science targets and one is sky Naturally this scheme is more efficient than taking a separate sky for each target but the sky is monitored less frequently Alternatively one can alternate between two or more different sky positions to improve the sky 24 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 subtraction skyl target field 1 target field 2 sky2 target field 2 target field 1 skyl The last pattern is in effect a 6 point sequence An example for this observation scheme is given in Fig 3 2 and the offsets are listed in Table 3 5 N
73. ase Usually a few tens or hundreds of ADU are considered acceptable by most users Finally the total integration time is accumulated by obtaining a certain number of images usually 18 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 specified by the total number of exposures and or the number of jittered images at each position depending on which template is used In case when relatively large integrations are necessary it is simply a matter of increasing the number of exposures However in the cases when the total required time can be accumulated in less than 5 7 exposures it might become difficult to create a good sky for the sky subtraction especially if the field is crowded because the sky image may contain residuals from the stellar images that will produce holes in the sky subtracted data This situation will require to increase artificially the number of exposures to 5 or 7 It might be possible to compensate this increase by decreasing correspondingly NDIT to keep the total integration time constant Still there will be some increase in the overheads Summarizing under average conditions for faint targets one can safely use DIT 30 15 and 10 sec for J H and K filters respectively The narrow band filters can tolerate DIT of 60 120sec The array level will be dominated entirely by the sky Brighter targets require to reduce these times down to 12 8 and 6sec for the broad band filters for 12 15 mag stars and all the way down
74. background is lower longer DITs are necessary For most observations this will mean a DIT between 10 and 60 seconds In addition shorter DIT means larger overheads because every DIT is associated with certain fixed overheads i e for reset and read out and the shorter the DIT the more DITs it will take to complete the desired total integration time Once the DIT is chosen the NDIT can be determined from the desired offsetting frequency During good weather conditions one can stay on the object as long as 2 min for broad band imaging 75 min for narrow band imaging and 15 min for spectroscopy before switching to the sky and in this case NDIT 2 min DIT is appropriate Note that in case of short DITs the NDIT may have to be reduced further by a factor of 1 2 1 3 to take into account the overheads If conditions are not so good e g the sky intensity is fluctuating rapidly or if one simply does not wish to integrate that long then a lower value of NDIT can be used As a general rule the switching frequency should be higher for larger fields of view or longer wavelengths Another consideration that has to be taken into account in choosing the NDIT is the need to have a sufficiently high number of jittered images 1 to create a good sky image without stellar residuals and 2 to remove well the array cosmetic defects and the cosmic rays albeit there are few cosmic rays because of the short integrations for individual images in the IR Th
75. calibrations the respective templates take frames with lamps off to be used as darks SOFT_spec_cal_Arcs This calibration template Table C 43 takes arc spectra with the calibration unit The calibration mirror is automatically inserted at the beginning of the template and is automatically removed at the end SOFT User s Manual 2 3 Parameter signature Exposure Name DIT NDIT Number of columns Number of rows First column of window First row of window Number of Exposure Spectro Mode Which slit Return to origin T F X offset list arcsec Y offset list arcsec Obs Type O or S Guiding N B S LSO MAN ESO 40100 0004 101 Value NGC6118 40 2 1024 1024 1 1 6 LONG_SLIT_BLUE long slit_1 T 0 60 0 120 0 60 0 0 60 60 60 60 OOSOSO B Table C 41 SOFLspec obs GenericSpectro Example Parameter signature Exposure Name DIT NDIT NSAMP NSAMPPIX Number of columns Number of rows First column of window First row of window Number of exposures Spectro Mode Which slit Return to origin T F X offset list arcsec Y offset list arcsec Obs Type O or S Guiding N B S Value NGC6118 40 2 4 10 1024 1024 1 1 6 LONG SLIT BLUE long slit_1 T 0 60 0 120 0 60 0 0 60 60 60 60 OOSOSO B Table C 42 SOFLspec obs GenSpecNonDestr Example 102 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 A 55 5 2 1990 West L
76. copic flats must be taken with the same slit as the observations Imaging flat fields can also be obtained from the twilight sky or from the observations themselves However dome flats represent better the low frequency sensitivity variations of the array Twilight spectroscopic flats cannot be used to flat field your data although they can be used to determine the slit transmission function for spectroscopy of extended sources The slit transmission function describes the transmission variation along the slit It is usually affected by magnetized or sticky dust particles that attach to the slit edges These particles can make the slit effectively narrower Normally steps are taken to clean the slit during interventions so the slit transfer function is highly uniform This effect can be neglected for spectroscopy of point sources Spectroscopic flat fields can be taken with the halogen lamp in the calibration unit Nasmyth flat Again one takes an image with the lamp on and off The flat is the difference between the two normalized to unity The optical path length between the Nasmyth flat field lamp and the front window of SOFI is considerably smaller that the path length between the incandescent lamp on the SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 35 floor of the dome and the front window of SOFI Thus one will see less of the strong atmospheric absorption features in the Nasmyth flats There are four templates to create dome flat
77. d 84 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Parameter signature Value Exposure Name NGC6118 DIT 6 NDIT 10 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Number of Exposures 4 NJITT 3 Jitter box width arcsec 20 Filter wheel 1 Ks Filter wheel 2 open Instrument Mode LARGE_FIELD IMAGING Combined offset T F F Return to origin T F T RA offset list arcsec 50 0 50 0 Dec offset list arcsec 50 50 0 50 Table C 21 SOFLimg obs AutoJitterArray and SOFLimg obs_AutoJitterArray_1 Example the guide probe necessary in case of observations with guiding or non combined offsets offsetting the telescope alone when guiding is not used i e the telescope pointing changes on average at least once every 5 10 minutes However with this flexibility comes complexity This is a complicated template and it is meant to be used in situations in which none of the other templates is suitable Telescope offsets and guiding options are defined as lists in SEQ OFFSETALPHA LIST SEQ OFFSETDELTA LIST and SEQ GUIDING LIST The offsets are relative to the previous position they are in RA and DEC and they are in arcsec There are three guiding options N B and S The option N must be used only for observations without guiding i e the offsets are non combined This works well for short total integrations up to 5 10 min at each individual position The template will fail if this o
78. d into an individual image List of Spectroscopic Modes B R Z J NB 1 061 H K Slit list 0 6 1 2 Table C 44 SOFLspec cal DomeFlats 104 Parameter signature Exposure Name Number of columns Number of rows First column of window First row of window NSAMP NSAMPPIX Number of exposures List of DITs List of NDIT Spectral Mode List Slit List SOFI User s Manual 2 3 Header Keyword DET EXP NAME DET WIN NX DET WIN NY DET WIN STARTX DET WIN STARTY DET NSAMP DET NSAMPIX SEQ NEXPO SEQ DIT LIST SEQ NDIT LIST SEQ SPECTROMODELIST SEQ SLIT LIST Value SOFI 1024 1024 1 1 4 4 1 NODEFAU NODEFAU NODEFAU NODEFAU LT LT LT LT LSO MAN ESO 40100 0004 Description File name prefix Number of columns in the window Number of rows in the window First column of window First row of window Number of Samples Sample Number per Reading Number of exposures Detector Integration Time individual exposure sec Number of DITs averaged into an individual image List of Spectroscopic Modes BRZJNB_1 061 H K Slit list 0 6 1 2 Table C 45 SOFLspec cal NonDestrDomeFlats Appendix D Frame Types There are several basis frame types which are identified by three keywords in the FITS header The full list is given in Table D 1 Keyword Value Type DPR CATG SCIENCE Exposure on target object DPR CATG OTHER Exposure off target sky DPR CATG CALIB Calibration frame
79. d the standard being on different parts of the slit Thus before dividing the standard into the science target one should use the atmospheric absorption features in both to realign the spectra The IRAF task elluric may be useful for this The true intrinsic spectrum of the standard is usually not known with any great precision As dis cussed in Section 3 4 3 usually the telluric standards are selected to be wither solar analog because a good Solar spectrum is available or early type stars because their spectra are easier to model The IR Solar spectrum is available from the National Solar Observatory An atlas of the solar spec trum in the infrared from 1850 to 9000 cm 1 Livingston W amp Wallace L N S O Technical Report 91 001 July 1991 A more detailed description how to use solar analogs as telluric standards and an IRAF based tool can be obtained from Maiolino Rieke amp Rieke 1996 AJ 111 537 An associated IRAF tool is available on the Sofl web site http www eso org sci facilities lasilla instruments sofi tools reduction sofi scripts Recently the fist library of flux calibrated near infrared spectra became available Rayner et al 2009 ApJS 185 289 Cushing et al ApJ 2005 623 1115 and it can be used to remove the artificial emissions for telluric of spectral types other than the Solar For an example on how to treat early spectral type telluric standards and for an empirical library of spectra of ea
80. de List Slit List Spectral Lamp List Parameter signature Exposure Name Number of columns Number of rows First column of window First row of window Number of exposures List of DITs List of NDIT Spectral Mode List Slit List Header Keyword DET EXP NAME DET WIN NX DET WIN NY DET WIN STARTX DET WIN STARTY SEQ NEXPO SEQ DIT LIST SEQ NDIT LIST SEQ SPECTROMODELIST SEQ SLIT LIST SEQ LAMP LIST LSO MAN ESO 40100 0004 Value SOFI 1024 1024 1 1 1 NODEFAULT NODEFAULT NODEFAULT NODEFAULT N Table C 43 SOFLspec cal Arcs Header Keyword DET EXP NAME DET WIN NX DET WIN NY DET WIN STARTX DET WIN STARTY SEQ NEXPO SEQ DIT LIST SEQ NDIT LIST SEQ SPECTROMODELIST SEQ SLIT LIST Value SOFI 1024 1024 1 1 1 NODEFAULT NODEFAULT NODEFAULT NODEFAULT 103 Description File name prefix Number of columns in the window Number of rows in the window First column of window First row of window Number of exposures Detector Integration Time individual exposure sec Number of DITs averaged into an individual image List of Spectroscopic Modes B R Z J NB 1 061 H K Slit list 0 6 1 2 Lamp list N none B both Xe Xenon Ne Neon Description File name prefix Number of columns in the window Number of rows in the window First column of window First row of window Number of exposures Detector Integration Time individual exposure sec Number of DITs average
81. defined by DET NSAMP The largest the number of samples the lowest the read out noise However DET NSAMP must not exceed 60 above this value in fact the glowing of the shift registers become the dominant source of noise For long integrations DET NSAMP between 20 and 40 is advisable A way to further low the read out noise keeping DET NSAMP small enough is to sample the video signal more than once The number of sampling of the video signal is DET NSAMPIX 4 is a good number for this parameter For long integrations DIT should not exceed 300 sec The following examples Table C 37 and C 38 show typical parameters values According to the NINT and number of cycles they produce 3x3x2 18 files each of them corresponding to the average of the DIT x NDIT SOFT_spec_obs_GenericSpectro and SOFI spec obs GenSpecNonDestr These observation templates Table C 39 and C 40 are for spectroscopy and they have the flexibility to do any sequence of telescope offsets whether they be combined offsets offset the telescope and guide probe or non combined offsets offset the telescope alone SOFL spec obs GenSpecNonDestr is identical to SOFIL spec obs GenericSpectro except for the non destructive readout more With this flexibility of these templates comes complexity They are meant to be used in situations that cannot be accommodated by the nodding templates described above i e when the target is too extended and to obtain sky spectrum one has to move f
82. des are included for the benefit of ESO Science Archive users 6 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Rotator angle 0 no 38 Object Instrument N 1 1 n E Figure 2 2 Orientation of SOFT for rotator angle 0 deg In imaging the North is to the left and the East to the bottom of the image The slit for the spectroscopy is oriented North South Nota Bene To align the North South axis of the field of view or the slit in case of spectroscopy long a certain Position Angle one has to apply in the acquisition template a rotation angle equal to this Position Angle Starting from Apr 1 2007 new templates are available to the users that allow them to rotate the instrument between the individual pointings in a single template as opposed to setting one position angle and acquiring all the data in the OB at that orientation These templates are particularly useful for observations of fields containing bright sources that would cause significant reflections and cross talk artifacts Due to the rotation the artifacts will be located on different place of the detector array Taking data at enough different position angles derotating aligning and combining the images will allow the users to reduce the artifacts The exact number of position angles will depend on the number and the brightness of the bright sources Note that the movement of the rotator will cause extra overhead The spectroscopic objective can be used to t
83. described fully in Appendix C The P2PP tool is a browser editor that enables observers to create lists of targets and observing descriptions and then associate them with an acquisition template to form an OB The day time telescope operator will start you up on the P2PP and will help you to optimize the observations but we highly recommend the users to read carefully the P2PP Users Manual available on the User Support Department USD web page at http www eso org sci observing phase2 USD html Once the OBs are created they can be stored either in the local cache which resides on the same machine where the OBs were created p2pp cache ID or they can be exported as ASCII files The exported format of the files allows to ftp them from one machine to another NOTA BENE Under no circumstances edit the OBs by hand with a text editor or with scripts The software is sensitive to the format of the OB and extra or not enough spaces tabs often cause crashes In visitor mode P2PP is used to select OBs for execution Observing blocks are selected by highlight ing them with the left mouse button in the P2PP window Once the OB is selected the instrument operator will transfer it to the BOB Broker of Observing Blocks where it will be executed Note that once an OB is transferred to BOB any changes that the observer make to the OB in the P2PP have no effect on the observations unless the OB is loaded into BOB again In general the hig
84. dicated that the monitoring of the sky is crucial for the accuracy of the photometry Furthermore having a reference source in the field so it is observed simultaneously with the target helps to improve the photometric accuracy further It is best if the reference source is of comparable brightens or brighter than the target It is advisable to split long time series into smaller bits typically 5 10 min The minimum separation between the different series without moving the telescope or a filter wheel is 8 sec SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 27 The data product is a 3 dimensional cube where every slice along the third axis is essentially 1 dimensional image or detector read The first two dimensions have the meaning of the usual rows and columns The file header contains the UT at the start of the sequence and the UT at the end as time of file creation The ESO time distribution system has absolute accuracy better than tenth of a sec and even much better relative accuracy This will be tested and more detail will be made available in the future To take full advantage of this more the targets have to be sufficiently bright to avoid the necessity to rebin the data later on The standard calibration plan for FastPhot includes darks and flats normally dome flats sky flats if possible taken with the same detector configuration This mode is not supported by the Sofl pipeline The time to execute a series of 1000 images
85. duce 6 fits files Each image is jittered regarding to the previous one and the jitter offsets are chosen randomly inside a box of 20 arcsec Jitter Box Width around the central position Each of them corresponds to the average of 6 exposures NDIT of 10sec DIT At the end of the exposures the telescope moves back to the preset position Return to origin True SOFLimg obs AutoJitterRot This observation template Table C 9 C 10 is identical with the SOFI_img_obs_AutoJitter with the additional capability to move the rotator by a given offset The offset is relative with respect to the current position The user is strongly advised to check if the rotator has sufficient range to complete the observations and to monitor the rotator angle variation during the execution of the template The offset parameter is a list the default value is list of one zero the valid range is from 360 to 360 degr SOFI_img_obs_AutoJtterOffset This observation template Table C 11 allows the user to move the telescope alternatively between the object and a nearby patches of sky However when pointing to the object the position of the telescope is still randomly distributed within a box of size SEQ JITTER WIDTH in arcsec The offset skies are at a constant distance defined by SEQ SKYTHROW in arcsec from the original telescope position but at an angle randomly distributed between 0 and 360 degrees i e the offset SOFT User s Manual 2 3 Para
86. e T guiding ON F guiding OFF Preset Telescope F T SEQ PRESET True T full preset F fine tunning of the pointing Save Image F T SEQ SAVE False T preserve the acq image F no Table C 6 SOFL_img acq_ MoveToPixel To move the target from one position of the array to another the operator simply draws an arrow on the screen with the left hand side button of the mouse At this point a window which lists the pixel co ordinates at the start and the end of the arrow will appear The operator can accept the offsets cancel or edit the co ordinates directly If the offsets are accepted the telescope offsets by the desired amount Finally the operator is given the possibility to redraw the arrow for refining the position of the target if necessary Once the operator is satisfied the template finishes Guiding is possible defaulted to the Telescope Control System s option Boz To Star parameter value B only if SEQ COMBINED OFFSET is set to T and if guiding is on before the start of the template The experience has shown that the guiding is necessary if the observer plans to stay at one pointing more than 5 10 min This is rarely the case in imaging mode but can often happen in spectroscopic observations The SEQ PRESET parameter allows the user to use the fine tunning pointing functionality of the template without presetting the telescope On other words the user can improve the position of the target on the array without ful
87. e from Ivanov et al 2004 ApJS 151 387 A list of O F and G type stars selected from the HIPPARCOS catalog with magnitudes appropriate for SOFI spectroscopy is available at the telescope and on the SOFI web page together with a web based tool for automated search for stars with given spectral type and brightness near the science target http www eso org sci facilities paranal sciops catsearch html A similar tool stdsopMain with extra capability to automatically create OBs for observing of telluric standards is available on the SOFI instrument workstation but we recommend that the users create their own OBs with the P2PP tool to retain consistent program ID Nota Bene Until recently IR spectrophotometric standards were not available How ever Cohen et al 2003 AJ 125 2645 published supertemplates for a series of 33 stars that span wavelength range from the UV to mid IR 40 jm The preliminary tests show excellent agreement between the synthetic photometry derived from these spectra and the observed magnitudes of the stars Pending further tests this spectral library may prove extremely useful for flux calibration of IR spectra 3 5 Calibration Frames 3 5 1 Darks Biases The concept of bias frames with IR arrays does not have the same meaning as CCD bias frames With IR arrays a zero second exposure is not possible It is better to think of all exposures without direct illumination as dark frames Note that in this contex
88. e low frequency variability improves to better than 1 Illumination Correction surfaces can be downloaded from the SOFI Web page The observatory staff produces them typically once a month Templates for observing a standard star in an ar ray of 16 positions are available in the impex directory of SOFI on wsofi in the calib directory SOFI_Illum_Correction_J SOFI Illum Correction H etc NOTA BENE The illumination correction frame refers only to the flat field that was used to create it and it must not be used with other flat fields The reason for this is that the illumination correction removes the effect of different dome screen illumination and in general every time the user obtains dome flats the intensity of the lamps is slightly different 3 5 4 Arcs Wavelength calibration can be done with the Xenon and Neon lamps in the adapter The Xenon lamp has a better distribution of lines and is by itself sufficient to calibrate the wavelength scale of data taken with the low resolution grisms Both the Neon and Xenon lamps should be used to calibrate the medium resolution grisms Arcs are taken with the template SOFI spec cal Arcs The parameters in this template are dis played in Table 3 14 In this template one can enter a list of DITs NDITs grisms slits and lamps If the number of exposures is greater that the number of elements in either of these lists the list is repeated until the the correct number of exposures have been comple
89. e observation The offsets in the list are relative to the previous position they are in RA and DEC and they are in arcsec Note that the telescope offset are executed before the corresponding image is taken In other words if you want the first image to be at the position you have acquired the target the first offset has to be 0 0 If the number of the images is longer than the number of the offsets the software loops back to the beginning of the list and executes again the first offset If the number of the images is shorter than the number of the offsets the software executes only as many offsets as necessary starting from the beginning of the list The RA and DEC lists can have different length but this can be extremely confusing It is good practice to use either lists with a single value or lists of equal length If SEQ COMBINED OFSET is set to True must be done by the observer during the preparation of the Observing Block and if guiding was started before the start of the template will be done by the operator after a warning from the observer the guiding is Boz To Star Given the short time spent on a single position typically 2 3 min determined by the sky level variations guiding is usually rarely used If SEQ RETURN is set to True T the telescope slews back to its original position at the end of the template If not the telescope is not moved Consider the following example Table C 14 it gives to you 5 imag
90. e some concrete examples In later sections we discuss some finer points The templates are described in detail in The SOFI TSF Parameters Reference Guide 3 2 Imaging It is not unusual for the objects of interest to be hundreds or even thousands of times fainter than the sky Under these conditions it has become standard practice to observe the source together with the inevitable sky and subtract from it an estimate of the sky Since the sky emission is generally 16 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 17 variable the only way to obtain good sky cancellation is to do this frequently The frequency depends on the wavelength of observation and respectively on the nature of the sky background emission and on meteorological conditions Ideally one would like to estimate the sky more quickly than the time scale of the sky variations While this could be done quickly with the traditional single and especially double channel photometers the overhead in observing with array detectors and the necessity of integrating sufficient photons to achieve background limited performance are such that the frequency is of the order of once per minute This sky subtraction technique has the additional advantage that it automatically removes offsets due to fixed electronic patterns bias and dark current NOTA BENE The sky and the object sky have to be sampled equally integrating more on the object sky than on the sky will not improve the overal
91. e the bulk of the observations will be done iii Good focus is very important Poor focus can result in distorted images if any aberrations are present In fact distorted images are a good sign that either the telescope needs to be refocused or another image analysis needs to be done The focus depends linearly with the temperature of the four serurriers The coefficient is 0 079 mm of M2 movement per degree In good seeing a displacement in M2 of 0 03 mm induces a noticeable degradation in the image quality So if the temperature has 46 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 changed by 0 5 degrees or more since the last focus one should refocus On La Silla the temperature changes quickest during the first hour of the night iv During long integrations greater that five minutes it is possible to run the NTT in closed loop mode that is the optics are adjusted during the observations For SOFT integrations are usually shorter and this is usually not needed NOTA BENE The best way to find out if it is time for another image analysis is to monitor the image quality of the incoming raw data and to request an image analysis if you notice degradation As a general guideline however you should trust the TiO s experience 4 3 2 Focusing The telescope and the instrument are focused separately Nominally the instrument is focused automatically to a pre defined collimator position depending on the instrument configuration
92. e very good for wavelength calibration Avoid the blended Q branch lines Sample spectra for Xenon Neon and the main OH emission lines are displayed in appendix A Predefined OBs for wavelength calibration with the correct exposure times are available at the tele 38 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Parameter signature Value Exposure Name Arc Calibration DIT LIST 2222 NDIT LIST 5555 Number of exposures 4 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Spectro Mode BR Slit List 1 Lamp List N Xe N Ne Table 3 14 Parameters in the SOFI_spec_cal_Arcs template with commonly used values scope 3 6 Finer Points Some of the information in this section is available through the manual but here it is concentrated for the convenience of the user 3 6 1 Choosing DIT NDIT and NINT The appropriate value of DIT depends on the intensity of the source and the background First and foremost the DIT must be kept short enough that moderately bright objects of scientific interest do not saturate On the other hand in order to maximize the S N ratio one would like to work in background limited instrument performance i e the sky intensity in a single frame should be sufficiently high that the sky shot noise will dominate over the detector read out noise In the Ks band where the sky is bright this can be reached with DITs of one second Through the spectroscopic modes where the
93. ector down but this does not reduce the minimal DIT because the windowing is applied to the data by the Sofl observing software after the entire array was read However the detector can be read only partially if the so called hardware windowing is applied At the moment this is available to the user only in FastPhot mode In addition to the windowing to speed up the photometry the new mode allows to save time by storing only the individual array readouts instead of their difference and by using the fast RdRstRd Read Reset Read readout mode The readout time is in order of 0 002 0 008 seconds i e between every integration a few miliseconds are spent to read the detector and to transfer the image The hardware window is centered at 512 512 at the detector to take advantage of the simultaneous work of the four ADCs analog to digital converters The window can be rectangular Due to the detector readout pattern windoing along the rows and columns have different effect on the DIT reductions If you need a rectangular window it is better to extend it along the rows and to have the narrow side along the columns There is no telescope movement during a time series so it is best to make sure the object is placed on an array location clear of bad pixels The pixel scale is 0 288 arcsec px so windows smaller than 4x4 px are likely to lead to aperture losses and will not allow to measure reliably the sky variations The preliminary tests in
94. ector defining the instrument induced polarization will rotate relative to the sky A method to eliminate the instrumental polariza tion is outlined by Sperello di Serego Alighieri 1989 Proceedings of 1st ESO ST ECF Data Analysis Workshop Technical reports Wolf Vanzi amp Ageorges 2002 have been released which describe in details this mode and its operation They are available on the SOFI web page http www eso org sci facilities lasilla instruments sofi tools report ps 2 5 The DCS Detector Control System The DCS is made up of the detector the front end electronics and the controller InfraRed Array Control Electronics or IRACE The IRACE controller controls the detector front end electronics and manages pre processing of the data before sending it to the SOFI workstation It includes an embedded Sparc and two sets of transputers Further details on the IRACE system can be found at http www eso org sci facilities develop detectors controllers irace html The amount of data pre processing depends on the readout mode The readout modes available with SOFI are discussed below The detector used by SOFI is a Rockwell Hg Cd Te 1024x1024 Hawaii array manufactured by Rock well Scientific with 18 5 micron pixels The array is read out in four quadrants The average quantum efficiency is 65 Fig 2 4 The dark current is very low 20 e hour and the readout noise with the IRACE controller in DCR Double Correlated Read mode i
95. ed before the start of the template If SEQ RETURN is set to True T the telescope slews back to its original position at the end of the template If not the telescope is not moved Note that the instrument mode is not a parameter of this template because all polarimetric obser vations are taken in Large Field mode This example Table C 32 gives the typical values of the parameters Note that to determine the Stokes parameters the template will have to be executed twice with a rotator offset of 45 degree C 3 3 SOFT Polarimetric Calibration Template SOFLimg cal PolarimDomeFlats This template Table C 33 takes polarimetric dome flats for a list of rotation angles 92 Parameter signature DET DIT DET NDIT Filter wheel 1 Filter wheel 2 Alpha Offset arcsec Delta Offset arcsec Add Velocity Alpha Add Delta Velocity Rotation Offset on Sky Combined offset F T Preset Telescope F T Save Image F T Parameter signature Exposure Name DET DIT DET NDIT Number of columns Number of rows First column of window First row of window Number of exposures Filter wheel 1 Filter wheel 2 Combined offset F T Return to Origin T F X offsets list arcsec Y offsets list arcsec Rotator Offset SOFI User s Manual 2 3 Header Keyword DET DIT DET NDIT INS FILT1 ID INS FILT2 ID TEL TARG OFFSETALPHA TEL TARG OFFSETDELTA TEL TARG ADDVELALPHA TEL TARG ADDVELDELTA TEL ROT OFFANGLE SEQ
96. eg you have to set TEL ROT OFFANGLEPA 32 deg There are three guiding options N B and S The option N must be used only for observations without guiding i e the offsets are non combined This works well for short total integrations up to 5 10 min at each individual position The template will fail if this option is selected AND the guiding is started The option B stands for Box To Star and it must be used with guiding i e the offsets are combined with with appropriate counter movement of the guiding arm to maintain the pointing This is necessary for long total integrations more than 5 10 min at each individual position The option S stands for Star To Box It has been disabled i e the template fails if the guiding option is set to S but it was kept in the software for historic reasons Note With large combined offsets the guide probe may not be able to follow the one guide star In such a case the guiding system will automatically find another star and resume guiding The observation type can be defined for each image and entered as a list in SEQ OBSTYPE LIST O stands for Object and S stands for Sky In addition to observations that require large sky offsets these templates are very useful for slit scanning across an object ie simulation of 3D spectroscopy by defining a list of offsets in the y direction perpendicular to the slit direction A third case when these templates can be used is when the user wants to ob
97. ent of most electronic components it remained unreproducible Usually it is confined to a single quadrant The noise amplitude is only a few ADU and it adds negligible error in imaging or low resolution spectroscopy mode However in high resolution spec troscopy mode especially when observing faint targets it is advisable to select nodding and jittering parameters placing the target in the unaffected quadrants 2 5 3 Windowed Reading The SOFT detector can be windowed Each window is defined by the starting pixel coordinates and the size of the windowed region The entire array is still read out by the IRACE controller however only the windowed section is transferred to the workstation This leads to a slight overhead decrease For position angle PA 0 deg starting coordinates of 1 1 correspond to the North East corner and they increase toward South West respectively To reduce further the overhead one has to resort to hardware windowing This is available only for the Fast Photometry Burst observations In this case only part of the array is read and transferred and the significant reduction of the overheads comes from reading a smaller fraction of the array To take advantage of having four Analog to Digital Converters the windowed part of the array is placed at the center of the detector 512 512 so the four ADC s can be used simultaneously The window size determines both the minimum DIT and the readout time 2 6 Calibration Un
98. ent pixel scales The pixel scales and the fields of view are summarized in Table 2 1 The large and small field objectives are used with the corresponding mask in the mask wheel that covers the same field of view The masks reduce the amount of stray light entering the instrument For a normal use of SOFI rotator angle 0 deg the orientation is showed in the Fig 2 2 You could modify the orientation simply using a rotator angle in the template This option could be useful in spectroscopy if you want to align two objects in the slit the positive angle is from the north to the east or in imaging if you want to map some asymmetric target ie an elongated galaxy A simple rule of thumb says to align the North South axis of the field of view or the slit in case of spectroscopy along an axis on the sky with a certain Position Angle one has to apply in the acquisition template a rotation angle equal to this Position Angle SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 5 COLLIMATOR LENS SOFI instrument concept SLIT MASK WHEEL Figure 2 1 Optical layout of SOFI Objective Pixel Scale Field of View Large Field Objective 0 288 4 92 x 4 92 Spectroscopic Objective 0 273 4 66 x 4 66 Small Field Objective decommissioned 0 144 2 46 x 2 46 Large Field Objective Focal Elongator 0 144 2 46 x 2 46 decommissioned Table 2 1 The fields of view and the pixel scales available with SOFI The decommissioned mo
99. entioned above will be avoided The only case when the dark bias subtraction is still necessary is the reduction of the flat field calibration images See the description of the special procedure for that in Section 5 2 5 5 2 4 Sky Subtraction This is a critical operation which depends on how many frames are available to create a sky image The experience shows that at least 5 7 images are necessary to achieve acceptable sky subtraction and this number depends strongly on the nature of the sky field if the field is crowded with stars i e the Galactic plane this may not be enough In such cases we recommend to make sure that at least 12 15 images at different locations are available In general the more frames you have the better the results but up to a point because of such effects as pupil rotation and increasing overheads There are two basic strategies for sky subtraction i creating a single sky image and then subtracting it from each individual frame Here is a step by step description of the procedure to remove the sky from a stack of n jittered frames where n is 54 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 large a For each image take the 10 images that were taken closest in time These images will be used to estimate the sky In case of jittered observations when the target field is on all images these 10 images will indeed include the target In case of observations alternating between the object and the
100. equence of frames will be generated The template SOFI spec cal DomeFlatNonDestr contains in addition the parameters of Table 3 11 It was pointed out earlier that shading one of the components that make up a dark frame is a function of the incident flux This means that the method of creating dome flats described above does not remove perfectly the shading pattern The effect for most observations is small and manifests itself as a discontinuity of a few percent in the ZP across the center of the array For most programs this is not a major problem Nevertheless we have developed a technique to remove this residual and this technique is embodied in the template SOFI_img_cal_SpecialDomeF lats This template in addition to the four frames taken with the template SOFI spec cal DomeFlats takes frames with the mask partially obscuring the array Therefore each exposure generates a sequence of 8 frames In addition to them 1 2 or more images are taken at the beginning in order to adjust and check the illumination level set to 4000 6000 ADU so the total number of the obtained images by this template may vary Note that only the last 8 images are actually the flat field Half of this sequence is shown in Fig 3 3 If the parameter Number of Exposures is gt 1 a corresponding sequence of frames will be generated so generally use 1 These frames are used to estimate the shading Further details are given in the section describing data reduction
101. es as needed until the correct number of frames have been acquired The lists can have any length however having lists of different lengths can become extremely confusing It is good practice to use either lists of one value and or lists of equal length SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 85 Parameter signature Header Keyword Value Description Exposure Name DET EXP NAME SOFI File name prefix DET DIT DET DIT NODEFAULT Detector Integration Time individual exposure sec Number of columns DET WIN NX 1024 Number of columns in the window Number of rows DET WIN NY 1024 Number of rows in the window First column of window DET WIN STARTX 1 First column of window First row of window DET WIN STARTY 1 First row of window Number of exposures SEQ NEXPO 1 Number of exposures in the sequence List if NDIT DET NDIT NODEFAULT Number of DITs averaged into an individual image Filter wheel 1 INS FILT1 ID NODEFAULT Filter wheel 1 position Filter wheel 2 INS FILT2 ID NODEFAULT Filter wheel 2 position Instrument Mode INS IMODE NODEFAULT Instrument Mode Return to Origin T F SEQ RETURN T Returns the telescope to the original pointing if True RA offsets list arcsec SEQ OFFSETALPHA LIST NODEFAULT list of offsets along RA Dec offsets list arcsec SEQ OFFSETDELTA LIST NODEFAULT list of offsets along Dec Obs Type O or S SEQ OBSTYPE LIST NODEFAULT O object S sky Guiding N B S SEQ GUIDING LIST NODEFAULT N no guiding
102. es at 5 different positions Each image is the average of 3 NDIT exposures of 20sec DIT The first image is taken at the position at which the target was acquired The following example Table C 15 will produce 10 images repeating the sequence of 5 offsets Note SOFT User s Manual 2 3 Parameter signature Exposure Name DET DIT DET NDIT Number of columns Number of rows First column of window First row of window Number of exposures Filter wheel 1 Filter wheel 2 Instrument Mode Combined offset F T Return to Origin T F RA offsets list arcsec Dec offsets list arcsec Header Keyword DET EXP NAME DET DIT DET NDIT DET WIN NX DET WIN NY DET WIN STARTX DET WIN STARTY SEQ NEXPO INS FILT1 ID INS FILT2 ID INS IMODE SEQ COMBINED OFFSET SEQ RETURN SEQ OFFSETALPHA LIST SEQ OFFSETDELTA LIST LSO MAN ESO 40100 0004 79 Value SOFI NODEFAULT NODEFAULT 1024 1024 1 1 1 NODEFAULT NODEFAULT NODEFAULT F T NODEFAULT NODEFAULT Table C 13 SOFLimg obs Jitter Description File name prefix Detector Integration Time individual exposure sec Number of DITs averaged into an individual image Number of columns in the window Number of rows in the window First column of window First row of window Number of exposures in the sequence Filter wheel 1 position Filter wheel 2 position Instrument Mode T guiding ON F guiding OFF Returns the telescope to the original pointing if T
103. eter SEQ JITTER WIDTH set to zero is equivalent to staring on the object For the pupil rotation see the discussion in Sec 3 6 Consider the following example Table C 12 the sky is taken at 600 arcsec of the object and a small jitter of 20 arcsec around the target is done The number of exposures corresponds to the TOTAL 78 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Parameter signature Value Exposure Name NGC6118 DIT 6 NDIT 10 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Number of Exposures 10 Filter wheel 1 Kg Filter wheel 2 open Instrument Mode LARGE FIELD IMAGING Combined offset T F F Jitter Box Width arcsec 20 Return to origin T F T Sky Offset Throw arcsec 600 Rotate Pupil T Table C 12 SOFLimg_obs_AutoJitterOffset Example number of exposures that means SCIENCE exposures SKY exposures So here you will have 5 images of your objects and 5 images of the sky Each image is the average of 10 NDIT exposures of 6 sec DIT No guiding is used SOFLimg obs Jitter This observation template Table C 13 allows the user to offset the telescope between exposures according to a list of predefined offsets SEQ OFFSETALPHA LIST and SEQ OFFSETDELTA LIST giving more freedom and flexibility Complicated patterns can be used pointing the telescope exactly on areas of interest However this comes to a price of more difficult preparation of th
104. every image has to be multiplied by that of the first masks that matches the location of the cluster and summed with that of the second masks that again matches the target quadrant Remember to produce those manipulations on a copy of the original data Finally a suitable low threshold in imcombine will make sure that the target has been masked out from the sky An alternative to this masking is to use a floating sky constructed from the six in case of the 4 point scheme images immediately preceding and succeeding every image Those images have the target located on different locations than on the sky subtracted image As usual the global average is preferable if the sky background is stable and the floating sky subtraction is preferable if the sky varies NOTA BENE The sky subtraction of every observation has to be considered individually The recipes provided here refer to the most general cases Every observation is different and covering all possibilities is out of the scope of this manual We provide some tools that help to carry out the sky subtraction Check the Data Reduction section of the SOFI web page and if necessary ask the telescope operator what options are available to receive help usually they are limited to calling Paranal on the phone or to sending an e mail to the Sofl Instrument Scientist the contact details are available on the SofI web page 5 2 5 Flat Fields and Illumination Corrections Creating a
105. f Imaging Template s to use Imaging of uncrowded fields SOFLimg_obs_AutoJitter or or point like objects SOFLimg obs_Jitter Imaging of fain objects SOFLimg obs_AutoJitterRot or around bright objects SOFLimg obs_JitterRot Imaging of crowded fields SOFLimg obs_AutoJitterOffset or or extended objects SOFLimg obs_JitterOffset Map of extended fields SOFLimg_obs_AutoJitterArray or SOFLimg obs_AutoJitterArray_1 Imaging requiring complex telescope offsets and or guiding options SOFLimg obs GenericImaging Fast Photometry SOFLimg_obs_FastPhotJitt Imaging Polarimetry SOFLimg obs Polarimetry Table C 2 Short guide for imaging and polarimetry templates Type of Spectroscopy Template s to use Spectroscopy of point like or moderately extended objects SOFTI_spec_obs_AutoNodOnSlit As above but in Non Destructive read out mode SOFI_spec_obs_AutoNodNonDestr Spectroscopy of extended objects i e wider than 2 arc minutes or complex sequences of slit positions SOFI_spec_obs_GenericSpectro As above but in Non Destructive read out mode SOFLspec_obs_GenSpecNonDestr Table C 3 Short guide for spectroscopic templates Type of calibration Template s to use Darks SOFLimg cal Darks Imaging Dome Flat Fields SOFLima cal DomeFlats Special Imaging Dome Flats SOFLima cal SpecialDomeFlats Standard Star imaging SOFLimg cal StandardStar Polarimetric Dome Flat Fields SOFLimg cal PolarimDomeFlats Arcs spectroscopy SOFLspec cal Arcs Spec
106. f the fact that the true intrinsic IR solar spectrum of the Sun is available It was corrected by observing the Sun at different zenith angles and extrapolating to secz 0 an extremely time consuming technique that is not reasonable to the usually faint science targets The IR Solar spectrum is available from the National Solar Observatory An atlas of the solar spectrum in the infrared from 1850 to 9000 cm 1 Livingston W amp Wallace L N S O Technical Report 7291 001 July 1991 A more detailed description on how to use solar analogs as telluric standards and an IRAF based tool can be obtained from Maiolino Rieke amp Rieke 1996 AJ 111 537 The tool is available at the SofI web page http www eso org sci facilities lasilla instruments sofi tools reduction sofi scripts in a tarball that also contains related material including lists of stars with reliable spectral class de terminations that can be used as telluric standards The final step is to flux calibrate the object spectrum The flux calibration is done by integrating the object spectrum over one or more broadband filter band passes and scaling the result with the respective broadband magnitudes This method typically gives accuracy of order of 5 1096 It is very different to what is done to flux calibrate spectra in the optical In the optical atmospheric absorption is a relatively smooth function of wavelength so it is sufficient to divide extracted spectra by a smoot
107. ficient to use the Xenon lamp for wavelength calibration For the medium resolution grism both the Xenon and Neon lamps should be used The main Xenon and Neon lines are identified in Appendix A For the medium resolution and red grisms a cubic fit to calibrate the dispersion is adequate For the blue grism a quadratic fit is better 5 3 6 Removing of the Atmospheric Absorption Features and Flux Calibration The IR spectra are dominated by atmospheric absorption features To remove these features it is customary to observe a star with a featureless spectrum at a similar airmass to that of the target This star is called telluric standard The reduction involves the following steps e the spectrum of the target has to be divided by the spectrum of the telluric standard This will remove the atmospheric absorption because presumably it affects both spectra in the same way e the science spectrum has to be multiplied by the true intrinsic spectrum of the standard This will remove the artificial emission features introduced to the target spectrum by the intrinsic absorptions in the spectrum of the standard Removing the atmospheric features is a critical operation In general the spectra of the atmospheric standard and the spectra of the science target will not have exactly the same wavelength scale They may differ by as much as half a pixel This could be caused by internal flexure within the instrument or by the science object an
108. fields two for imaging SOFI img cal DomeFlats and SOFL img cal SpecialDomeFlats and two for spectroscopy SOFI spec cal DomeFlats and SOFL spec cal DomeFlatNonDestr Please note that the last template must be used ONLY to flat frames taken with the NDR Mode Table 3 13 gives an example for imaging dome flats Parameter signature Value Exposure Name Imaging Dome Flats Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Number of exposures 1 DIT LIST 3 NDIT LIST 60 Filter wheel 1 J Filter wheel 2 open Instrument Mode LARGE FIELD IMAGING Table 3 13 Parameters in the SOFLimg cal DomeFlats template with commonly used values The template actually takes four exposures in total the first with the lamp off then two exposures with the lamp on and then a final exposure with the lamp off The intensity of the dome flat field lamp is controlled by a panel on the tcs machine The instrument operator will adjust the voltage until a level of a few thousand usually 3000 4000 ADU counts is reached when the lamp is on In this template one can enter a list of DITs and NDITs If the number of exposures is greater that the number of elements in either of these lists the list is repeated until the the correct number of exposures have been completed Note that each exposure generates a sequence of 4 frames as explained before If the parameter Number of Exposures is gt 1 a corresponding s
109. fix Number of columns DET WIN NX 1024 Number of columns in the window Number of rows DET WIN NY 1024 Number of rows in the window First column of window DET WIN STARTX 1 First column of window First row of window DET WIN STARTY 1 First row of window Number of exposures SEQ NEXPO 1 Number of exposures in the sequence List of DITs SEQ DIT LIST NODEFAULT Detector Integration Time individual exposure sec List if NDIT SEQ NDIT LIST NODEFAULT Number of DITs averaged into an individual image Filter wheel 1 INS FILT1 ID NODEFAULT Filter wheel 1 position Filter wheel 2 INS FILT2 ID NODEFAULT Filter wheel 2 position Instrument Mode INS IMODE NODEFAULT Instrument Mode Table C 27 SOFLimg cal DomeFlats and SOFLimg cal SpecialDomeFlats unlike the other templates where the offsets are defined along the RA and Dec here they are defined along detector rows and columns and are entered into SEQ OFFSETX LIST and SEQ OFFSETY LIST Offsets are in arcsec Guiding is possible defaulted to Box To Star if SEQ COMB OFFSET is set to True T and if guiding was started before the start of the template If SEQ RETURN is set to True T the telescope slews back to its original position at the end of the template If not the telescope is not moved Although it is possible to observe standard stars with any of the imaging science templates using this one is recommended because it sets up properly the header keywords necessary for the pipeline
110. focus of the NTT The light from the tertiary mirror of the telescope enters the front window which has no optical power Immediately after the front window is a cryogenically cooled mask wheel which coincides with the telescope focus The mask wheel contains several masks one for each imaging objective three for long slit spectroscopy a pinhole mask and a special mask used with polarimetry The optical layout is shown in Fig 2 1 The mask wheel is followed by a collimating lens used to focus the instrument two filter wheels a grism wheel an objective wheel and then the detector itself The re imaged pupil of the telescope the primary mirror is located on a slightly undersized stop just before the grism wheel The first filter wheel contains the standard broad band near IR filters several narrow band filters two order sorting filters for low resolution spectroscopy an open position and a fully closed position The second filter wheel contains more narrow band filters a focus pyramid an open position and a fully closed position The grism wheel contains three grisms for long slit spectroscopy a Wollaston prism for imaging polarimetry an open position and a fully closed position The objective wheel contains two objectives for imaging at 07288 large field and 0 144 small field decommissioned per pixel a spectroscopic objective an open position and a fully closed position 2 2 Imaging SOFI offers imaging at several differ
111. g dashed red lines show the soon to be commissioned J filter top panel and the typical broad K filter Persson et al 1998 AJ 116 2475 see their Table 10 The dotted blue line is the atmospheric transmission model for Mauna Kea for Am 1 0 and water vapor column of 1mm Lord S D 1992 NASA Technical Memorandum 103957 courtesy of Gemini Observatory The data for the plot and a SuperMongo script are available from the SOFI web page covers the region from 1 53 to 2 52 microns Both grisms are made of KRS5 refractive index 2 44 and the entrance surface of both is inclined to the optical axis The blaze angle of the grooves is equal to the apex angle of the prism The wavelength ranges the resolving powers and the resolutions of the grisms are given in Table 2 3 The medium resolution grism gives about twice the resolving power of the two low resolution grisms It is used with the H and K filters as order sorting filters in the 3rd and 4th orders to cover respectively the H and K atmospheric transmission windows The wavelength ranges the resolving powers and the resolutions of the grisms are given in Table 2 4 The wavelength ranges are defined by the H and K filters The grism can also be used in the higher orders with the J and Z filters However there is significant overlap between these orders so the useful wavelength range is limited Furthermore the line profile degrades in the blue so that the resolution is not significant
112. g Slit Curvature Spectra taken with all SOFI grisms show slit curvature For the red and blue grisms the curvature amounts to a few pixels from the middle of the slit to the edge and is well fitted with a quadratic For point source observations removing the slit curvature is an unnecessary step However extended objects may require to correct it in order to avoid degradation of the spectral resolution The correction is done by doing a 2 dimensional wavelength solution to arc spectra and both MIDAS and IRAF have tasks to do this For the red and blue grisms a quadratic in the slit direction and a cubic or quadratic in the wavelength direction is adequate There are no cross terms An excellent description of the method and the IRAF tools to do it can be found in the cookbook 2D Frutti Reductions with IRAF by Mario Hamuy and Lisa Wells The relevant section is cited verbatim as Appendix A in A User s Guide to Reducing Slit Spectra with IRAF by Phil Massey Frank Valdes and Jeannette Barnes These documents are available on the WWW but for user s convenience we have placed copies of them at http www eso org sci facilities lasilla instruments sofi doc index html For the medium resolution grism the slit curvature is larger As yet we have not tried to do a 2 dimensional wavelength solution to arc spectra taken with this grism 58 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 5 3 5 Arcs For the red and blue low resolution grisms it is suf
113. g slit spectroscopy with the medium resolution grism DARK Dark exposures IMAGING LF FE Imaging with the large field objective and the focal elongator decommissioned UNDEFINED Undefined Mode Table 3 15 The currently supported SOFI observing modes keywords either when BOB is executing the template or in the FITS image headers where some of these keywords are stored As an example setting the filter on filter wheel 1 is done with the signature Filter wheel 1 whereas the corresponding keyword is INS FILT1 ID 3 6 5 File Naming and Exposure Number The parameter Exposure Name determines the name the output files will be given on the instru ment workstation The raw data become available to the user under different names consisting of an abbreviation of the instrument and a time stamp indicating the time the file was created i e SOFT 2005 02 03T10 17 22 354 fits where the time is UT However the exposure name is stored in the fits header and some users may find it convenient to use it in order to rename the files to somewhat more intuitive names for easier data reduction The exposure names consist of a base name defined by the parameter Exposure Name plus an extension of the type _0003 fits The extension is automatically set If the base file name is found on disk the sequential number is automatically incremented For example if there exists a file fileName_0024 fits on disk the next file generated by the tem
114. ge H and K band spectroscopy with fixed width slits of 0 6 1 and 2 arc seconds and slit length of 4 92 arcmin e 0 9 2 5 micron imaging polarimetry with the large field objective 0 288 arcsec per pixel and the set of filters available in imaging mode This manual is divided into several chapters For proposal writers it is sufficient to read chapter 2 giving a general description of the instrument For those who will be observing at the telescope it is sufficient to read up to the end of chapter 4 Chapter 3 discusses how to use observation templates to set up observations and chapter 4 describes how to use the instrument at the telescope For those who will reduce data taken with SOFI it is sufficient to read chapters 2 3 and 5 Chapter 5 discusses how to reduce data taken with SOFI At the end there are appendixes which contain various information useful for observers detailed template descriptions and examples fits headers information calibration data etc 1 1 A First and Final Word The authors of the manual hope that you find this document useful in writing SOFI proposals or for preparing for your SOFI run The manual is continually evolving Some sections still need to be written and some figures still need to be included If you have any suggestions on how to improve SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 the manual please contact the La Silla Science Operations department lasilla eso org or the La Silla coordi
115. ge 3 min and all other presets 6 min the more challenging spectroscopic acquisitions of faint targets can require up to 5 10 minutes e the time necessary to offset the telescope between different jitter positions and any remaining time to complete the last DIT that was started during the telescope offset recall that the sequencer is not interrupted unless the DIT is changed in case of guiding there is an extra overhead necessary to reacquire the guiding star therefore the intervals between the offsets have to be as long as possible to minimize these losses and this can be achieved with increasing NDIT i e NDITxDIT 2 3 min for imaging in case of excellent sky conditions but not too long as to compromise the sky subtraction for practical purposes the average telescope movement overhead is estimate to be 15 sec e there is a 0 1 second delay between the reset of the array and the first read at every DITand 1 7 sec to read the array this implies that to minimize the overhead the user should try to use as long DIT as possible but not too long as to saturate the detector or to work in the non linear regime above 10000 ADU therefore each DIT costs 1 8 sec of overhead if the entire detector is being read e the movement of the telescope rotator that is used to change the orientation of SOFI is relatively slow but it can not be quantified because it is never known in advance what would be the preceding rotator position and how much it w
116. ges into the combination step However the imcombine task can carry out only integer shifts and this will degrade the spatial resolution of the image This is the reason we recommend a 2 step shifting fist with fractional offsets and then with integer offsets see Section 5 2 6 The only possibility to apply the bad pixel mask in this case is before any of the offsets as described here 5 2 3 Subtracting the Dark Bias Frame First and foremost a SOFI dark frame changes with the DIT so if a dark frame is subtracted from other frames the DIT must be identical Secondly the underlying bias pattern to any image is a function of the flux incident on the array so it is not always useful to subtract separately the dark frame In fact the bias and the dark contributions to the total signal are almost impossible to separate without special efforts In cases where the mean count level is low for example narrow band images the dark frame is a good approximation to the underlying bias pattern In cases where the mean flux level is high for example broad band images the dark frame is a poor approximation to the underlying bias pattern so subtracting it serves no real purpose Therefore we suggest to avoid separate dark subtraction Instead the user can subtract the darks together with the sky as described in the next section The advantage of this approach is that the overall illumination is nearly constant so the varying bias level problem m
117. good flat field for SOFI data is difficult The simplest way of creating flat fields is to subtract an image of the dome flat field screen with the dome lamp off from an image with the dome lamp on This flat field has two short comings First the shade pattern of the array is a function of the overall flux so the shade pattern in the image with the lamp on is different from that in the image with the lamp off Thus the difference of the two will contain a residual shade pattern Second the illumination of the dome panel is slightly different from that of the sky Furthermore it changes with time because of the aging of the lamps used to illuminate the screen and in the case of observations with A22 3 micron due to the variations of the dome temperature Both these effects are at the 1 296 level and both can be removed The residual shade pattern can be removed if the SOFI img cal SpecialDomeFlat template was used see sec 3 5 2 This template takes the usual sequence of images with the dome lamp on and off and in addition it takes the same set of images with the mask wheel vignetting the array Figure 3 3 The vignetted part of the array is relatively free of scattered light so it can be used to estimate and remove the shade pattern However this estimate of the shade pattern is valid for the frame that is vignetted and it is not valid for un vignetted frame The difference can be estimated from a region that is common to both the vignetted
118. gure 4 2 Real Time Display with the Store Fixed Pattern option enabled Note the offset between the negative and the positive star pattern This offset is necessary to avoid self cancellation of the objects in the field The green dot on the top of the image flashes every time when the RTD updates the image The yellow arrow is a remnant from a previous acquisition it indicates the direction and the size of the offset that was carried out e The pick object sub window can determine the centroid and FWHM of selected objects in pixels It is useful for monitoring of the image quality e The statistics sub window can be used to display the statistics of a region that can be defined by the cursor It is useful for checking if the counts in the core of an image are below the non linearity limit of 10 000 ADU e The cuts sub window can be used to plot a trace that can be defined by the cursor which is especially useful to check the count levels in spectra and the general background level Zooming within this sub window is possible with the left mouse button the right mouse button un zooms Last but not least the RTD display is also used with the acquisition templates to place objects of interest into slits in spectroscopic modes into the clear portion of the polarimetric mask in polari metric mode or to a specific position on the array in imaging mode The interaction between the RID and the acquisition templates is done by the TIO after the
119. he overheads can easily reach 20096 Furthermore the internal reflections inside the telescope and the instrument will cause optical ghosts from the bright source that are not negligible and therefore will also require the minimum possible DIT to minimize the ghosts On the contrary the faint star calls for just the opposite strategy use the longest DIT that does not saturate the detector array just from the sky which often can dominate the signal even at the peak pixel of the faint star s image and move the telescope as little as possible to minimize the overheads The minimum DIT in combination with the simple jittering and offset templates can certainly help to minimize the crosstalk and the ghosts but they will not move them across the field of view In other words the affected areas will be the same on all jittered or offsetted images with respect to the bright source This may significantly affect the scientific output of the observations i e the depth of the image will vary there will be cross talk produced star like artifacts that will move together with the bright star making any common proper motion studies unreliable One solution to this problems is to brake the observations into a few series taken at different position angles i e three series of images divided by 120 degr While this will certainly help to minimize the 26 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 effects an alternative is to acquire all the i
120. he subtraction of the sky will also result in the removal of the contribution of the sky polarization to the net polarization of the object Therefore the sky has to be estimated and subtracted independently for both the upper and the lower field Finally the remaining instrumental polarization is expected to be mainly caused by the reflection on the tertiary mirror M3 It should therefore depend on the altitude angle of the telescope However the remaining instrument polarization was found to be 0 396 for the K band and 0 4 for the J band see Appendix B of the technical report for more details Since the statistical error of the results from which these limits have been derived is in the same order of magnitude a possible contribution by the mirror M3 could not be extracted If you plan to do polarimetric observations with SOFI we recommend to read carefully the three following reports http www eso org sci facilities lasilla instruments sofi tools report ps http www eso org sci facilities lasilla instruments sofi tecdoc tech rep polarimetry ps http www eso org sci facilities lasilla instruments sofi tools reduction polarimetry http www eso org sci facilities lasilla instruments sofi tools reduction polarimetry sofipolmode html Please note that this manual gives only a short description consult the polarimetry reports from the SOFI web page for more details Appendix A Calibration Arcs The adapter contains both
121. hed standard star spectrum and then multiply by the smoothed absolute energy distribution of the standard In the IR accurate spectroscopic standards do not SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 33 exist For an example of IR flux calibration in the more general case of extended objects see Ivanov et al 2000 ApJ 545 190 The choice of which standard to use depends on which part of the spectrum is of interest All stars have absorption lines so the idea is to choose a star which does not have strong features near the wavelength of interest Hot stars provide relatively featureless spectra however they have strong hydrogen absorption lines so they should not be used as flux calibrators if the region around the hydrogen lines is of interest Later type stars such as G stars have weaker hydrogen lines but are contaminated by multiple weak absorption lines T hese stars have the additional advantage of being very numerous Stars of later type should not be used as these stars contain numerous weak lines throughout their spectra However such stars have very weak hydrogen absorption and may be usefully employed to determine the strength of hydrogen absorption in the hot stars In general IR standards are significantly brighter than optical standards nevertheless stars should be fainter than seventh magnitude A moderately high resolution IR spectral library for stars with well known parameters including spectral types is availabl
122. her Alternatively the standard can be observed in several different parts of the array A pattern of five exposures with the standard observed once in the center of the array and once in each of the quadrants is one example Table 3 7 demonstrates the usage of the SOFI img cal StandardStar template In this template the offsets are along array columns and rows and the units are arc seconds un like other templates where the offsets are given along the RA and Dec The offsets are relative This template is very similar to the template SOFL img obs Jitter Indeed one can use the SOFLimg obs Jitter template to do the same observation however we strongly encourage ob servers to use the SOFL img cal StandardStar template as it is both easier to use and it tags the resulting images as standard star frames for archiving purposes and for easier identification by the zero point calculation tool available at the telescope In general is not required for standard star observations and it is useful to set the Return to Origin parameter to true SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 29 Parameter signature Value Exposure Name sj9104 DIT 5 NDIT 10 Number of exposures 5 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Filter wheel 1 J Filter wheel 2 open Instrument Mode LARGE FIELD IMAGING Combined Offset T F F Return to Origin T F T X offset list arcsec 0 75 150 0 150 Y offset list
123. hlighting of the OBs selected for execution is all the observer needs to do Some observer intervention may be required during the execution of an observing block For example the user may need to select the correct object to be put in the slit or to indicate an offset during the acquisition Everything else is either automatic or it is done by the support staff of the observatory The users are advised to use the available time for inspection of the raw data BOB is a very versatile tool It can be used to display the contents of an OB it can skip templates within an observing block it can be used to pause at a template and it even allows to edit the parameters of an OB but only in the templates that have not been started yet Thanks to this facility it is possible to fine tune the execution of an OB in real time However with this freedom comes the danger to make a mistake especially during a long and stressful observing night Therefore we recommend to make changes within the P2PP and then to re load the edited OB into BOB This provides an extra layer of verification because the P2PP has some built in tools for verification of the OBs BOB runs a Tcl script which sends commands to the OS Observing Software which then sends commands to the TCS Telescope Control System the DCS Detector Control System and the ICS Instrument Control System The status of the instrument is displayed in the OS GUI Graphical User interface and the obtained
124. humidity clouds cover etc Thus in order to accumulate sufficient photons without saturating the detector SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 11 60 F 50 E gt Q E 40 F Figure 2 4 Quantum Efficiency of the SOFI detector at T 78 K The peak Q E is at 1 970 um and the long wavelength cut off is at 2 579 um the observer should acquires many individual integrations and then averages them on the fly into a final frame The control system allows to do exactly that In this context we define DIT the Detector Integration Time as the amount of time during which the signal is integrated onto the detector diodes and NDIT as the number of detector integrations that are obtained and averaged together These averaged frames make up the raw data and in normal co add mode they are the smallest block of data presented to the user So that the total exposure time of a single image single raw data file available to the user is NDIT x DIT Note that the RTD Real Time Display can show either a single DITor the averaged NDIT x DIT Additionally some of the science templates allow the user to obtain several consecutive exposures of NDITxDIT as defined by NINT and NJIT parameters NOTA BENE The counts in a raw data file always correspond to DIT seconds However a single raw data file represents total integration time of NDIT xDIT because the counts in the file are the average of NDIT sub integra
125. ience Templates o o e C 2 3 SOFI Imaging Calibration Templates ens C 3 SOFI Polarimetric Template 2er C 3 1 SOFI Polarimetric Acquisition Template len C 3 2 SOFI Polarimetric Science Template 2l C 3 3 SOFI Polarimetric Calibration Template o 44 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 C 3 4 SOFT Spectroscopic Templates C 3 5 SOFT Spectroscopic Acquisition Templates C 3 6 SOFI Spectroscopic Science Templates C 3 7 SOFI Spectroscopic Calibration Templates D Frame Types E Photometric Standards vii 94 94 95 100 105 106 List of Figures 2 1 2 2 2 3 2 4 2 5 3 1 3 2 3 3 3 4 4 4 2 A 1 A 2 A 3 Optical layout of SOF 2522020442820 8645 bP OX Bex E ep bee Ros Orientation of SOFT for rotator angle 0 deg In imaging the North is to the left and the East to the bottom of the image The slit for the spectroscopy is oriented North South Nota Bene To align the North South axis of the field of view or the slit in case of spectroscopy long a certain Position Angle one has to apply in the acquisition template a rotation angle equal to this Position Angle o SOFI filters The solid red lines indicate the currently available broad band J J H and K filters The short dashed magenta lines are the available narrow band filters The long dashed red lines sh
126. ill have to move so this overhead is simply included in the average acquisition time in cases of templates that rotate SOFI between exposures it is assumed that the angle between the individual exposures will be always small i e less than 15 deg which takes less than 10 sec typically this adds negligible extra time to the time that the telescope takes to move and so it is ignored It is difficult to give a precise estimate of the instrumental overhead but here is a common example A one hour exposure made up of 60 one minute unguided exposures with typical for H K or K DIT 10 and NDIT 6 with telescope offsets for jittering in between will take a total of 80 minutes Guided exposures will take 35 to 5096 overheads Chapter 3 Observing in the IR 3 1 The IR Sky Observing in the IR is more complex than observing in the optical The difference arises from a higher and more variable background by stronger atmospheric absorption and telluric emission throughout the 1 to 2 5 micron wavelength region Short ward of 2 3 microns the background is dominated by non thermal emission principally by aurora OH and O emission lines The vibrationally excited OH lines are highly variable on a time scale of a few minutes Pronounced diurnal variations also occur The lines are strongest just after sunset and weakest a few hours after midnight A complete description and an atlas of the sky emission lines can be found in the paper by Rousselo
127. imaging filters together with the atmospheric transmission are shown in Fig 2 3 see also Appendix B Filter Filter Central Wavelength Width Peak Transmission Name Wheel um um Z 1 0 9 0 140 J 1 1 247 0 290 Js 1 1 240 0 160 H 1 1 653 0 297 83 K 1 2 162 0 275 88 NB 1 061 1 1 061 0 010 NB 1 083 Hel J 2 1 083 0 016 61 NB 1 187 1 1 187 0 010 NB 1 215 2 1 215 0 018 NB 1 257 Fell J 2 1 257 0 019 NB 1 282 PG 2 1 282 0 019 NB 1 644 Fell H 2 1 644 0 025 NB 1 710 2 1 710 0 026 NB 2 059 Hel K 2 2 059 0 028 81 NB 2 090 1 2 090 0 020 NB 2 124 H S1 2 2 124 0 028 78 NB 2 170 Bry 2 2 167 0 028 71 NB 2 195 1 2 195 0 030 NB 2 248 2 2 248 0 030 NB 2 280 2 2 280 0 030 69 NB 2 336 CO 2 2 336 0 031 GBF 1 0 925 cut on GRF 1 1 424 cut on Open 1 and 2 Closed 1 and 2 Table 2 2 The broad and narrow band filters available with SOFT The K short or K filter is different from both the standard K filter and the K filter defined by Wainscoat and Cowie 1992 AJ 103 332 The long wavelength edge of the K filter is similar to that of the K filter but the short wavelength edge is similar to that of the K filter Thus the K filter avoids both the atmospheric absorption feature at 1 9 wm and radiation from the thermal background beyond 2 3 ym The difference between K and K is given by K K 0 005 J K Similarly the J and J filters differ mostly in the long waveleng
128. ime individual exposure sec NDIT DET NDIT NODEFAULT Number of DITs averaged into an individual image Filter wheel 1 INS FILT1 ID NODEFAULT Filter wheel 1 position Filter wheel 2 INS FILT2 ID NODEFAULT Filter wheel 2 position Which Slit INS WHICHSLIT NODEFAULT Slit Alpha Offset arcsec TEL TARG OFFSETALPHA 0 0 Alpha Offset for sky subtraction arcsec Delta Offset arcsec TEL TARG OFFSETDELTA 0 0 Delta Offset for sky subtraction arcsec Add Velocity Alpha TEL TARG ADDVELALPHA 0 0 Additional tracking velocity in RA arcsec sec Add Delta Velocity TEL TARG ADDVELDELTA 0 0 Additional tracking velocity in DEC arcsec sec Rotation Offset on Sky TEL ROT OFFANGLE 0 0 Rotation offset on the sky degrees Combined offset F T SEQ COMBINED OFFSET False T guiding ON F guiding OFF Preset Telescope F T SEQ PRESET True T full preset F fine tunning of the pointing Save Image F T SEQ SAVE False T preserve the acq image F no Table C 34 SOFL_img acq_ MoveToSlit template without presetting the telescope In other words the user can change the position of the target on the array without full telescope preset This option is particularly useful if acquisition was aborted for some reason or if multiple OBs to observe the same target are executed one after another In this case it is advisable to re acquire the target with a false preset every 1 2 hours to ensure it has not drifted away from the slit The par
129. img obs AutoJitterArray template with commonly used values 3 2 5 Imaging of Moderately Large Object This section describes imaging of moderately large objects i e objects comparable to the SOFI field of view We discuss this case after the mapping of large areas because it uses the same template SOFI img obs AutoJitterArray 1 as the area mapping There are two distinct cases i Observations of semi extended objects as large of a quarter to half of the SOFI filed of view these may be moderately distant galaxies small Milky Way clusters or LMC SMC clusters They are too big for simple jittering mode observations but still leave room for more efficient observing strategy where the observer does not need to sample clear sky 5096 of the observing time Instead the user can observe the sky simultaneously with the target adopting clever offset pattern that would move the target in the centers of the four quadrants or in the centers of the two halves of the array In the latter case a suitable rotation offset may be necessary to apply during the acquisition so the side of the array is aligned with the major axis of the object A typical example of a 4 point observation is shown in Fig 3 1 The figure shows the target a round object with diameter 90 arcsec as it will appear on the RTD and on the SOFI images The best choice is to use the template SOFL img obs AutoJitterArray 1l As it is described in Section 3 2 4 this template
130. in the large field imaging mode the grism wheel is set to open the mask wheel is set to the large field mask and the objective wheel is set to the wide field objective In addition the collimator used for focusing the instrument is moved to a predefined position depending on the mode and the filter The standard modes available with SOFI are listed in the following Table 3 15 3 6 4 Template Parameters Signatures and Keywords Template parameters can be described either by their signature which are aimed at being self explanatory or by their keyword which are more meaningful to the OS Observing Software When a template is created with P2PP the signature form of the parameter is displayed When the OB is passed to the BOB Broker of Observing Blocks for execution signatures are translated into keywords Although users will mostly be faced with signatures they will later see the corresponding 40 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Observing Mode Description LARGE FIELD IMAGING Imaging with the large field objective LARGE FIELD SO IMAGING Imaging with the large field objective SMALL FIELD IMAGING Imaging with the small field objective decommissioned POLARIMETRY Polarimetric mode LONG SLIT RED Long slit spectroscopy with the red grism LONG SLIT BLUE Long slit spectroscopy with the blue grism LONG_SLIT_K 2 0 2 3 micron long slit spectroscopy with the medium resolution grism LONG SLIT H 1 5 1 8 micron lon
131. inst larger defocusing that would turn the stars into doughnuts because this increases the relative contribution of the sky degrading the photometric accuracy 106 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 No HST RA J2000DEC J o J H oH K c K Ks oc Ks 9101 P525 E 00 24 28 3 07 49 02 11 622 0 005 11 298 0 005 11 223 0 008 11 223 9103 S294 D 00 33 15 2 39 24 10 10 932 0 006 10 657 0 004 10 596 0 005 10 594 9104 754 C 01 03 15 8 04 20 44 11 045 0 005 10 750 0 005 10 693 0 010 10 695 9105 P530 D 02 33 32 1 06 25 38 11 309 0 010 10 975 0 006 10 897 0 006 10 910 9106 S301 D 03 26 53 9 39 50 38 12 153 0 007 11 842 0 005 11 772 0 010 11 788 9108 P533 D 03 41 02 4 06 56 13 11 737 0 009 11 431 0 006 11 337 0 008 11 336 9109 S055 D 04 18 18 9 69 27 35 11 552 0 002 11 326 0 002 11 255 0 027 11 269 9111 S361 D 04 49 54 6 35 11 17 11 246 0 006 11 031 0 006 10 992 0 033 10 980 9113 S252 D 05 10 25 6 44 52 46 11 059 0 005 10 776 0 005 10 708 0 034 10 713 9115 S363 D 05 36 44 8 34 46 39 12 069 0 007 11 874 0 005 11 826 0 007 11 831 9116 S840 F 9118 S842 E 9119 121 E 9121 S255 5 9125 S005 D 9129 S209 D 9132 S312 T 9133 5495 E 9134 P545 C 9135 S705 D 9136 S165 E 9137 S372 5 9138 S852 C 9140 262 E 9141 S708 D 9143 P550 C 9144 S264 D 9146 S217 D 9147 S064 F 9150 S791 C 9153 P499 E 9154 S008 D 9155 S867 V 9157 S273 E 9160 S870 T 9164 P565 C 9170 S875 C 9172 S279 F 9173 S024 D 9175 S071 D 91
132. ipeline yuca D bee ia dod debug A ud d 40 27 Phe Archive ore Ree ae EGER IRAE Am COs 4 6 3 The Calibration Plan e 4 7 At the End of the Night and at the End of Your Run leen Data Reduction 5h Basic Concepts s lios Be eel Ae Bebo eho eh ES aD eee ee bod ey 5 2 Imagl g ax uet eR Boh ah doen SB AS peers eed m heh ah ee Meo hd xs 5 2 1 Inter quadrant Row Cross Talk 2e 5 2 2 Masking the Bad Pixels ee teda 5 2 3 Subtracting the Dark Bias Frame o o e o 5 2 4 Sky Subtraction opo dh al bk 4 dd beue 5 2 5 Flat Fields and Illumination Corrections 00 5 2 6 Image Alignment and Combination ees 5 3 Long Slit Spectroscopy eee eres 5 8 1 Inter quadrant Row Cross Talk o e e e 5 9 2 OY Subtraction bs tar a er ey Roe Log ges Dido Flat Fields fs 20a ou hohe ae EE E ea bu ue HEN A 5 3 4 Removing Slit Curvature lee DISD ALCS ost hee a Bea ae Mee A ha S UY Ba eal ne GS 2 5 3 6 Removing of the Atmospheric Absorption Features and Flux Calibration 5 8 7 Alignment and Combination 2 2 0 0 0 eh 544 Polarimetry x4 esae Seo ee he ele ae ed Noe EA vue n d da A Calibration Arcs B Atmospheric Absorption C SOFI Templates A Reference Guide Cl General Points ee Ae ay ee a ee wk gd Rk Gd fa Dh De C 2 SOFI Imaging Templates ee C 2 1 SOFI Imaging Acquisition Templates o oo o C 2 2 SOFI Imaging Sc
133. istributed randomly within a box whose size defined by SEQ JITTER WIDTH in arc seconds with the condition that the distance between any two points in a series of ten values is greater than certain minimum specified withing the template This is intentionally done to ensure that the 5 frames before and after any frame are not too close spatially and can be safely used for creating a sky image without large residuals for the sky subtraction By default there is no telescope offset before the first exposure If SEQ RETURN is set to True T the telescope slews back to its original position at the end of the template if not the telescope is not SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Parameter signature Value Exposure Name NGC6118 DIT 20 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Number of Exposures T List of NDIT 3232323 Filter wheel 1 Ks Filter wheel 2 open Instrument Mode LARGE FIELD IMAGING Return to origin T F T RA offset list arcsec DEC offset list arcsec 0 150 140 150 140 250 250 0 150 140 150 140 250 250 Obs Type O or S OSOSOSO Guiding N B S B Table C 23 SOFLimg obs GenericImaging Example moved This feature is useful if more exposure of the same field are taken i e with different filters without running the corresponding acquisition template In this case the parameters should be set to True To minimize the
134. it The SOFI calibration unit is located inside the telescope adapter It contains a halogen lamp for internal spectroscopic flat fielding Xenon and Neon lamps for wavelength calibration Line identifi cations are given in the Appendix A The halogen lamp is used to flat field spectroscopic data These flats are called Nasmyth flats Note 14 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 that the Nasmyth flats introduce an extra spatial response by the lamp that has to be removed by additional night sky spectroscopic flat The halogen lamp is not suitable for flat fielding of imaging data Use the dome flat field lamp for this see Sec 3 5 2 for details 2 7 Instrument Performance and the Exposure Time Calculator Table 2 5 we list the approximate zero point of the broad band filters with the large field objective as measured over many nights starting from July 1998 to May 2005 Limits are similar for the small field and spectroscopic objectives For more recent detection limits in the spectroscopic modes and for detection limits with the refer to the SOFI web page Filter ZP Average Background Detection Limit Mag sq arc second Point Source Extended Source Z 22 6 is E xi J 23 2 15 5 16 1 22 7 22 1 Js 23 1 is Er n H 23 0 13 4 14 7 21 8 21 4 K 224 12 8 13 3 20 8 20 4 Table 2 5 Measured SOFI performance for the broad band filters 1 hour exposure The point source detection limits are based on the following assumptions
135. ith 30 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 offsets in the X direction to help with sky subtraction To measure the Stokes parameters and hence the degree and direction of linear polarization one needs at least one additional set of observations with a position angle different from the first The usual practice is to take two sets of observations with a rotational offset of 45 degrees This can be done with the Rotator Offset parameter The approximate rotation centers for the imaging modes are listed in Table 3 9 Objective Rotation Center x y Large Field 517 504 Small Field 501 505 Spectroscopic Objective 537 502 Table 3 9 Approximate mechanical rotation centers for the imaging modes Note that these values change after each instrument intervention There are two ways to do this In the first method one includes two observation templates for each polarimetric observation one with the Rotator Offset angle set to zero and the second with the Rotator Offset angle set to 45 Alternatively one can set the rotator angle through the acquisition template See Appendix C and keep the Rotator Offset angle zero in both cases Since the rotator axis does not correspond to the center of the array the later method is preferable when one is trying to determine the linear polarization of a single object For those who wish to map the polarization over a large field either method can be used 3 4 Spect
136. k scattered light contribution Note that there is no need to take the arc calibration with non destructive readout because the DIT can always be increased to ensure that the arc spectrum has a sufficient signal to noise ratio There is also no reason to match the readout mode of the arc calibration with the readout mode of the data SOFI spec cal DomeFlats and SOFI spec cal NonDestrDomeFlats Spectroscopic dome flats are taken with these two calibration templates Table C 44 and C 45 The only difference between the two templates is the readout mode it has to match the readout mode used to obtain the scientific observations For each element in SEQ DIT LIST and SEQ NDIT LIST the templates take four images the first with the dome lamp off the next two with the dome lamp on and the fourth one with the dome lamp off The intensity of the dome lamp is controlled manually Note that in case of SOFLspec cal NonDestrDomeFlats it is not crucial that NSAMP and NSAMPPIX are the same for the science frames and calibrations In fact this is sometimes impossible because the DIT of the scientific exposure is rather long often in order of 5 10 min Good values for the parameters of the NonDestrDomeFlat templates are DIT 10 NSAMP 6 NSAMPPIX 4 SOFT User s Manual 2 3 Parameter signature Exposure Name Number of columns Number of rows First column of window First row of window Number of exposures List of DITs List of NDIT Spectral Mo
137. l images are obtained the uniformity of the flat field illumination can be checked Several standard star lists are currently available but it should be noted that each list was created using detectors and filters that differ from the detectors and filters used with SOFI Note also that most standards were observed with single channel photometers with very wide apertures Thus close companion stars were probably included The standards listed in Carter and Meadows 1995 MNRAS 276 734 have proved to be very useful although they may prove to be too bright for SOFT and to require defocussing the telescope The more recent NICMOS standards by Person et al 1998 AJ 116 2475 are more suitable as one does not need to defocus the telescope except for seeing better than 0 6 arcsec They are the standards of choice for most observers Predefined OBs are available at the telescope for most of these stars A list of references to these and other standard star lists as well as useful papers regarding the transformations between one photometric system and another are available from the SOFI web page It is important that you take care where the standard star lies on the array Use the RTD Real Time Display to find areas that are clear of bad pixels Furthermore check if other objects are in the field The offset between exposures should not be such that other objects in the field interfere with flux of the standard when one frame is subtracted from anot
138. l signal to noise ratio because the noise will be dominated by the sky There are two standard techniques to estimate the sky The first is appropriate for angularly large objects or crowded fields and the second is appropriate for angularly small in comparison with the field of view objects or uncrowded fields 3 2 1 Selecting the best DIT and NDIT Selecting the best DIT and NDIT is a complex optimization problem and it depends on the nature of the program targets required signal to noise frequency of sky sampling etc Therefore it is hard to give general suggestions and the users should exercise their judgment and discuss their choices with the telescope operator who is supporting their run The first constraint is to keep the signal from the target on the linear part of the detector array dynamic range which is below 10 000 ADU The minimum DIT of 1 182 sec allows to observe without problem stars of 10 mag under average seeing humidity conditions Keeping in the linear regime any bright stars that may have randomly fallen into the filed is also desirable because these stars will become a source of horizontal lines due to the cross talk The cross talk is correctable well only if the sources that cause it are not saturated Unfortunately it is not always possible to use small DIT because the smaller DIT increases greatly the overhead up to 50 in the cases of 1 2 2 sec DIT For comparison observations with DIT 8 sec and NDIT 8 o
139. l telescope preset This option is particularly useful if acquisition was aborted for some reason or if multiple OBs to observe the same target are executed one after another The parameter SEQ SAVE allows to save an acquisition image at the prize of some additional overhead for file transfer and saving The interactive pop up windows are usually displayed before new images have arrived on the RTD Therefore operators are strongly advised to carefully check that a new image has arrived before clicking on these windows e g for storing a fixed pattern for changing the DIT NDIT configuration The arriving of a new image on the RTD is marked by a flashing green dot in the middle of the upper part of the RTD window 74 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 C 2 2 SOFT Imaging Science Templates SOFIimg_obs_AutoJitter This observation template Table C 7 offsets the telescope between exposures according to a random pattern of offsets automatically generated within the template It is ideal for long integrations on empty fields or fields containing isolated point sources and does not require a long list of offsets to be defined by the observer The offsets are distributed randomly within a box whose size is defined by SEQ JITTER WIDTH in arc seconds with the condition that the distance between any two points in a series of ten values is greater than certain minimum specified withing the template This is intentionally done t
140. lay with the Store Fixed Pattern option enabled Note the offset between the negative and the positive star pattern This offset is necessary to avoid self cancellation of the objects in the field The green dot on the top of the image flashes every time when the RTD updates the image The yellow arrow is a remnant from a previous acquisition it indicates the direction and the size of the offset that was Carried outre o A A oe ee a oa a ee A Xenon arc spectrum taken with the blue grism The main lines are marked A Xenon arc spectrum taken with the red grism The main lines are marked A Xenon and Neon arc spectrum taken with the medium resolution grism at the Z atmospheric window The main lines are marked viii SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 ix AA A 5 A 6 B 1 C 1 E 1 E 2 E 3 E 4 E 5 A Xenon and Neon arc spectrum taken with the medium resolution grism at the J atmospheric window The main lines are marked o 65 A Xenon and Neon arc spectrum taken with the medium resolution grism at the H atmospheric window The main lines are marked o 66 A Xenon and Neon arc spectrum taken with the medium resolution grism at the K atmospheric window The main lines are marked o 67 The atmospheric transmission at a resolution of 8A o o o 69 Positions of the offsets listed in Tables C 42 and C
141. lly the offsets for each individual image However this is inconvenient if the number of images is large To simplify the matter we suggest to the user to define a closed loop where after four images the telescope points back to the original position This means that the sum of the offsets along the RA is zero same as the sum of the offsets along the DEC Fortunately the autojitter feature of the template ensures that the the next batches of four images will not be taken at exactly the same positions as the previous ones In the example the jitter box is 20 arcsec and for simplicity it is shown only on the first of the four positions Four dashed circles show where the target would be if the random jitter offsets move the target center to the edges of the jitter box This has to be remembered well because too large a jitter box may move the target out of the field of view or it may lead to overlapping of the target edges in sequential images The overlap is better to be avoided because it may affect the sky subtraction The user should try to keep the target as close to the center as possible without allowing the overlap SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 23 Parameter signature Value Exposure Name SOFLA point DIT 6 NDIT 10 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 NJITT 12 NEXPO 4 Jitter Box Width arcsec 20 Filter wheel 1 K Filter wheel 2 open Instrument
142. ly better than that obtained with the low resolution blue grism Three slits of different fixed widths 0 6 1 0 and 2 0 arc seconds are available The slit length is 290 arc seconds SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 9 Grism Order Sorting Wavelength Resolving Dispersion Number Filter Range microns Power A pixel Blue GBF 0 95 1 64 930 6 96 Red GRF 1 53 2 52 980 10 22 Table 2 3 The wavelength range and the resolution of the low resolution grisms The resolution is that measured for the 0 6 arc second slit The resolution scales inversely with the slit width Grism Order Order Sorting Wavelength Resolving Dispersion Name Filter Range microns Power A pixel 3 Ks 2 00 2 30 2200 4 62 4 H 1 50 1 80 1500 3 43 5 J 1 20 1 28 1400 2 71 6 J 1 17 1 24 1400 2 22 T Z 0 89 0 93 1400 1 87 8 Z 0 86 0 95 1400 1 58 Table 2 4 The wavelength range and the resolution of the medium resolution grism The resolution is that measured for the 0 6 arc second slit The resolution scales inversely with the slit width 2 4 Polarimetry SOFI offers imaging polarimetry A Wollaston prism in the grism wheel splits the incoming parallel beam into two beams that are perpendicularly polarized The beams are separated by 48 arc seconds Thus an image taken with the Wollaston prism will contain two images of every object To avoid sources overlapping a special mask consisting of alternating opaque and transmitting strips can be inserted at
143. mages of the object at different angles True it will add an extra overhead for moving the rotator between the exposures but this extra overhead is partially compensated by having only one acquisition as opposed to the few extra acquisitions that will be associated with the different series of images Some extra taxes to be paid when using this strategy are e complication in the data reduction which will include not just the usual image alignment but derotation as well e degradation of the image quality toward the edges of the field of view because the field distor tions differ in the different parts of the array e depending on the used rotation procedure the corners of the field of view may be lost so the final image will be a circle with 4 92 arcmin diameter in case of Large Field mode and not a 4 92x4 92 arcmin square Given these limitations the templates that rotate the instrument for every exposure are suitable only for imaging of faint targets within 1 arcmin from a bright source The templates that perform the observations described here are called SOFLimg obs AutoJitterRot and SOFLimg obs JitterRot 3 2 7 Objects with Fast Variability The typical observational problems addressed here are occultations transits and fast variables In the normal case the minimal DIT is 1 182sec and it is determined by the fraction of the array being read i e the detector windowing The normal templates do allow to window the det
144. meter signature Exposure Name DET DIT DET NDIT Number of columns Number of rows First column of window First row of window Number of exposures Filter wheel 1 Filter wheel 2 Instrument Mode Combined offset F T Jitter Box Width arcsec Return to Origin T F Header Keyword DET EXP NAME DET DIT DET NDIT DET WIN NX DET WIN NY DET WIN STARTX DET WIN STARTY SEQ NEXPO INS FILT1 ID INS FILT2 ID INS IMODE SEQ JITTER WIDTH SEQ RETURN LSO MAN ESO 40100 0004 75 Value SOFI NODEFAULT NODEFAULT 1024 1024 1 1 1 NODEFAULT NODEFAULT NODEFAULT SEQ COMBINED OFFSET F 40 T Table C 7 SOFLimg obs AutoJitter Parameter signature Exposure Name DIT NDIT Number of columns Number of rows First column of window First row of window Number of Exposures Filter wheel 1 Filter wheel 2 Instrument Mode Combined offset T F Jitter Box Width arcsec Return to origin T F Value NGC6118 10 6 1024 1024 1 a 6 Ks open Description File name prefix Detector Integration Time individual exposure sec Number of DITs averaged into an individual image Number of columns in the window Number of rows in the window First column of window First row of window Number of exposures in the sequence Filter wheel 1 position Filter wheel 2 position Instrument Mode T guiding ON F guiding OFF Jitter box size Returns the telescope to the original pointing if True
145. mination correction surface refers to a particular special dome flat and it can not be used with another special dome flat Special Dome Flats and Illumination Correction Surfaces for all broad band filters for Large Field mode are prepared by the observatory staff monthly The latest versions can be downloaded from the Web Page of SOFI ftp ftp eso org web sci facilities lasilla sofi fits Archive and fits images for a number of 16x16 grids can be found at ftp ftp eso org web sci facilities lasilla sofi fits Archive Illum Corr An IRAF script for producing Illumination Correction Surfaces illumination cl is offered to the users It can be downloaded from the SOFI web page http www eso org sci facilities lasilla instruments sofi tools reduction sofi scripts Twilight sky flats and flats created from the observations themselves can also be used to flat field the data However these frames suffer from the same problems as the uncorrected dome flats and in the case of the shade pattern it is not clear how they can be corrected Last but not least the Special Dome Flats and the Illumination Correction Surfaces are mode and filter specific ie they have to be prepared separately for each instrument mode and filter combination 5 2 6 Image Alignment and Combination Most data reduction packages have tools for alignment of images and for their combination For example in IRAF one can use imexam imalign and or imcombine We refe
146. move the cross talk only if the source that causes it is not saturated In case of saturation the counts of the source will be decreased by the non linearity and the subtraction of the ghost will be incomplete However even in those cases the users can improve the data The cross talk manifests itself on the final reduced image as i residual of the ghost and ii residual of the line leaks i e horizontal lines with increased counts While there is nothing one can do for the former effect short of masking the latter one can be removed This can be achieved in a few steps i block average the image along the horizontal lines note that the IRAF task blkavg is not well suited for this purpose because it does not allow to use median averaging or to impose rejection and lower threshold limits Instead it is better to use imcombine feeding it with a list that contains each row of the image as individual entry final image 1 final image 2 final image 3 etc The use of a median averaging and a suitable rejection algorithm with low value of the upper threshold is necessary to exclude from the averaging any sources The result from this step will be a 1 dimensional vector image ii expand the vector image to the full size of the final image by replicating it the IRAF task blkrep is well suited for this purpose The result from this step is a 2 dimensional image with a size matching the size of the final image iii subtract the im
147. nDestr 32 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 3 4 3 Telluric Standards and Flux Calibration The issue of flux calibrating IR spectra is still very much an issue of some debate This section is written not as the definitive method to flux calibrate IR spectra but rather it is written to illustrate the problems involved and various methods for their solution As the IR window is dominated by time varying atmospheric absorption features that depend non linearly with airmass it is necessary to divide object spectra with the spectrum of what we shall call a telluric standard The standard should be at an airmass that is as close as possible to that of the science target usually Asecz lt 0 05 0 1 it should be observed immediately before or after the target and it should be observed with the same instrument set up grism slit objective etc From a technical point of view the telluric standards are observed in exactly the same way as the science targets Since they are usually bright stars with no nearby companions of comparable magnitude the template SOFI spec obs AutoNodOnSlit is perfectly suited for this It is very important that object and standard star spectra are accurately aligned A small misalign ment will result in poor cancellation of the atmospheric absorption features Misalignment could be caused by slit misalignment in between the observations by instrument flexure or when the seeing is smaller than the slit
148. nator ls coord eso org Nota Bene 1 2 1 2 3 4 5 1 3 Some of the SOFI modes may be decommissioned during the life time of the instrument See the SOFI web page for the latest instrument status Their description remains in the manual for the benefit of users who may download and use old SOFI data from the ESO archive The SOFI Template Signature Files hereafter TSF Parameters Reference Guide is suppressed by this manual see Appendix C Those of you who are successful in being awarded SOFI time must read this document and the related P2PP Users Manual before coming to observe with SOFI This document contains a data reduction cook book Chapter 5 with detailed description of the individual data reduction steps The users must read it carefully before preparing their observations because the good understanding of the data reduction will help you to plan the observations better and more efficiently There is also a WEB page dedicated to SOFI http www eso org sci facilities lasilla instruments sofi index html It is accessible go to Instruments from the main web page of the Science Operations depart ment for La Silla http www eso org lasilla sciops There you will find the most up to date information about the instrument recent news effi ciency measurements and other useful data that do not easily fit into this manual or a subject of frequent changes This WEB page is updated regularly Applicable documents V
149. ng ON F guiding OFF Returns the telescope to the original pointing if True list of offsets along rows list of offsets along columns Rotator offset degrees SOFT User s Manual 2 3 Parameter signature Exposure Name DIT NDIT Number o Number of rows First column of window First row of window Number of Exposures Filter wheel 1 Filter wheel 2 Instrument Mode Combined offset F T Return to origin T F X offset list arcsec Y offset list arcsec Rotator offset columns LSO MAN ESO 40100 0004 93 Value NGC6118 20 3 1024 1024 1 1 5 Ks open LARGE FIELD IMAGING T T 0 50 0 100 0 0 50 100 0 100 0 Table C 32 SOFLimg obs Polarimetry Example Parameter signature Exposure Name DIT individual exposure NDIT number of DIT Number of columns Number of rows First column of window First row of window Number of exposures List of Rotator angles Filter wheel 1 Filter wheel 2 Return to Origin T F Header Keyword Value Description DET EXP NAME SOFI File name prefix DET DIT NODEFAULT Detector Integration Time individual exposure sec DET NDIT NODEFAULT Number of DITs averaged into an individual image DET WIN NX 1024 Number of columns in the window DET WIN NY 1024 Number of rows in the window DET WIN STARTX 1 First column of window DET WIN STARTY 1 First row of window SEQ NEXPO 1 Number of exposures in the sequence SEQ ROT OFFANGLE NODEFAULT Rotator offset lis
150. nts with an error of a few arc second Most observers find this template sufficient for the purposes of their programs SOFI_img_acq MoveToPixel This acquisition template Table C 6 has two parts First it presets the telescope to the coordinates of the Target associated with the Observation Block Next it requests the operator to interact with the RTD so that the user specified object is moved to a user specified location on the array In order for objects to be clearly seen the telescope can do a small offset after the preset The size and direction of the offset are template parameters TEL TARG OFFSETALPHA and TEL TARG OFFSETDELTA The operator is then prompted to store a fixed pattern that is subsequently subtracted from incoming images via the real time display RTD features After storing the fixed pattern the telescope returns to the original position At this point in time the operator is offered the possibility to change DIT and NDIT This is useful if the target can not be identify securely because of the low signal to noise If the operator changes the values the telescope offsets again and the operator is required to store another fixed pattern before the telescope returns to the nominal position This loop can continue until the operator has identified the target Parameter signature Header Keyword Value Description DIT DET DIT NODEFAULT Detector Integration Time individual exposure sec NDIT DET NDIT NODEFAULT N
151. o ensure that the 5 frames before and after any frame are not too close spatially and can be safely used for creating a sky image without large residuals for the sky subtraction By default there is no telescope offset before the first exposure If SEQ RETURN is set to True T the telescope slews back to its original position at the end of the template if not the telescope is not moved This feature is useful if more exposure of the same field are taken i e with different filters without running the corresponding acquisition template In this case the parameters should be set to True If SEQ COMBINED OFSET is set to T and if guiding was started before the start of the template guiding is Boz To Star Considering that in imaging mode the time between the offsets is limited typically to 1 3 min by the sky background variation the guiding is rarely used in imaging mode The value of SEQ JITTER WIDTH corresponds to the full width of the box in which the offsets are generated The choice of box size is a compromise between to constraints 1 Too wide a box may lead to insufficient image overlap The signal to noise ratio will degrade toward the edged of the final combined image 2 Too small a value may lead to poor sky subtraction near extended objects including bright stars with noticeable wings Setting this parameter to zero is equivalent to staring the observations Consider the following example Table C 8 the template will pro
152. oD sto 0 eo co o c co st LO C2 cD cO o C ES st a e QQ aO O NM co H c yaj e NO Q2 i e DO M ea en e Qu Q2 QQ QQ Na cw N NA Qu o o o o o oo wo o o o QD ou Z A 6x10 E i y ied 01 mS A El o ES E o F 24x10 3 sed V E 2x104 remm L 1 L L 1 L L 1 L L 1 2x104 2 1x10 2 2x104 2 3x10 Wavelength A Figure A 6 A Xenon and Neon arc spectrum taken with the medium resolution grism at the K atmospheric window The main lines are marked Appendix B Atmospheric Absorption The atmospheric transmission in the 0 8 to 2 5 micron region is plotted as a function of wavelength Fig B 1 together with pass bands for the SOFI filters 68 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 Transmission OS T 0 85 0 9 0 95 1 1 05 14 1 15 1 2 Transmission Transmission Transmission oe 2 1 2 15 22 2 25 2 3 2 35 24 2 45 2 5 Wavelength microns Figure B 1 The atmospheric transmission at a resolution of 8 Appendix C SOFI Templates A Reference Guide C 1 General Points This section provides a summary of the currently released SOFI templates It assumes that the reader is already familiar with the capabilities of the instrument Information about SOFI may be found on line at the NTT Web site http www ls eso org Note that undocumented and unreleased prototype templates may be available at the
153. of window First row of window DET WIN STARTY 1 First row of window Number of exposures SEQ NEXPO 1 Number of exposures in the sequence NJITT SEQ NJITT 1 Number of Jittered exposures around each Array position Jitter Box Width arcsec SEQ JITTER WIDTH 40 Jitter box size Filter wheel 1 INS FILT1 ID NODEFAULT Filter wheel 1 position Filter wheel 2 INS FILT2 ID NODEFAULT Filter wheel 2 position Instrument Mode INS IMODE NODEFAULT Instrument Mode Combined offset F T SEQ COMBINED OFFSET F T guiding ON F guiding OFF Return to Origin T F SEQ RETURN T Returns the telescope to the original pointing if True RA offsets list arcsec SEQ OFFSETALPHA LIST NODEFAULT list of offsets along RA Dec offsets list arcsec SEQ OFFSETDELTA LIST NODEFAULT list of offsets along Dec Table C 20 SOFLimg_obs_AutoJitterArray and SOFLimg_obs_AutoJitterArray_1 rameter set and it takes the same number of images as the original template but in a different order exposures are taken at the positions of the array as defined by the offsets lists then the random offset is applied and the entire array pattern is executed again This template is preferable to the original one because it gives better sky subtraction Also if the weather conditions are poor i e the sky background variations are strong there are thin clouds etc the entire mapped area is imaged under relatively more similar conditions in comparison with the SOFI_img_obs_AutoJitterArray tem
154. om stellar images will remain on the sky subtracted images Indeed with only the preceding and the succeeding images to subtract the sky it is impossible to remove the negative stars This difference with the previous method makes us to consider it separately Note that the negative stars can be dealt with later on the stage of combining the shifted object images Given enough object images and sufficiently low crowding this technique works well This method is applicable for observations taken in both simple jittering mode and with offsets on a clear sky region A special attention is required to observations obtained following the schemes described in Sec tion 3 2 5 As pointed there the sky sampling in those observing strategies is somewhat rarefied and SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 55 the user should check the residuals from the sky subtraction In case of observations in which the target has been kept on the array during every image the user should mask the target out before constructing a sky image This can be done in many ways perhaps the most straightforward solution is to follow the same procedure as for masking out the bad pixels In the case of 4 point observations described in Section 3 2 5 you have to create four pairs of masks one with zeros in the target quadrant and one everywhere else and another with a large negative value ie 1e6 at the target quadrant and zero everywhere else Then
155. ost of the imaging templates here the offsets are defined along detector rows and columns in arcsec so that the users can move the object easily along the strips of the polarimetric mask Most importantly the user have to decide which angle are necessary one should observe the same object at several angles to determine the Stokes parameters cf section 2 4 All frames can be given an orientation relative to the previous position angle by setting SEQ ROT OFFANGLE parameter in degrees When the template starts the instrument is rotated on the sky by SEQ ROT OFFANGLE and remains at this position until the end of the template After the last exposure the instrument is rotated back to the original position With this scheme it is possible for the user to sample the object and the sky as desired for one rotator position and then restart the template with another orientation on the sky for another series of exposures At least two different orientations separated by 45 degrees are required for computing the Stokes parameters This implies that the template must be called at least twice within an Observation Block or with two different Observation Blocks both with two different rotator positions The most likely situation will be to set SEQ ROT OFFANGLE to 0 degrees in the first template and then to 45 degrees in the second template Guiding is possible defaulted to Box To Star if SEQ COMBINED OFSET is set to T and if guiding was start
156. ote that the rotation offset has to be defined in the acquisition template Also the acquisition has to put the telescope at the location marked on the figure with zero However the first image in the sequence is actually a sky on the location marked with number one The parameter NEXPO has to be set to six The sequence in which the images are taken is shown in the figure on the right The target is marked with a thick dark line and it is covered by two overlapping fields The jitter boxes around each position are not shown One 6 point cycle will obtain two images of 1 min at each target field 12 cycles will accumulate 24 min of integration on each of the two galaxy fields and 12 min on each of the two sky fields The total observation will require about 1 5hr including the overheads 1000 E North 4 10 P o 500 H A o EI L o PA 30 exe po f p 7 Eo o Sod ius n n e s 0 o El eo j t S 0 2 6 8 12 e L o o F Nn o u 120 arcsec 500 r S F V 1 7 13 1000 1000 500 0 500 1000 R A arcsec Figure 3 2 Example of a 6 point observation of semi extended objects alternating between two target fields and one of two different skies Nota bene The reduction of data obtained with the patterns described in this subsection is usually more complicated and requires special attention and more manual intervention than the classical observations In addition the
157. ow the soon to be commissioned J filter top panel and the typical broad K filter Persson et al 1998 AJ 116 2475 see their Table 10 The dotted blue line is the atmospheric transmission model for Mauna Kea for Am 1 0 and water vapor column of 1mm Lord S D 1992 NASA Technical Memorandum 103957 courtesy of Gemini Observatory The data for the plot and a SuperMongo script are available from the SOFT web page Quantum Efficiency of the SOFT detector at T 78K The peak Q E is at 1 970 um and the long wavelength cut off is at 2 579 um oes Deviation from the linearity of the SOFI detector measured fraction of the expected level for linear response versus the measured illumination level in log of ADUs for May 1052012 EE nS A SAS t4 bue dpa tex HUE Example of 4 point observation scheme of a semi extended objects without extra overhead for observation of clear sky Remember than although the target is moving Example of a 6 point observation of semi extended objects alternating between two target fields and one of two different skies 22e Examples of Special Dome Flat images From left to right lamp off lamp off with mask lamp on with mask lamp On rss Stability of the Special Dome Flat images accuracy of the photometry as a function ofthe flat field age a ke eet y be he ts eto t XO LS OS OLSOELD star de Sk ve bee hob a rai ee Se eC WANTS Real Time Disp
158. perator upon arriving e The visitors are expected to read carefully the La Silla web page to acquaintance themselves with the rules of the operation of the observatory and the instrument they are going to use e The visitors should not edit by hand an OB or write a script to produce them as it may cause software crashes Detailed visitor guidelines are available at the La Sills web page http www eso org sci facilities lasilla sciops observing VA GeneralInfo html 4 2 The VLT environment P2PP BOB OS TCS DCS ICS Observations are described by observing blocks OBs essentially scripts for driving the telescope and the instruments The OBs consist of one or more templates in effect commands in the scripts The templates have a number of parameters much like the functions in computer programming 42 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 43 A typical science OB is made up of four components a target an acquisition template an observation description and a set of constraints The observation description itself is made up of one or more observation and calibration templates The constraints describe the acceptable conditions under which an observation can be carried out in service mode and therefore the constraints section can be ignored by the visitor mode observers Templates are the simplest unit of an observation They are split into three categories acquisition observing and calibration The templates are
159. plate is fileName_0025 fits If no file with the same base name is found on disk the sequential number is automatically set to 0001 see for an example Table 3 16 NOTA BENE Do not use spaces slashes or other non alphanumeric characters in file names and target names otherwise the SOFT OS will report an error However the underscore symbol _ is acceptable Signature Keyword Default Description Input Value Exposure Name DET EXP NAME Exposure Base filename ngc1068 Number of Exposures SEQ NEXPO Number of exposures for 10 the template Table 3 16 File naming signatures and keywords The first filename will be ngc1068_0001 fits if no file with the ngc1068 string was found on disk and the last file will be ngc1068_0010 fits SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 41 3 6 6 Detector Window In all templates except acquisition templates it is possible to window the detector The window is described with the coordinates of the origin and the window size Table 3 17 Note that the entire array is read the windowing parameters only determine which part of the image will be stored in the raw data file Therefore the windowing does not decrease appreciably the overheads Signature Keyword Default Value Explanation Number of rows DET WIN NX 1024 Number of rows Number of columns DET WIN NX 1024 Number of columns First column of window DET WIN STARTX 1 First column of window First row of window DET
160. ption is selected AND the guiding is started The option B stands for Box To Star and it must be used with guiding i e the offsets are combined with with appropriate counter movement of the guiding arm to maintain the pointing This is necessary for long total integrations more than 5 10 min at each individual position The option S stands for Star To Box It has been disabled i e the template fails if the guiding option is set to S but it was kept in the software for historic reasons Note With large combined offsets the guide probe may not be able to follow the one guide star In such a case the guiding system will automatically find another star and resume guiding If any of the entries in SEQ GUIDING LIST are different from N and if guiding is off at the beginning of the template a pop up window prompts the operator to start guiding A list is also provided for NDIT SEQ NDIT LIST but not for DIT The observation type can be defined for each image and is entered as a list in SEQ OBSTYPE LIST O stands for Object and S stands for Sky This is to make it essayer for the user to classify the images during the data reduction The total number of exposures is defined in SEQ NEXPO This number can differ from the number of elements in the aforementioned lists Lists do not need to have the same length If the number of exposures is larger than the number of elements in a list the list is restarted from the beginning as many tim
161. r Example SOFT User s Manual 2 3 Parameter signature Exposure Name DET DIT DET NDIT Number of columns Number of rows First column of window First row of window Number of exposures Spectro Mode Which Slit Return to Origin T F X offset list arcsec Y offset list arcsec Observation Type S or O Guiding N B S Parameter signature Exposure Name DET DIT DET NDIT NSAMP NSAMPPIX Number of columns Number of rows First column of window First row of window Number of exposures Spectro Mode Which Slit Return to Origin T F X offset list arcsec Y offset list arcsec Observation Type S or O Guiding N B S Header Keyword DET EXP NAME DET DIT DET NDIT DET WIN NX DET WIN NY DET WIN STARTX DET WIN STARTY SEQ NEXPO INS SMODE INS WHICHSLIT SEQ RETURN SEQ OFFSETX LIST SEQ OFFSETY LIST SEQ OBSTYPE LIST SEQ GUIDING LIST Header Keyword DET EXP NAME DET DIT DET NDIT DET NSAMP DET NSAMPIX DET WIN NX DET WIN NY DET WIN STARTX DET WIN STARTY SEQ NEXPO INS SMODE INS WHICHSLIT SEQ RETURN SEQ OFFSETX LIST SEQ OFFSETY LIST SEQ OBSTYPE LIST SEQ GUIDING LIST LSO MAN ESO 40100 0004 99 Value Description SOFI File name prefix NODEFAULT Detector Integration Time individual exposure sec NODEFAULT Number of DITs averaged into an individual image 1024 Number of columns in the window 1024 Number of rows in the window 1
162. r 64sec of integration in total require about 80sec so the overhead is 25 The user may choose to accept the large overheads or the presence of increased lines in the image that are hard but not completely impossible to correct The sky background is another factor that has to be accounted for when selecting a DIT It is the strongest in K band when it amounts to 400 800 ADU per second depending strongly on the humidity The sky background can easily saturate the array by itself if the user selects a big DIT of 15 30 sec or more depending on the filter Furthermore this background is not constant It varies on a time scale of a 1 3 minutes becoming a source of systematic uncertainties To account for them the user must monitor these variations on the same time scale This is done by alternatively observing the target and a clear sky field next to the target the next sections contain some useful tips how to optimize these sampling in case of different types of targets The frequency of the sky sampling is determined by the product DIT xNDIT plus the overhead The user should try to keep this time close to 1 2 min in average sky conditions or 3 min in exceptionally good conditions To verify the choice of sky sampling the observer should frequently subtract sequential images from one another and monitor how large the average residual is Ideally it should be smaller than or comparable to the expected Poisson noise but this is rarely the c
163. r the user to the corre sponding user manuals Note that in case of IRAF the task imcombine can not handle fractional offsets The workaround is to use a 2 step shifting procedure First shift the images by the fractional part only with the task imshift If you intend to do photometry on the final image set the interpolation parameter of imshift to linear in order to conserve the flux during the shifting In addition set the boundary value that will be given to pixels with no value to constant and set that constant to a large negative value i e 1e6 Later during the combination these pixels can be excluded with an appropriate lower cut off threshold If you use any of the outer options for the boundary pixels you will end up creating data The second step of the shift can be carried out with imcombine using the offset facility of this task The final image will have the largest possible dimensions and the signal to noise ratio will most likely degrade toward the edges because of the smaller and smaller number of images that cover the outer regions Remember to set the lower threshold of imcombine appropriately to exclude both the marked bad pixels and the pixels that were given bogus values at the edges of the shifted images SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 57 5 3 Long Slit Spectroscopy 5 3 1 Inter quadrant Row Cross Talk This step is a must for reducing spectroscopic observations because unlike the imaging
164. ree jittered images is the lowest meaningful number to ensure the array cosmetics removes well but having 5 7 or more images SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 39 is preferable especially if they have to be used for creating a sky frame An additional restriction on the offsetting frequency applies to imaging observations Unlike equa torial telescopes the pupil plane of the NTT rotates relative to the image plane and for reasons which may be due to internal reflections within the objectives this causes pupil ghosts images of the telescope support structure near the primary mirror to appear in sky subtracted images In other words the traces of the spider supporting the secondary mirror do not fully subtract on images taken more than 5 10 min apart The effect is greatest when the paral lactic angle changes quickest and this occurs when the telescope is near the meridian The effect is worst for images taken in K and the large field objective To minimize this effect the beam switching frequency should be such that the parallactic angle does not change by more than 0 5 degrees between exposures For jittered images the residual effect is at or below the shot noise level when applying standard reduction techniques which involve running sky measurements over 10 frames For applications requiring large object sky offsets it is advisable to use a strategy which automatically equalizes the parallactic angle of the two This s
165. rk frames both for imaging and spectroscopy The instrument is set to the DARK mode where both filter wheels and the grism wheel are in the closed position The number of frames is defined in SEQ NEXPO Dark frames can be taken with different DIT and NDIT values defined as lists in SEQ DIT LIST and SEQ NDIT LIST Note that SEQ DIT LIST and SEQ NDIT LIST do not need to have the same length If the number of exposures is larger than the number of elements in a list then the list is restarted from the beginning as many times as needed until the correct number of frames have been acquired This template should be used to take dark frames at the end of the night for all the detector integration times that have been used during the night though it is not strictly necessary since the dark signal will be removed along with the subtraction of the sky background SOFI img cal DomeFlats and SOFI img cal SpecialDomeFlats Imaging dome flats are taken with these calibration template Table C 27 For each element in SEQ DIT LIST and SEQ NDIT LIST four images are taken one with the dome lamp off two with the dome lamp on and a fourth with the dome lamp off The intensity of the dome lamp is controlled manually SOFLimg cal SpecialDomeFlats takes an additional set of images with the aperture wheel partially masking the array These images are used to estimate the bias pattern The reasons for using this template rather than the first are explained in
166. rly type stars refer to Hanson et al 1996 ApJS 107 281 Good empirical IR spectrophotometric standards are not available The only remaining way to do flux calibration is to observe the science target in one or more broad band filters and to scale the spectrum so that it agrees with the broadband fluxes This method can achieve an absolute calibration of 5 10 For an example of IR flux calibration in the more general case of extended objects see Ivanov et al 2000 ApJ 545 190 See also the discussion on the flux calibration in Section 3 4 3 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 59 5 3 7 Alignment and Combination There are two general approaches to combine spectra i to combine 2 dimensional images after an appropriate geometric correction that will align the wavelength axis and straighten the spectra ii to combine 1 dimensional spectra usually after wavelength calibration that were extracted from the 2 dimensional images Every one of these methods has certain advantages and disadvantages and the user will have to decide what is the best way to reduce the data Obviously the first method is easier for reducing ex tended objects because it will become trivial to extract spectra at different locations of these objects However the geometric correction involves heavy modification of the data and some degradation of the resolution may be possible The second method is easier to apply on point source spectra and i
167. rompted to store a fixed pattern that is subsequently subtracted from the incoming images via the real time display RTD features After storing the fixed pattern the telescope returns to the original position At this point in time the operator is offered the possibility to change DIT and NDIT This is useful if the target can not be identify securely because of the low signal to noise If the operator changes the values the telescope offsets again and the operator is required to store another fixed pattern before the telescope returns to the nominal position This loop can continue until the operator has identified the target To move the target from one position of the array to another the operator has two options 1 If the target must be centered on the slit one can use the build in feature by clicking on the center button on the RTD dialog box The software will draw an arrow from the location of the selected object to the center of the slit and the operator will be asked to confirm the offset This option is faster and it works well if the science observation is done via simple nodding along the slit 2 The operator can simply draw an arrow on the screen holding the left hand side button of the mouse At this point a window which lists the pixel co ordinates at the start and the end of the arrow will appear The operator can accept the offsets cancel or edit the co ordinates manually If the offsets are accepted the telescope offsets
168. roscopic arcs e The exposure time keyword represents the total integration time for a single image i e DITXNDIT As the image you receive is the average of NDIT exposures of DIT seconds the correct number to use for flux calibration is DIT and not the value given in the exposure time keyword 5 2 Imaging 5 2 1 Inter quadrant Row Cross Talk This is the first step of the data reduction but some specific cases described below require post processing of the final image to remove cross talk residuals A bright source imaged on the array produces a ghost that affects all the lines where the source is and all the corresponding lines in the other half of the detector For instance if the bright source is on row 300 the cross talk affects the row 300 512 812 and vice versa if the source is on row 750 the other affected row will be 750 512 238 The cross talk is seen either on high S N data or if the bright source reaches near saturation level 51 52 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Though the effect is not completely understood it is well described and can be easily corrected The intensity of the ghost is in fact 1 4 x 107 times the integrated flux of the line An IRAF script for removing the cross talk is available from the SOFI web page http www eso org sci facilities lasilla instruments sofi tools reduction sofi scripts The script has to be given input and output lists of images The script can re
169. roscopy 3 4 1 Small Objects and Uncrowded Fields In spectroscopy like imaging accurate sky cancellation is important If the object is small enough the object can be observed at different slit positions Sky cancellation is then achieved by subtracting one frame from another This is a very efficient method because it is not necessary to spend extra time in tegrating off the target This technique is embodied in the template SOFI spec obs AutoNodOnSlit Typical parameters for this template are listed in Table 3 10 In this template the telescope nods the object between two positions along the slit that are Nod Throw Along Slit arc seconds apart For convenience we will call one of these positions position A and the other position B At position A the object is observed for 9 minutes and the observer will receive 3 frames specified by NINT each the average of 3 specified by NDIT exposures of 60 seconds specified by DIT The telescope then moves to position B where it again integrates for nine minutes producing three frames each the average of three 60 second exposures This completes one cycle The number of A B pairs also called cycles is defined by the parameter Number of AB or BA cycles In this particular case the number of cycles is 4 so the total exposure time is 605 x 3x3x2x4 72 minutes If the number of cycles is greater than 1 the telescopes moves in the following pattern A B B A A B B A etc Between
170. rue list of offsets along RA list of offsets along Dec Parameter signature Value Exposure Name NGC6118 DIT 20 NDIT 3 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Number of Exposures 5 Filter wheel 1 J Filter wheel 2 open Instrument Mode LARGE_FIELD_IMAGING Combined offset F T T Return to origin T F T RA offset list arcsec 0 50 0 100 0 DEC offset list arcsec 0 50 100 0 100 Table C 14 SOFLimg obs Jitter Example 80 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Parameter signature Value Number of Exposures 10 RA offset list arcsec 0 5 10 2 5 5 DEC offset list arcsec 0 5 10 2 5 5 Table C 15 SOFLimg obs Jitter Offset Example Parameter signature Header Keyword Value Description Exposure Name DET EXP NAME SOFI File name prefix DET DIT DET DIT NODEFAULT Detector Integration Time individual exposure sec DET NDIT DET NDIT NODEFAULT Number of DITs averaged into an individual image Number of columns DET WIN NX 1024 Number of columns in the window Number of rows DET WIN NY 1024 Number of rows in the window First column of window DET WIN STARTX 1 First column of window First row of window DET WIN STARTY 1 First row of window Number of exposures SEQ NEXPO 1 Number of exposures in the sequence Filter wheel 1 INS FILT1 ID NODEFAULT Filter wheel 1 position Filter wheel 2 INS FILT2 ID NODEFAULT Filter wheel 2 position In
171. s 12e In NDR Non Destructive Read values as low as 3e7 have been reached with integrations of one minute About 0 1 of the pixels are bad An updated mask of bad pixels is available in the SOFI web page The gain of the array 5 4e ADU The well depth of the array is around 170 000 electrons 32 000 ADU Although the array non linearity is limited to less than 1 5 for a signal up to 10 000 ADU Fig 2 5 we recommend that observers keep the exposure short enough so that the background does not exceed 6 000 ADU This is due to the bias of the array which has a complicated dependence on the flux when the flux is above 6 000 ADU NOTA BENE A star of K 10 mag produces 9 000 10 000 ADU at the central pixel under seeing 0 6 0 7 arcsec and average conditions in Large Field imaging mode This is the case with many of the photometric standards and since these values are close to the non linearity limit the user may consider defocusing the telescope for standard star observations or switching to one of the imaging modes with finer pixel scale for observations that can not The DCS has to cope with backgrounds that range from a fraction of a ADU sec pix as seen in the spectroscopic modes to six hundred ADU sec pix as seen during summer with the K filter and the large field objective The high count rate from the background particularly at K limits a single integration to about 10 seconds and it can drop down even to 6 seconds depending on the
172. se source The minimum integration time for SOFI in this mode is 1 182 seconds e Non Destructive Read In Non Destructive Read or NDR the array is sampled several times after the reset We will call the number of samples NSAMP In other words the array is read NSAMPtimes during each individual DIT The flux in each pixel is computed by fitting a linear function to the voltage as a function of time The fitted slope equivalent to the photon rate is then multiplied by the integration time There are several ways the signal can be read in the NDR mode The one used in SOFI concentrates the read out at the beginning and at the end of the integration in this way the noise is minimized This mode is called Fowler sampling Unfortunately with the number of readings the glowing produced by the heat from the shift registers at the border of the array increases For NSAMP 60 the photon noise of the glowing starts to compete with the read out noise and above 60 becomes the dominant source of noise For this reason the NSAMP must not exceed 60 To further minimize the noise keeping NSAMP small the analog signal can be sampled several times for each reading this is done with the parameter NSAMPIX The read out noise is reduced approximately by the square root of the number of samples NSAMP This method is also less susceptible to 50 Hz pickup SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 13 noise The drawback of the NDR is that it takes slightl
173. strument Mode INS IMODE NODEFAULT Instrument Mode Combined offset F T SEQ COMBINED OFFSET F T guiding ON F guiding OFF Return to Origin T F SEQ RETURN T Returns the telescope to the original pointing if True RA offsets list arcsec SEQ OFFSETALPHA LIST NODEFAULT list of offsets along RA Dec offsets list arcsec SEQ OFFSETDELTA LIST NODEFAULT list of offsets along Dec Rotation offsets list SEQ OFFANGLE 0 List of rotator offsets on degr on Sky the sky Table C 16 SOFLimg obs JitterRot that the sums of the offsets along RA and along Dec are non zero so the second offset pattern will be shifted relative to the first one by 2 5 arc to the East and 2 5 arcsec to the North This example illustrates an interesting usage of this template to define a closed loop offset sequence that will bring the telescope at the original position or close to the original position i e 2 5 arcsec away to imitate jittering minimizing the effects of the array cosmetic defects Therefore these template can execute sequences of the type target1 sky target2 target1 sky target2 saving from the time spent on the sky Unlike some other templates such as the SOFLimg obs AutoJitterArray the freedom to choose the offsets means that the targets and the sky can form an irregular pattern SOFLimg obs JitterRot This observation template Table C 16 C 17 is identical with the SOFLimg obs Jitter with the additional capability to move
174. t degrees INS FILT1 ID NODEFAULT Filter wheel 1 position INS FILT2 ID NODEFAULT Filter wheel 2 position SEQ RETURN T Returns the telescope to the original pointing if True Table C 33 SOFL img cal PolarimDomeFlats 94 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 C 3 4 SOFT Spectroscopic Templates C 3 5 SOFT Spectroscopic Acquisition Templates SOFT_img acq MoveToSlit This acquisition template Table C 34 is very similar to the SOFI_img acq MoveToPixel template The selected slit is drawn on the real time display RTD superimposed on the image of the field In most cases operators will use the option to move the selected object to the center of the slit Since each slit has a different position on the detector the slit name i e the 0 6 1 or 2 arcsec slits is explicitly stated in the template via the INS WHICHSLIT parameter This acquisition has two parts First it presets the telescope to the coordinates of the Target associated with the Observation Block Next it takes and displays an image on the RID Finally it requests the operator to interact with the RTD so that the user specified object is moved to either a center of the slit or to a user specified location on the slit To make the target clearly visible the telescope can do a small offset after the preset to obtain a sky image The size and direction of the offset are template parameters TEL TARG OFFSETALPHA and TEL TARG OFFSETDELTA The operator is then p
175. t has an added advantage of giving a straightforward estimate of the observational error this is just the standard deviation obtained during the combination 5 4 Polarimetry As in the case of imaging you should start to use a bad pixel mask Some bad pixel masks can be retrieved from http www eso org sci facilities lasilla instruments sofi tools reduction bad pix html Then you have to apply the flat field to correct for the pixel to pixel sensitivity variations across the chip The flat field exposures should be performed using the same filter as the object exposures and the Wollaston prism in the optical path Since the whole instrument rotates in order to provide a certain orientation of the Wollaston prism separate exposures at different orientations are not required We recommend to take dome flats because when a sky map is used some ghosts of the object on consecutive images may result in a wrong flat field especially at the position s of the object Exposures with the lamp OFF and with the lamp ON must be combined to obtain the final flat field FF lamp ON a lamp OFF a lamp ON b lamp OFF b The final flat field is polarized i e the median value of the upper field is different from the median value of the lower field However this problem can be overcome by an independent normalization of the flat field in each field A further artificial polarization is introduced due to the deviation of the transmission ratio of
176. t et al 2000 A amp A 354 1134 Long ward of 2 3 microns the background is dominated by thermal emission from both the telescope and the sky and is principally a function of the temperature The background in K can vary by a factor of two between the winter and summer months but is more stable than the J or H band background on minute long time scale It also depends on the cleanliness of the primary mirror Imaging in broadband K and the wide field objective can result in backgrounds of 600 700 ADU sec depending strongly on the temperature and humidity The IR window between 1 and 2 5 microns contains many absorption features that are primarily due to water vapor and carbon dioxide in the atmosphere These features are time varying and they depend non linearly with airmass The atmosphere between the J and H bands and between the H and K bands is almost completely opaque The atmospheric transmittance between 1 and 2 5 microns as seen by SOFT is plotted in Appendix B As the amount of water vapor varies so will the amount of absorption The edges of the atmospheric windows are highly variable which is important for the stability of the photometry in J and K filters These difficulties have led to the development of specific observing techniques for the IR These techniques are encapsulated in the templates that are used to control SOFI and the telescope In this section we link common observational scenarios with specific templates and we giv
177. t the biases are an integral and indistinguishable part of the darks Dark frames are taken with the template SOFI_img_cal_Darks The parameters for this template are listed in Table 3 12 In this template one can enter a list of DITs and NDITs If the number of exposures is greater that the number of elements in either of these lists the list is repeated until the correct number of exposures have been completed The structure seen in the darks is quite complicated In general it is not a linear function of time The signal is made of several components shading a component which depends on the DIT and on the incident flux heat from the readout amplifiers commonly referred to as amplifier glow and classical dark current from the random generation of electron hole pairs The heat status from the readout amplifiers is a function of the number of reads only whereas the dark current is a linear 34 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Parameter signature Value Exposure Name Dark_Frames DIT LIST 10 10 20 20 10 10 NDIT LIST 663636 Number of exposures 12 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 Table 3 12 Parameters in the SOFI img cal Darks template with commonly used values function of time Other than depending on the DIT the shading pattern seen in darks frames also depends on the incident flux Thus subtracting a dark frame from a science frame with the same DIT
178. tain spectra at more than two positions across the slit then the offsets will be only in the x direction These templates cannot by nature be easily optimized After each exposure one or more DITs are skipped At the end of the templates the telescope is returned to the original position if SEQ RETURN is set to true T If not the telescope is not moved The lists can have any length however having lists of different lengths can become extremely con fusing It is good practice to use either lists of one value or lists of equal length The following examples Table C 41 and C 42 demonstrate a typical usage of the generic templates Six images will be obtained 1 the object at the acquisition position 2 the object in an offset position 15 arcsec 3 the sky the object is off the slit 4 the object at another offset position 5 the sky again and 6 the object at the original position A Fig with the object positions is also shown Each of the obtained 6 files is the average of 2 NDIT exposures of 40 sec DIT The guiding option asks to put the box to the star C 3 7 SOFT Spectroscopic Calibration Templates SOFI_img_cal_Darks The template Table C 26 for imaging darks can be used for obtaining spectroscopic darks if neces sary However notice that in practice this is never necessary because i in case of observations the dark including the bias is subtracted together with the sky and ii in case of the
179. ted The choices for the lamps are Xe Xenon Ne Neon B Both and N None The choices for the slit are 0 6 1 and 2 The choices for the spectral mode are B and R for low resolution grisms H and K for the high resolution grism Although the template enables one to do this it is best to keep things simple i e just take arcs with one slit and if another slit is required run the template again with that slit The Xenon lamp warms up quickly and produces high background level in the arcs notably when calibrating the red grism To minimize the background the user may want to split the time when the lamp is on giving the lamp some time to cool To do this just decrease the number of exposures and repeat the OBs with a few minutes delay as many times as the NEXP was reduced Nota Bene Remember to take images with the lamps off to remove the pattern due to dark bias scattered light and thermal background This makes it easier to identify the weaker emission lines In general you should avoid taking arcs during the night since the high illumination levels of the strongest emission lines in the arc images may cause persistence problems It is therefore advisable in such cases that you use the arcs in the morning after your run to do the wavelength calibration and then use the atmosphere to check the zero point of the calibration during the night Both OH emission and sharp atmospheric absorption lines can be used The P1 branch lines of OH ar
180. telescope Contact the SOFI support astronomer for further details SOFT has three types of templates observation templates OT for science observations Calibration templates CT for calibration exposures and Acquisition templates AT for target acquisition Tables C 1 C 2 C 3 and C 4 associates templates with observational scenarios Not all parameters have default values In those cases where there is no default value the value is NODEFAULT IMPORTANT For those parameters that require a string to be entered for example the target name or the exposure name spaces and special characters must NOT be used only letters numbers and underscore symbol are allowed The SOFI templates are maintained by the SOFI instrument scientist and by the Science Operations team The following sections will review the SOFI templates in order of the instrument modes imaging spectroscopy polarimetry This Appendix supersedes the SOFI Template Manual ESO document LSO MAN ESO 40100 0007 Type of Acquisition Template s to use Simple telescope preset point and shoot SOFLimg acq Preset Preset telescope and move an object onto a pixel SOFLimg acq MoveToPixel Preset telescope and center an object in a slit SOFLimg acq MoveToSlit Preset telescope and position an object for polarimetry SOFI_img_acq_Polarimetry Table C 1 Short guide for acquisition templates 70 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 71 Type o
181. th edge the transmission of the broader J filter is in practice limited by the atmosphere and therefore the photometry in the J filter is less stable than with the narrower J one However the J filter allows about 20 less photons to reach the detector than the J filter 2 3 Long Slit Spectroscopy SOFI offers low and medium resolution long slit spectroscopy The large field objective is used in this mode instead of the spectroscopic one because the former is achromatic There are three grisms available with SOFI two low resolution grisms and a medium resolution grism Of the two low resolution grisms one covers the region from 0 95 to 1 64 microns and the second 8 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 LT T T T T T T T T T T T T T T T mee S F 1 06 1 19 Fe II 1 26 J ET r 1 08 1 21 PB 1 28 4 dgpee MUN n e c ie 7 E EF r pee a j d H g 05r RB C E L 0 1 1 1 2 1 3 1 4 FE T I T T T T T T T T T T T T T TJ g C J 9 L Fe I 1 64 1 71 n r es 7 Rz Is 7 E E t i5 a S 0 5 F pov pO E L E F L 4 4 amp 0 1 1 1 paar jaa 1 1 6 L7 1 8 ET T T T T T T T T I B L Hel2 06 H 2 12 2 19 2 25 CO 2 34 7 a iL EL UNE UG a RM l 1 z 0 5 F a L E E Figure 2 3 SOFT filters The solid red lines indicate the currently available broad band J Js H and K filters The short dashed magenta lines are the available narrow band filters The lon
182. the Wollaston prism ideal 50 50 intensity upper lower beam Three alternatives for the correction of the wrong transmission ratio are proposed to see them in detail please read the technical report available at http www eso org sci facilities lasilla instruments sofi tools reduction Ageorges SofI Polarimetry pdf e Use the transmission ratios between the two beams of the Wollaston of Ck 0 968 and C 0 954 respectively for Ks and J Ratios for the other filters have not been determined as of yet e Since the sky is also affected by this effect a good measure for the intensity ratio of both beams can be derived from the median of the lower upper image This method may not be applicable if the polarized object covers a large fraction of the lower upper image because in this case the median value may be influenced by a possibly polarized object and moreover this method 60 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 only works properly if the background radiation of the sky is unpolarized so the images are not affected by scattered moonlight e Fitting a cos 20 function to the intensity of the object on the image as a function of the ori entation 0 of the Wollaston prism see Ageorges 2000 http www eso org sci facilities lasilla instruments sofi tools reduction polarimetry The next step is the sky subtraction since the intensity of the background radiation is an additive component to the intensity of the object t
183. the humidity is low and it is cold it has been possible to achieve better 28 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 than 196 photometry however on most nights this should be considered as the best limit and the typical accuracy is 3 596 Good planning of the observation and careful data reduction has allowed some users to reach with SOFI relative photometry of 0 3 To get good photometry you should choose standard stars that are as close as possible to your program objects and you should observe them before and after you observe your program objects The classical optical method of determining extinction coefficients by observing standards over a wide range of airmasses works less well in the infra red as the extinction coefficients are generally not a linear function of airmass and may vary with the conditions i e with the humidity During the night you should observe at least three different standards and you should observe at least one standard every two hours On nights where the humidity which is a rough measure of the water vapor column is varying considerably let s say by 4096 in one hour you will need to observe standards more frequently When observing standards two or more images typically five one in the center of the array and four in the centers of each quadrant of the standard star are obtained with a telescope shift s in between In this way one image can be used as the reference sky of the other s If severa
184. the spectra are always nearly aligned along the array columns so a ghost of the upper half of the spectrum always appears on the top of the lower one and vice versa unlike the imaging when the observers may be fortunate not to have bright star ghosts on top of their targets or they can even rotate the instrument to avoid it The effect is particularly dangerous when one looks for emission lines because their ghosts are emission line like and can be confusing More details on the cross talk can be found in Section 5 2 1 5 3 2 Sky Subtraction In most cases the user simply has to subtract one image from another This should remove most of the night sky emission The subtraction of the average of every preceding and succeeding image works particularly well The residuals can be removed during the extraction of a 1 dimensional spectrum The tool for sky subtracting of imaging data works for the spectroscopy as well 5 3 3 Flat Fields Spectroscopic flats are taken with the dome flat field screen Like imaging flats one subtracts an image with the dome lamp off from an image with the dome lamp on These flats also suffer from the shortcoming that the shade pattern is not perfectly removed However the slits do not cover the entire chip there is a region of approximately 50 pixels wide which is free of direct illumination and it may be possible to use this region to estimate the residual shade pattern and to correct the flat 5 3 4 Removin
185. tions each of DITseconds 2 5 1 Readout Modes Unlike optical CCDs the charges in individual pixels of infrared detector arrays are not shifted from pixel to pixel during the read out process Instead each pixel is independently read and each column is individually reset This enables one to develop several readout methods For SOFI two readout methods are available DCR Double Correlated Read and NDR Non Destructive Read DCR 12 SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 Figure 2 5 Deviation from the linearity of the SOFI detector measured fraction of the expected level for linear response versus the measured illumination level in log of ADUs for May 10 2012 incurs less overheads and is suitable for modes where the dominant source of errors is the Poisson noise of the sky emission imaging and low resolution spectroscopy NDR has a lower read out noise and is suitable for modes for which the read out noise is comparable to the other sources of error low and medium resolution spectroscopy e Double Correlated Read Here the voltage is sampled twice once at the beginning of the integration and a second time at the end of the integration This method is called Double Correlated Read or DCR for short It is commonly used in high background situations where integrations are forced to be short In principle this method is susceptible to 1 f noise however the Poisson noise from the background will be by far the most important noi
186. to obtain a single image at each offset position then to shift the entire offset pattern by a small jitter and to repeat it This can be done with the tem plate SOFL img obs AutoJitterArray 1 that has identical parameters takes the same number of images on the same locations as SOFI img obs AutoJitterArray but differs only in the order in which these images are taken The template SOFL img obs AutoJitterArray may be preferable if it is more important to make sure that at least some of the offset positions are observed with the required depth while the template SOFLimg obs AutoJitterArray 1l may be preferable if covering the entire mapped area albeit at a shallower depth has higher priority than the depth of the mosaic The parameters of both templates are listed in Table 3 3 1 As of 10 05 2012 these two template must never be used with guiding option set to S See the template descritption for details SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 21 Parameter signature Value Exposure Name SOFI Map DIT 10 NDIT 6 Number of columns 1024 Number of rows 1024 First column of window 1 First row of window 1 NJITT 6 NEXPO 9 Jitter Box Width arcsec 40 Filter wheel 1 K Filter wheel 2 open Instrument Mode LARGE FIELD IMAGING Combined Offset T F F Return to Origin T F T RA Offset List arcsec 0 450 450 0 450 450 0 450 450 DEC Offset List arcsec 000 4500 0 450 0 0 Table 3 3 Parameters of the SOFL
187. trategy can be employed in two SOFI templates SOFL img obs AutoJitterOffset and SOFI img obs JitterOffset just set the parameter rotate pupil to true It is possible to reduce the effect of the ghosts during the data reduction by running sky measurements over only 2 4 frames but this makes the data reduction somewhat more complicated Finally in the spectroscopic mode only one can accumulate a number of NINT exposures in one nodding position A or B before moving the telescope to the other side of the nod This feature allows to minimize the overheads for moving the telescope between the two nodding positions These overheads are bigger than the ones related to the small jitter offsets 3 6 2 Autoguiding The NTT tracks very well over most of the sky For IR imaging and some spectroscopic observa tions the exposures are short enough that there is no need to guide Guiding is necessary only for spectroscopy of faint targets with exposures of 10 15 min or longer Nota Bene If you choose to guide you should be aware that the guiding mode is currently fixed to bor2star as opposed to star2box and this option has to be chosen in templates that allow to specify the guiding mode 3 6 3 SOFI Observing Modes For SOFI there are several standard observing modes For example there are modes for wide field imaging and long slit spectroscopy The observing mode defines which optical elements are inserted into the beam As an example
188. troduction Tet A First and Final Word s 322 e Rt he do Tio ec NS RUN 1 2 Applicable documents 1 8 Abbreviations and Acronyms es SOFI Son of ISAAC 2A Optical Eayout xe er A EI qe a A eee RR eu 2 2 1 AMAGING 2 5 to sa ibd accio cs dien od e de Sec edt oh hide dere das ok Medos dee E yl 2 9 Long Slt Spectroscopy 5 ea soe tage ik ea ae ae BP dee a geared e 2 4 Polares toe Aa eL eat ed dos aedes tede be MS Ska i 2 5 The DCS Detector Control System en 2 5 1 Readout Modes 3 2 m ERO a Ai RAV Be eee quad 2 5 2 Features on the Detector 2 5 3 Windowed Reading 0 000 eee ee ee 2 6 Calibration Unit 2533 exa ee Re tal in Goin ok ok dI RARI len ele Ra 2 7 Instrument Performance and the Exposure Time Calculator 2 8 Instrumental Overheads les Observing in the IR Bel The RE SkY A ee 2 citer et Riad nicotine t on i tere ote Bae Gade Reto ES 3 2 Imagmg 20 tot d sere uude un Rudy A e e A 3 2 1 Selecting the best DIT and NDIT lees 3 2 2 Small Objects or Uncrowded Fields 22s 3 2 3 Large Objects or Crowded Fields cnn 23 224 Maps of barge Fields 2 weg PPR A WW eET3Gede3 3 2 5 Imaging of Moderately Large Object 22e 3 2 6 Faint Objects Around Bright Objects 0 o e 3 2 7 Objects with Fast Variability e 3 2 8 Standard Stars ais dier prie a iuni ad da a amp
189. troscopic Dome Flats SOFL spec_cal_DomeFlats or for different readout modes SOFLspec cal DomeFlatsNonDestr Spectroscopic Adapter Flats SOFLspec cal Flats Table C 4 Short guide for calibration templates 72 SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 C 2 SOFT Imaging Templates C 2 1 SOFT Imaging Acquisition Templates Acquisition Templates determine how a target field is acquired by the telescope In case of SofI some instrument parameters can be set up in the Acquisition Templates such as the instrument mode There are two acquisition template for simple imaging SOFI_img_acq_Preset This acquisition template Table C 5 does a simple telescope preset i e pointing in case of imaging observations to the coordinates of the Target associated with the Observing Block The Differential tracking rates can be set using TEL TARG ADDVELALPHA and TEL TARG ADDVELDELTA There is no operator intervention no acquisition image is saved on disk To save time the instrument and detector parameters most importantly the instrument mode and the filters can be set to values that will be used in the first Observation Template attached with this Observing Block An exception is the NDIT which should be set to 1 to save time because even after the telescope points at the target the next template will not start until the NDITxDIT integrations are completed It is the best template to use when pointing is not critical The NTT typically poi
190. u will find the form on the La Silla home page http www eso org sci facilities lasilla sciops EoM Chapter 5 Data Reduction This chapter does not aim to be the ultimate guide to reducing SOFI data but it does outline the general steps it does provides useful tips and in effect it can be used as a mini cook book It represents the experience the NTT team has gained in reducing data from SOFI This section of the manual will evolve with time and we are very keen to hear of any suggestions that people may have The first and the last principle of data reduction is to look at the data The data do contain all the answers you seek We urge the users to experiment with different techniques combination and rejection algorithms and to look at the quality of the final data product We also warn the users that pipelines as good as they might be are dangerous tools because they separate the observer from the data Therefore the products from any automatic data reduction tools must be treated with caution First and foremost one must understand the data the specific instrumental effects the stability of the weather conditions during the observations etc 5 1 Basic Concepts Here we list some basic concepts for the SOFI data e There are two read out methods available with SOFI NDR non destructive and DCR double correlated Data taken in one readout method should NOT be used to calibrate data taken in another an exception are the spect
191. umber of DITs averaged into an individual image Filter wheel 1 INS FILT1 ID NODEFAULT Filter wheel 1 position Filter wheel 2 INS FILT2 ID NODEFAULT Filter wheel 2 position Instrument Mode INS IMODE NODEFAULT Instrument Mode Add Velocity Alpha TEL TARG ADDVELALPHA 0 0 Additional tracking velocity in RA arcsec sec Add Delta Velocity TEL TARG ADDVELDELTA 0 0 Additional tracking velocity in DEC arcsec sec Rotation Offset on Sky TEL ROT OFFANGLE 0 0 Rotation offset on the sky degrees Table C 5 SOFLimg acq Preset SOFT User s Manual 2 3 LSO MAN ESO 40100 0004 73 Parameter signature Header Keyword Value Description DIT DET DIT NODEFAULT Detector Integration Time individual exposure sec NDIT DET NDIT NODEFAULT Number of DITs averaged into an individual image Filter wheel 1 INS FILT1 ID NODEFAULT Filter wheel 1 position Filter wheel 2 INS FILT2 ID NODEFAULT Filter wheel 2 position Instrument Mode INS IMODE NODEFAULT Instrument Mode Alpha Offset arcsec TEL TARG OFFSETALPHA 0 0 Alpha Offset for sky subtraction arcsec Delta Offset arcsec TEL TARG OFFSETDELTA 0 0 Delta Offset for sky subtraction arcsec Add Velocity Alpha TEL TARG ADDVELALPHA 0 0 Additional tracking velocity in RA arcsec sec Add Delta Velocity TEL TARG ADDVELDELTA 0 0 Additional tracking velocity in DEC arcsec sec Rotation Offset on Sky TEL ROT OFFANGLE 0 0 Rotation offset on the sky degrees Combined offset F T SEQ COMBINED OFFSET Fals
192. urther than the slit length allows Telescope offsets and guiding options are defined as lists in and SEQ OFFSETX LIST SEQ OFFSETY LIST and SEQ GUIDING LIST Telescope offsets are relative to the previous position they are defined along detector lines X and columns Y and they are in arcsec The slit is oriented along the N S direction i e along the X axis N to the left negative and S to the right positive and E W along the Y axis E down negative and W up positive for Rotation Offset on 98 SOFI User s Manual 2 3 Parameter signature Exposure Name DIT NDIT Number of columns Number of rows First column of window First row of window Spectro Mode Which slit Combined offset F T Jitter Box Width arcsec Return to origin T F Nod Throw arcsec NINT Number of AB or BA cycles LSO MAN ESO 40100 0004 Value NGC6118 40 2 1024 1024 1 1 LONG SLIT BLUE long slit_1 T 20 T 100 3 3 Table C 37 SOFLspec obs AutoNodOnSlit Example Parameter signature Exposure Name DIT NDIT NSAMP NSAMPPIX Number of columns Number of rows First column of window First row of window Spectro Mode Which slit Combined offset F T Jitter Box Width arcsec Return to origin T F Nod Throw arcsec NINT Number of AB or BA cycles Value NGC6118 50 2 30 4 1024 1024 1 1 LONG SLIT BLUE long slit_1 T 20 T 100 3 3 Table C 38 SOFI_spec_obs_AutoNodNonDest
193. user specifies his her requirements 4 6 The Data Flow Path The raw data passes from the detector to the IRACE controller where pre processing of the data occurs Exactly what the pre processing does depends on the readout method For example in non destructive readout mode the pre processing involves fitting a line through the individual readouts and multiplying the slope by the DIT SOFI User s Manual 2 3 LSO MAN ESO 40100 0004 49 Next the data is written to the disk of the instrument workstation wsofi as FITS files and it is displayed on the RTD Finally copies of the data are sent to the archive machine wgddhs and to the off line computer wg5off During this transfer the file names are changed from exposure name plus number to instrument name plus time stamp The archival machine is used to produce the data package the user is given at noon after the last night of the observing run The offline machine is where the user can access the data in a directory data raw Y Y Y Y MM DD The data transfer processes are transparent to the user 4 6 1 Sofl Pipeline A VLT compliant SOFI pipeline is available at the NTT It carries out on line data reduction and derives zero points automatically in nearly real time The observer can find the products on wgdoff in the data reduced Y Y Y Y MM DD sub directory The products of the on line pipeline are not science grade because they were produced with out of date calibrations They
194. we recommend that you set this option to T true For a more detailed discussion about guiding options the algorithm used to compute the offsets and the reasoning behind the option to rotate the pupil please refer to Sec 3 6 If you wish to enter the offsets manually use the template SOFI img obs JitterOffset If one wishes to use a more complex pattern that does not involve observing the object and the sky alterna tively use either SOFI img obs Jitter or SOFI img obs GenericImaging These templates are discussed fully in The SOFI TSF Parameters Reference Guide NOTA BENE The offset to a clear sky field introduces extra 130 150 overhead because we have to spend the same amount of time on the sky as on the object sky plus the usual 30 5096 overheads for readout telescope offsets etc Therefore it is recommended if possible to take advantage of the SOFI Large Field mode and to observe such targets in the same way as compact targets see Sec 3 2 2 3 2 4 Maps of Large Fields To cover a large area of the sky with a map the template SOFI img obs AutoJitterArray is available This template allows to define an array of positions NEXPO through a list of offsets in RA and DEC and to randomly jitter number of jitter positions defined by NJIT around each of the offset positions The NJIT images around each offset are completed before moving to the next offset position However in some cases it is preferable first
195. y longer to process the data although the disadvantage becomes smaller as the exposure time increases The shortest integration time in this mode is given by 1 64 x NSAMP NDR is only available for the spectroscopic modes For most applications we recommend NSAMP 30 and NSAMPIX 4 There is a noiseless component of the signal introduced by the reset which is somewhat unstable a short time after the reset Thus there is a small delay of 100 ms between the reset and the first read This component decays with time for this reason dark frames taken with long DIT have negative counts The Hawaii array like the NICMOS III array is read out simultaneously in four quadrants This leads to a characteristic jump in the count level at rows 1 and 513 This jump depends on the DIT and on the incident flux on the array The jump is stable with time and it disappears after the sky subtraction 2 5 2 Features on the Detector There is a number of features of the detector The most prominent ones are a few circular depressions in the quantum efficiency with radius if 20 40 pixels Their presence is not a source of concern because they flat field out In addition there are some variable features i e dark regions caused by dust particles that have fallen inside the instrument Rotation of the instrument can remove them Occasionally the detector array shows elevated noise pattern The effect is probably caused by in terference and despite the replacem

SofI Manual

Contents

Download Pdf Manuals

Related Search

Related Contents