Home

Manual for the Ares Frequency Response Measurement Module

1. 0 15 Time seconds The above behavior has been seen when using a single USB sound card for both the playback and recording When a second USB sound card was purchased and one was used for playback and the other for recording the problem disappeared Playback device This section is where you select what sound card is to be used for the output signal which channels left or right the signal should be played out of and the units to assume for the output signal The playback device does not have to be the same sound card as the recording device Output channels This option controls whether the output signal is played out of the left right or left and right channels simultaneously left
2. mm Fm module 1 Frequency Response Measurement Mm File Edit View NewModule Modules Tools Time data Window Help L 2 bed A QO Acquisition Setup Plotting Files Clipboard Type of measurement to perform left chan RMS V right chan RMS V stepped sine measurement Frequency parameters start 100 Hz stop 5000 Hz points 100 C log frequency spacing fixed level RMS output 0 7 data set name data set 1 data set comment Transfer function graphs v E a on amp Gaai v o a amp O C magnitude C phase C real amp imag left inputfright input C left inputfoutput C right inputfoutput Other graphs magnitude SNR CI THD N C time data Cl time envelope left input channel right input channel C output channel Plotting Options 2000 3000 C dB magnitude log frequency Frequency Hz Ready adj output for L const using L70 from data set adj output for R const using R O from data set adj output for L const using meas L O xfer fn adj output for R const using meas R70 xfer fn adj output using freg vs amplitude data file fixed level A fixed level keeps the output voltage constant for all frequencies The output level is specified in the RMS output control that will appear This control is shown below Note that the level is in RMS units and the output calibration value entered into the Acquisition soundcard calibration secti
3. C magnitude C phase C real amp imag left input right input C left input output C right input output Other graphs magnitude SNR CI THD N C time data C time envelope C left input channel C right input channel output channel Plotting Options 2000 3000 C dB magnitude log frequency Frequency Hz Ready data set name data set comment Each data set is given a name and contains comment text Once a measurement is successfully completed the data is copied to the last position in the module s data set buffer The data set will be given the name specified in the data set name field and will be given the comment specified in the comment field If you don t change the data set name they Il all be given the default value of data set 1 which isn t very informative so it s a good idea to update the name for each measurement If you want to rename or delete a data set you can do so in the Files Clipboard section Copyright 2012 2013 McIntosh Applied Engineering LLC 25 Perform measurement Pressing the Perform measurement button will start a frequency response measurement Once the measurement begins the button will change to Stop measurement which will allow you to terminate the measurement before it finishes If the SNR levels are too low the Stepped Sine Measurement can sometimes take a long time trying to find the signal buried in the noise In this case terminating the measu
4. Critical to the measurement of Hsrmpr is that the pressure Pypp is entirely a product of the HATS artificial mouth and that the pressure Pgrp is produced entirely by the DUT s speaker However in practice there is a significant second path from the MRP to ERP directly through the air as sketched below Purp N excitation signal Vam into artificial mouth Let s call the transfer function under this condition Hoff Copyright 2012 2013 McIntosh Applied Engineering LLC 43 H Porp a P MRP indicating that this path is measured with the DUT turned off If the transfer function Perpr PMmRP is measured with the DUT turned on it will be the sum of Hspyp and Horr or H on H smr H orr Thus the desired transfer function can be computed from the two measurements Hon and Hoff as H sour d o H og It is possible for Hoge to be zero One such case is when the DUT fully occludes the ear so essentially no sound exterior to the DUT makes it into the artificial ear An example of this would be a Bluetooth headset with an earpiece that seals into the ear canal somewhat like an ear plug Another possibility is when a sealed ear simulator such as the Type 1 B amp K 4185 is fully sealed probably with putty to a phone s earpiece In these cases Hoff 0 and so a single Hon measurement needs to be taken Note that if a Type 1 ear simulator is used then it should be safe to assume that a HATS isn t being used In this case ex
5. amp right ha left amp right left only right onl Output units This option specifies the units to be used for the output channel Two units are available voltage or digital A calibration value for both can be specified in the soundcard calibration selection in the Acquisition section digital level 0 1 Copyright 2012 2013 McIntosh Applied Engineering LLC 10 Play fullscale chirp out L amp R chans This button causes a full scale 1 to 1 normalized digital units chirp signal to be played out of the playback soundcard s outputs This is useful for setting amplifier gains for an external power amplifier that s being used to drive an excitation speaker Itis recommended that you turn the power amplifier gain to its minimum value press the Play fullscale chirp out L amp R chans button Slowly turn the gain on the amplifier to the highest value without excessive distortion of the speaker alternatively you can place a volt meter on the speaker and turn the gain up until the maximum drive voltage is achieved This way you can ensure that the measurement module won t overdrive the speaker during a measurement The frequency range that s used for the chirp is taken from the starting and ending frequencies specified in the Acqusition stepped sine or chirp measurement sections The chirp is played until the button is pressed again RS232 COM port for B amp K NEXUS This specifies the COM port that s being
6. 0 00088609 RMS using 0 339348 V DFS cal Factor The values have been copied to clipboard The 1kHz filter before RMS calc option can be selected to apply a 1 kHz bandpass filter to the data before the RMS level is computed As discussed above this can increase the calibration accuracy of a manually calibrated microphone The time domain data that s acquired during the RMS measurement will be plotted in the right side of the Ares window An example of this is shown below When performing a calibration the quality of the calibration signal should be observed by using the right click graph menu and Zooming in on the data However in the case shown below the inputs only had noise on them so a nice IkHz calibration tone is not present You would not want to use such a signal for calibration Copyright 2012 2013 McIntosh Applied Engineering LLC 16 AD NST module 1 Frequency Response Measurement h File Edit Yiew New Module Modules Tools Time data window Help DG id yade n ___ o o Sound Card Data Acquisition Setup Plotting Files Clipboard Type of measurement to perform soundcard calibration lt i Maximum peak to base input level leftchannel 0 34064 right channel 0 339348 Input voltage calibration measurement a Manual mic sensitivities ViPa v ViPa Measure RMS levels of record device inputs C apply 1kHz filter before RMS calc lized RMS values me
7. 44 1kHz configuration Go to the control panel select sound then recording right click on the sound card and select properties select advanced and then select 2 channel 16 bit 44100 Hz CD Quality from the drop down box The dialog boxes for doing this in Windows 7 are shown below Sound x pa Microphone Properties Playback Recording Sounds Communications General Listen Levels Advanced Select a recording device below to modify its settings Default Format r Select the sample rate and bit depth to be used when runnin in shared mode Microphone 2 USB Audio C AA Default Device channe gt bit 44100 Hz CD Quality r Microphone channel 16 bit 11025 Hz Dictation Qua Realtek High Definition Aud ODEC Quality i ion Audi channel 16 bit 16000 Hz Tape Recorder Quality a Read channel 16 bit 22050 Hz AM Radio Quality channel 16 bit 32000 Hz FM Radio Quality annel 16 bit 44100 Hz CD Quality i a ecorder Quality annel 16 bit 22050 Hz AM Radio Quality i FM Radio Quali Configure Set Default Properties estore Defaults ok Cancel ji w Ga an Also look at the recording settings and turn off any advanced features such as bass boost or dynamic gain adjustments The sound card should be just a simple input device with no special processing features enabled Many sound cards are configured for their analog inputs to be compa
8. E T E E E E 4 PSD a ee VCE E E E E A eeenees 4 TG U1 016 K o 8 EAEE EEA E VE ee EE mT ER A OI EE EE A A E T S E AE E E T T 4 P a A A eho anieenstoediat 4 Payback ad VCE sa inane sucht E 5 ETS KE ge E TEE I A ater A E E E T E E E E E ETE 5 Recordino device sorene a O eo eta eatin 5 SOU IAC AM e EAEE E A TTE A E A T OE A EEA AEE A EN 5 PE aeaeo e E E S 5 LoT ET E e PEE SENA N E ES A A EA E I E EE E ENA A NE E E TE N E T 6 SCP OONO Durie a A N a Pe Nites eae tee ct 7 Recor UNE CCVICC onsi a E cor beetle E Lead aoa 7 Playpack d VICO s eE E E E 10 RS COM Port Tor BS K NEXUS irrena a a A 11 INCOUISTUON CONTO esnan a T a a a E 11 soundcard Can Dra Oese e N 11 Stepped sme Measure iiasr aa a a a a sat 18 CHM eaS ire aeaa aaa a trae meee erate Senna meena att 29 WAY Te measurement eier a E eh dani N 30 PIONS C ONTO Dese a a E eeedna etna eleaets 32 Files C UpPDOar dc CONTOS eae EEE EE 34 TOSA aa a eer ON trea E E E E MORO E er Pa 35 RIR Calcul ator eaea Mu oniaden tienen Melienkaan iia 36 SCR CCUNO enean pen sutwstedwdetesusia NN i cdadons laden cutdadstelascumeus dads 37 GWG aC UN At OR cor Galaga eects aiaeare tensa aien arasetae aa a ene seamen ama ttenees 40 SENIK C I a opener eee te N es ete mete ener ots E me er wet Matteo cere nT 43 Channel Noe Measurement ririri a a E E 48 Copyright 2012 2013 McIntosh Applied Engineering LLC Document Revision History Version Date Comments 1 00 October 2012 Initial documentation release 1 01 April 20
9. adjustment Calibration for digital inputs For digital input signals the calibration button changes to Digital input calibration measurement and a digital drive level is required to be specified A similar calibration is performed but in this case the user is required to place the specified digital drive level on the soundcard s inputs How this is done is left to the user However in almost all circumstances the maximum peak to base input level for digital systems will be 1 i e unity The option to calibrate digital levels is only provided in the unlikely event that there is a digital level mismatch between the sound card and the digital system it s connected to The recorded signal is multiplied by the calibration value So if a full scale digital signal from the system being measured only registers as a 0 5 DFS signal on the PC sound cards the calibration constant should be set to 2 0 This way a 0 5 DFS signal ready from the input will be scaled up to a 1 0 DFS signal As an example of when being able to set the DFS calibration value is useful consider a connection with a Bluetooth headset made using a Rohde Schwarz CBT The CBT has analog outputs for the uplink signal to the headset and analog inputs for the downlink signal to the headset Thus an analog signal is being used to represent a digital Bluetooth signal The following was found for the system maximum input level for the CBT for the downl
10. and if it s shorter the WAV data will be looped until the desired time is achieved Playing a WAV file out can be useful for measuring the maximum SPL that can be generated by an audio device Copyright 2012 2013 McIntosh Applied Engineering LLC 49 For more information about Ares and acoustical measurement and modeling tools and services contact MAE at info MAELLC COM support MAELLC COM 678 234 5079 Or see us at MAELLC COM Copyright 2012 2013 McIntosh Applied Engineering LLC 50
11. essentially be DFS Pa since it was measured for an input level of 1 Pa Note that if the H1 measurement was performed with a different set of measurement frequencies the data will be interpolated for the frequency values for the current measurement So it is not necessary for the H1 measurement to have the same frequency settings as the current measurement It s possible that the output level computed using this method may exceed the output capability of the playback soundcard device for some frequencies In this case Ares will warn you that the required output level exceeds the systems capability and will ask if you want to continue or abort If you continue Ares will choose the maximum output possible for those frequencies Copyright 2012 2013 McIntosh Applied Engineering LLC 21 For the measurement shown below a target level of 0 02 was selected for the left channel input but the system couldn t achieve that level for all frequencies The arrows point to the frequencies where the 0 02 level couldn t be achieved 5 module 1 Frequency Response Measurement DER iB File Edit View New Module Modules Tools Timedata Window Help X m ENR TO Acquisition Setup Plotting Files Clipboard i Type of measurement to perform left eo RMS Pa 2 Vv chirp measurement v right chan RMS V Frequency parameters start 100 stop 5000 T chirp 5 periods 20 log frequency spacing ladj output for L
12. has a series of utilities for computing metrics from the data sets These are mostly focused on telephony metrics such as RLR Receive Loudness Rating and SLR Send Loudness Rating Time data Window Help RLR Calculator SRL Calculator TCL Calculator STMR Calculator Channel Noise Measurement The tools require that you specify the transfer function from the data sets for the analysis Since each data set holds three transfer functions specifying a transfer set requires two list boxes one to specify the data set and the second to specify the transfer function Select data set then Select the transfer function leftinput data set 1 right input Ce leftinput data set 3 Aa a Aak data set 4 output data set 5 right input output The two list boxes for this selection are shown below for the RLR Calculator Data set w Receive Send xfer tn data set 1 4g Select data set Transfer function to use For cale Left Right transfer Function hm S Select transfer function This data set transfer function method is used throughout the tools Some terminology for the tools DUT Device Under Test Uplink The audio signal sent from the DUT to the network Downlink The audio signal sent from the network to the DUT Network The system that the DUT is communicating with could be a closed system created with a PC computer and a network simulator Side Tone The uplink signal that s played back out the DUT s earpiece speaker t
13. level 0 207836 DFS 0 1 normalized 13 6456 dB DFS 0 1 normalized 7 49558 dBm0 Above text copied to clipboard TCLw Calculator Computes the TCLw weighted Terminal Coupling Loss according to ITU P 342 specification TCLw is obtained by measuring the transfer Functionbetween a downlink into a device under test DUT and the return uplink signal from the DUT The needed transfer Function is transfer Function uplink From DUT downlink to DUT The signals should both be digital and the uplink signal should be played at a level of 15 dBmd Note according to the ETSI Digital Enhanced Cordless Telecommunication DECT TBR10 requirements document 15 dBm0 corresponds to a sine wave of 21 15 dB or an RMS value of 0 0876 data set with transfer Funciton data set 1 v uplink from DUT downlink to DUT xfer Fn Left Right transfer Function v C wide band audio Calculate TCLw Tew lo Data set comment Append TCLw to data set comment To obtain the TCLw value 1 measure H Uner DyNeT 2 specify the transfer function H by using the data set with transfer function and uplink from DUT downlink to DUT xfer fn list boxes 3 check the wide band audio option 1f using wide band data 4 press the Calculate TCLw button The TCLw will appear in the edit field to the left of the calculate button The comment field for the data set that was selected with the data set with transfer function list box is shown a
14. minimum time frame to use for the short time FFTs The FFTs will be rounded up to the closest power of two samples Due to this power of two rounding up the actually FFT time used will likely be larger than that specified by T FFT frame The tables below give a short list of the FFT time and of FFT points for four common sample rates The T FFT frame time will be rounded up to these values The sample rate that s used is taken from the WAV file For example a 44 1 kHz sample rate WAV file the 200 ms FFT frame time will be rounded up to 372 ms and will result in an FFT size of 16384 points jsamplerate 8kHz 16 KHz The entire WAV file is played once Using a shorter WAV file will result in a shorter measurement time but less averaging The number of short time FFT averages will roughly be the length of the WAV file in seconds divided by the length of time of each FFT frame 22 05 KHz _ 44 1kHz Plotting Controls The Plotting tab displays a list of all of the data sets in the current module and allows for their data to be plotted To plot the data for a given set click on the set name in the Data set list and then select the plot data for selected data set check box that appears below the data set list Uncheck this selection to stop plotting the data for that set The plotting controls are similar to what was already covered in the Acquisition section The time data options are no longer available because the
15. right channel The data set EQ list box will list all of the data sets held in the module s buffer Once a data set is selected the left input output or the right input output transfer function will be used to compute the output voltage for each frequency If H is the selected transfer function so that _ Input channel output channel and Tres is the target RMS value for the desired input channel then the output channel level will be Trus RMS output gt Tams output Trus input output input Note that the units of Tams will be equal to the units of the input channel from the data set that H is taken from and not the units that the input channel are presently set to for the current measurement So in the controls above the data set 1 left output transfer function was measured with the left input channel being set to measure a microphone response Pascals It is important to understand that it is not required for the current measurement to have the same input units as the input units when H was measured Copyright 2012 2013 McIntosh Applied Engineering LLC 20 Consider the following situation A common audio measurement is to measure the response of a microphone relative to the response of a reference microphone as shown below I input signal from test microphone Via Ip input signal from reference microphone O output signal to excitation speaker To see the response of the test microphone one w
16. samples and so the recorded signal will actually contain less data than the signal that was played out This can be seen in the graph below where the recorded central section green curve is only about 70 ms long instead of 100ms If this happens the sound card configuration cannot be used Copyright 2012 2013 McIntosh Applied Engineering LLC 9 10000
17. specifying the amplitude at 500 1000 and 2000 Hz 500 5 1000 2 2000 2 Copyright 2012 2013 McIntosh Applied Engineering LLC 24 For frequencies not specified the data will be linearly interpolated and or extrapolated to obtain levels for other frequencies For the data above the output level will be a constant 0 2 for all frequencies above 1000 Hz For frequencies between 500 and 1000 Hz the output will be interpolated between 0 5 and 0 2 For frequencies below 500 Hz the output will be extrapolated from the 500 and 1000 Hz data points Note that the extrapolation operation can cause the output to easily go beyond the limits of the system Ares will detect these overdrive points and limit the output The output magnitude for the freq amp txt data for a measurement from 100 to 5000 Hz is shown below Note that for frequencies below 150 Hz the output was limited to a value of 0 7071 ao E module 1 Frequency Response Measurement Bm File Edit view NewModule Modules Tools Time data Window Help DE A IEE O 2 Acquisition Setup Plotting Files Clipboard output chan RMS V Type of measurement to perform stepped sine measurement Frequency parameters start stop points log frequency spacing adj output using freq vs amplitude data file freq amp txt ka data set name data set 1 data set comment Transfer function graphs ks E f H on SS m _ vo m eS a O
18. the data should at least cover 176 Hz to 4590 Hz For wideband telephony bandwidth the data should at least cover 88 Hz to 9000 Hz C use voltage to MRP pressure refrence transfer Function Data set w Uplink output xfer fn data set 1 Uplink output transfer Function Left Right transfer Function C wide band audio Calculate SLR SLR o Data set comment f Append SLR to data set comment The SLR calculator uses the specification given in ITU P 79 To compute SLR the transfer function H should be the uplink signal to the telephony network Unger over the microphone signal Pmr recorded at MRP Microphone Reference Point H U ver P MRP where Ungr 1s a digital signal for digital telephone networks and a voltage for analog telephone networks The drawing below shows a rough sketch of what a configuration might look like for an SLR measurement Uner uplink signal from DUT to network lt a aha PMRP 8 o voltage from mic T at MRP t wired or wireless connection to DUT V am artificial mouth acoustic DUT device under test such as source telephone or headset HATS or other artificial ear simulator Copyright 2012 2013 McIntosh Applied Engineering LLC 38 If it s possible to measure H directly then to compute the SLR 1 measure H of Unger PMRP 2 uncheck use voltage to MRP pressure reference transfer function 3 use the Data set w Uplink output xfer fn and Upli
19. time data is not stored in the data sets Time data is only available immediately after a successful measurement in the Acquisition section Copyright 2012 2013 McIntosh Applied Engineering LLC 32 SES ee module 1 Frequency Response Measurement Mm File Edit View New Module Modules Tools Time data Window Help DE A ERE OA Acquisition Setup Plotting Files Clipboard 20 Transfer function graphs left right chans V V data set 1 magnitude C phase left right chans V V data set 1 Oreal amp imag left input right input O left input output C right input output Other graphs magnitude SNR C THD N left input channel C right input channel Transfer function dB C output channel Plotting Options dB magnitude log frequency Data sets data set 1 0 6000 data set 1 Frequency Hz data set 1 left chan RMS V data set 1 left chan RMS V data set 1 plot data for selected data set Channel magnitude dB Frequency Hz By right clicking on a graph a menu will pop up with some useful features leFe fright chans wiwi data set 1 Copy trace data to clipboard as tab delimited text for amp Copy graph foorn Autoscale Copy trace data to clipboard as tab delimited text for All of the curve will appear in the sub menu listed by their legend text If one of th
20. use voltage to MRP pressure reference transfer function option 4 use the Data set w output to MRP xfer fn and Output to MRP transfer function list boxes to select H2 5 use the Data set w Uplink output xfer fn and Uplink output transfer function list boxes to select H1 6 if this is a wide band measurement check wide band audio 5 press the Calculate SLR button The result will appear in the SLR edit field TCLw Calculator The TCLw Calculator computes the weighted Terminal Coupling Loss according to the ITU P 342 specification This is essentially the uplink echo returned from a telephone device from a downlink signal Dyer downlink signal from network gt Uer uplink signal from DUT t wired or wireless HATS an A connection to DUT Head And baa DUT device under test such as Torso Simulator telephone or headset Let H be the transfer function of the uplink signal over the downlink signal or H U ver D NET When the downlink signal Dyner is played through the DUT s speaker into the HATS ear some portion of that audio will reach the DUT s microphone The DUT will attempt to use an echo canceller to eliminate the Dyer signal from the microphone signal but typically it s not able to eliminate all of it so some portion Dyer will appear on the uplink Ungr signal With modern echo cancellers their behavior can be nonlinear and time dependent so obtaining a single accurate measurement for H can be difficu
21. uses a Windows audio WAV file as the output signal from the playback sound card device It is intended to be used for measuring the transfer function of telephony systems that are designed to efficiently reproduce speech signals It is important for this WAV file to be a speech signal because the input and output signals are time aligned by using the WAV file envelope The peaks and valleys of a real speech are critical for achieving good time alignment of these signals If the WAV file doesn t have significant amplitude modulation the time alignment won t work properly and the measurement s accuracy will suffer Pink noise or a constant amplitude chirp signal do not have sufficient amplitude modulations and may not work well with this measurement technique The screen capture below shows the time domain output signal for a WAV file that contained speech As can be seen there is significant amplitude modulation that will allow for an accurate time alignment of the signals module 1 Frequency Response Measurement ile Edit View NewModule Modules Tools Timedata Window Help Acquisition Setup Plotting Files Clipboard 1 Type of measurement to perform WAY file measurement WAY file larry english 44p1khz wav WAY gain 1 Calculate WAY file metrics 5 n start 100 Hz xi stop 5000 Hz Octave bands 1 12th octaves R40 vv TFFT frame 200 ms s N amp ata set E ata set co ent 5 j li S e cd
22. 15 Separated manual into individual documents Copyright 2012 2013 McIntosh Applied Engineering LLC Frequency Response Module The Frequency Response Module uses two channel sound cards to perform simple frequency response measurements of a linear system The module has a series of tools to compute telephony metrics from the measured responses These metrics include SLR Send Loudness Rating RLR Receive Loudness Rating TCLw weighted Terminal Coupling Loss STMR Side Tone Masking Rating and channel noise The basic system configuration is shown below Recording device eft input sound card right input PC Playback device linear sound card OU Put system left or right Three transfer functions are measured leftinput right input leftinput output and right input output where input and output are relative to the PC sound cards and not to the linear system This will be confusing for users who are used to studying linear systems but it s consistent with the other Ares measurement modules and with the PC sound card I O The three magnitude spectrums are also measured left input spectrum right input spectrum output spectrum Two metrics are also computed for the sine wave based measruement methods SNR Signal to Noise Ratio and TND N Total Harmonic Distortion plus Noise Copyright 2012 2013 McIntosh Applied Engineering LLC Note The Frequency Response Measurement module does not have a s
23. 2 n unction gra agnitu a right input output ther graphs agnitude The WAV file measurement only uses power spectrums of the output and input signals The advantage of this is that it s largely unaffected by any nonlinear processing features of modern speech vocoders The disadvantage is that this measurement technique doesn t generate complex transfer functions SNR or THD N measurements WAV file measurement kd WAY file larry english 44p1khz wayv Calculate VWYAY file metrics start 100 stop 5O00 Octave bands T FFT frame 700 ms t Copyright 2012 2013 McIntosh Applied Engineering LLC 30 WAV gain The WAV gain is used to change the WAV file level that s played out of the playback soundcard device by multiplying every sample in the WAV file by this value If the gain causes the samples to clip 1 e exceed the digital output range you ll be alerted to that condition after the WAV measurement is completed To see how much gain the WAV file can be given without clipping it during playback use the Calculate WAV file metrics button and observe the head room without gain This gives the maximum gain before clipping Note that even if the WAV file output is clipped Ares will used the clipped waveform to compute
24. Ares Acoustical System Frequency Response Module User Manual Acoustical and Audio Measurement and Design Tools McIntosh Applied Engineering LLC MAE MAELLC COM info MAELLC COM support MAELLC COM Copyright 2012 2015 by McIntosh Applied Engineering LLC Eden Prairie MN USA Copyright 2012 2013 McIntosh Applied Engineering LLC Notice Ares is copyrighted and licensed by McIntosh Applied Engineering LLC MAE Ares and all of its hardware and software components are provided as is MAE makes no representations or watranties concerning the compatibility of Ares to the user s computer system or any potential damage caused to any computer peripherals digital storage systems or personal physical safety Further the accuracies of the measurement modeling and design components are not warrantied and should not be used as the sole source of evaluating an engineered system or component for commercial suitability or physical safety The user agrees not to distribute the Ares software to any non licensed third party attempt to disable the licensing system or reverse engineer the operation of the program or its hardware components Table of Contents INO MCCS tscatihtae techs eatin tee an ales etaeicataaeientiate enone team taht eo ealeemaa ahaa ieitin 2 T DEO COMON ree E S ONN 2 Pocument Revision Histone a A A E E 3 Frequency Response Module siure aa a N 4 Recordino CS VIG 6 ssec a a aa a S 4 OUI Keire PEEP IET E PE A I E E E
25. Output transfer function The H2o transfer function Pprp V am Is Selected by enabling the use unit turned off transfer function to correct for direct MRP to ear audio path specifying the name of the data set containing H2 in the Data set unit turned off list box specifying which transfer function in the associated transfer function list box Again we ll assume that the DRP artificial ear microphone was placed onto the left input channel and so the H2o transfer function will be the Left Output transfer function The H2 transfer function Pprp V am 1s selected by specifying the name of the data set containing H2 1n the Data set w unit turned on list box specifying which transfer function in the associated transfer function list box Again it s assumed that we re using narrow band data so wide band audio is unchecked The STMR dialog box controls will be set as shown below Pressing Calculate STMR will compute the STMR value use output voltage to MRP pressure refrence transfer Function Data set w output to MRP xfer Fn Prorpy am Output to MRP transfer Function Left utput transfer Function use unit turned off transfer Function to correct For direct MRP to ear audio path Data set w unit turned off Pdrp f var off transfer Function Left Output transfer Function Data set w unit turned on Pdro Yam on transfer Function LeFt Qutput transfer Function correction Type 3 3 B amp K HATS Sumulato
26. alibration w Recording dewice M Audio Transit USB w Maximum peak to base input lewel left chan digital level 0 1 T let channel 1 87 DFS F right chan pot used 1kHz peakto base digital cal sine level digital drive 1 C release sound card resource J DFS v Digital input calibration measurement Play Record pulse from soundcards Playback device Measure RMS levels of record device inputs M Audio Transit USB E apply 1kHz filter before RMS calc Output channels left amp right w Maximum peak to base output level Output units digital level 0 1 max output 1 27 DFS w m E I I E f I Manual mic sensitivities If the left or right input channel is specified to have a mic manual cal value input a section will appear between the Input voltage calibration measurement and the Measure RMS levels of record device inputs button This section shown below will require that the user manually enter the microphone sensitivity for each channel that s configured with the mic manual cal value option Input voltage calibration measurement Manual mic sensitivities L mic cal 1 ViPa al R mic cal 1 WPa v Measure RMS levels of record device inputs Copyright 2012 2013 McIntosh Applied Engineering LLC 15 The module doesn t have a direct means to calibrate a microphone but the Measure RMS levels of record device inputs button can be used to read the signal level from a microphone when a calibrator is placed on the micr
27. ase input levels This is the maximum voltage that can be sampled by the sound card inputs without clipping The Input voltage calibration measurement button can be used to perform a calibration of the input levels by recording a known voltage signal on the inputs It requires that a 1 kHz 100 mV peak to base signal be connected to the sound card inputs The MAE200 ACVC1 A generates such a signal and is available from MAE A photograph of the MAE200 ACVC1 A connected to the inputs of a USB sound card is shown below along with screen captures of the resultant calibration Note that the left channel s maximum input was 0 340624 and the right was 0 339337 While these values are very close it s important that they re not identical MARS MARS1 module 1 Frequency Response Measurement iE Eile Edit view NewModule Modules Tools Time data MER A 4 3 PS ep 2 Acquisition Setup Plotting Files Clipboard Type of measurement to perform Window Help And Card Data soundcard calibration v Maximum peak to base input level 0 1 left channel 0 340624 v right channel 0 339337 Y v Input voltage calibration measuremen 0 1 O apply 1kHz filter before RMS calc 02 Maximum peak to base output level recorded data 0 1 normalized levels max output 1 Vv v Type of measurement to perform A Left channel soundcard calibration v Right channel ri
28. asured on sound card inputs Left channel 0 00258623 RMS DFS 0 000880972 RMS using 0 34064 V DFS cal Factor Right channel 0 00261116 RMS DFS 0 00088609 RMS using 0 339348 V DFS cal factor recorded data 0 1 normalized levels Maximum peak to base output level The values have been copied to clipboard max output 1 y Peak to base output level for cal output level 0 1 DFS Play output calibration tone Cal output by recording output via loopback eady Maximum peak to base output level The output calibration is specified by the maximum peak to base signal that can be generated by the playback device s output This value is entered into the max value parameter Two buttons are provided to provide two means for calculating the output calibration value While either technique can be used if the sound card inputs are voltages and have already been calculated the second technique Cal output by recording output via loopback 1s probably the preferred technique Play output calibration tone This button will play a 1 kHz tone out of the playback device The level is specified in the Output level parameter By measuring the AC level generated by the tone with a volt meter the output calibration value can be computed Before the tone is played a dialog box is presented to the user giving details on how to compute the calibration value Cal output by recording output via loopback If the sound card inputs hav
29. citing the DUT with an artificial mouth and getting a good MRP pressure measurement becomes a problem However 1f you understand that the main function of HATS is to properly produce the path loss from MRP to the device s microphone then it should be possible to configure a standalone artificial mouth MRP microphone and DUT to simulate this path loss But the path loss from the MRP to the DUT microphone must be known As was the case with the SLR Calculator it may not be possible to measure the MRP microphone Pyrp signal with the DUT in place In this situation the transfer function from the artificial mouth voltage Vam to the MRP microphone pressure Pyrp can be used to aid in making the Hor and Hon measurements Call this the Pyrp V ay transfer function H1 H1 Purp Vam which will be measured with the DUT not in place This provides a good estimate of the Purp pressure for a given Va signal Since we assumed that the DUT interferes with the MRP microphone when it s in place on the HATS Pyrp can no longer be measured So H1 will have to serve as the Pyrp estimate given a Vam signal Then with the DUT in place the transfer functions from the artificial mouth to the artificial ear measurements are taken with the DUT off and on Call these transfer functions H2 and H2 or Re me H2 y off with DUT off lt AM Copyright 2012 2013 McIntosh Applied Engineering LLC 44 and H2 Pere with DUT on Vig Th
30. const using LFO from data set x RMS target 0 02 data set EQ 4 data set 1 data set name data set 1 data set comment Perform measurement Transfer function graphs Channel magnitude C magnitude C phase CO real amp imag left inputfright input C left inputfoutput C right inputfoutput Other graphs magnitude SNR CI THD N Cl time data C time envelope left input channel right input channel C output channel Plotting Options Cl dB magnitude log frequency 2000 3000 Frequency Hz The output voltage has been added to the magnitude plot shown below The frequencies where the desired left input level couldn t be achieved are the frequencies where the RMS output level was limited to 0 707 Vrms H module 1 Frequency Response Measurement fm Eile Edit View NewModule Modules Tools Timedata Window Help BREA 3 e Acquisition Setup Plotting Files Clipboard 0 8 ee RMS Pa right chan RMS V output chan RMS V Type of measurement to perform chirp measurement j Frequency parameters start 100 stop 5000 T chirp 5 periods 20 log frequency spacing adj output for L const using L70 from data set v RMS target 0 02 Pa i data set EQ 4 data set 1 o data set name data set 1 data set comment Perform measurement Transfer function graphs C magnitude C phase C real amp imag left inputfright input C
31. e MRP microphone was put onto the right channel Again it s assumed that we re using narrow band data so wide band audio is unchecked The STMR dialog box controls will be set as shown below Pressing Calculate STMR will compute the STMR value use unit turned off transfer Function to correct for direckt MRP to ear audio path Data set w unit turned off Pdrp Pmrp off w transfer Function Left Right transfer Function t Data set w unit turned on Fdrp Prorp on nt transfer Function Left Right transfer Function Ww correction Type 3 3 BRE HATS Surmulator DAP to ERP w wide band audio Example 3 Hoff is non zero and MRP mic is removed for DUT to be in place In this example the MRP microphone must be removed in order for the DUT to be put in place on a HATS so the H1 transfer function of Pmrp V am must be measured and then the subsequent measurements H2o and H2on will be Pprp Vam with the unit off and on respectively The H1 transfer function Pyrp V am is specified by enabling use output voltage to MRP pressure reference transfer function specify the name of the data set containing the H1 in the Data set w output to MRP xfer fn list box specifying which transfer function in the Output to MRP transfer function list box Copyright 2012 2013 McIntosh Applied Engineering LLC 47 We ll assume that the DRP artificial ear microphone was placed onto the left input channel and so the H1 transfer function will be the Left
32. e been calibrated the sound card s output can be calibrated by connecting it to the left channel sound card input When this button is pressed a dialog box will appear giving the user instructions on how to configure the system for calibration Care must be taken to ensure that the sound card output doesn t clip the sound card input If it does the user will be prompted with a clipping warning You should reduce the output level value and try the calibration again Copyright 2012 2013 McIntosh Applied Engineering LLC 17 stepped sine measurement The stepped sine measurement acquisition section provides an interface to measure a response using pure tones A series of tones from the starting frequency to the stopping frequency are generated The number of frequencies are specified along with whether they are linearly or logarithmically distributed A screen capture of the controls is shown below along with a graph of the left input right input transfer function magnitude that was last measured Ss eS module 1 Frequency Response Measurement Bm File Edit View NewModule Modules Tools Time data Window Help DE A ha i Op Acquisition Setup Plotting Files Clipboard lefricht ol NN left nght chans V V Type of measurement to perform stepped sine measurement v Frequency parameters l start 100 stop 5000 points 100 C log frequency spacing fixed level RMS output 0 7 data set name data set 1 data set com
33. e curve names is selected the data for that curve will be copied to the clipboard in text format It can then be pasted into a text editor Excel or any other text based program Copy graph This option will cause the graph to be copied to the clipboard in a metafile format It can then be pasted into applications that support graphical displays such as Word PowerPoint Excel etc Zoom After the Zoom option is selected the mouse s left button can be used to click and drag a zoom box over the graph data After the left button is released the area in the zoom box will be the new display area for the graph Copyright 2012 2013 McIntosh Applied Engineering LLC 33 Autoscale This option allows you to zoom out and autoscale all of the graph data It will return the graph to its default x and y axis scales Files Clipboard controls The Files Clipboard controls allow for individual data sets to be saved to or loaded from ARES FR files Features to edit the name and comment or delete the data set are also available When the data set is selected in the Data sets list the comments for that data set are shown to the right of the controls mm module 1 Frequency Response Measurement Im File Edit View New Module Modules Tools Time data Window Help L iF led SS BO Acquisition Setup Plotting Files Clipboard Comment for data set data set 3 3rd measurement of microphone transfer function Data sets in impedanc
34. e tube data set 1 data set 2 data set 4 comments for selected data set Copy data set to clipboard as text Cl use Matlab format else Excel The Copy data set to clipboard as text button is a convenient way to export all of a data set s data including the name and comment fields to the clipboard By default the values are exported as tab delimited values which can easily be pasted into Excel A format convenient for Matlab is also available by selecting the use Matlab format else Excel option Note that using the right click menu on the graphs to copy the data for a single curve is a good alternative to using this feature to export all of the data set s data Using the right click menu allows you to just export the data you re interested in as opposed to the entire data set When the Save or Save as commands in the main File menu are used it saves all of the modules to a single ARES file including all of the data in the modules So rather than exporting or importing data sets one at a time all of the data from the Frequency Response Module can be saved to one ARES file Saving all of the data to a single ARES file is the preferred method for saving data However saving too much data in a single file is an unwise practice because if that file gets accidentally deleted or corrupted you could lose a day s or week s worth of work Copyright 2012 2013 McIntosh Applied Engineering LLC 34 Tools menu The Tools menu
35. easurement are shown below Rather than specifying a number of frequency points the chirp measurement requires the length of time the chirp 1s to last and the number of sine wave periods that are to be averaged to generate a single frequency measurement chirp measurement w Frequency parameters start 100 Hz w stop 5000 Hz ne periods 20 C log frequency spacing The T chirp parameter specifies how long the chirp signal will be The longer the chirp signal the more frequencies that will be resolved for the measurement The periods parameter specifies how many sine wave periods are used to generate a single frequency For the parameters shown above Ares will generate 641 discrete frequencies for the chirp measurement in 1 5 the time it would take the stepped sine measurement to obtain 100 frequency points The chirp measurement has all the same features as the stepped sine measurement but it doesn t provide as good of a THD N measurement Stepped sine should be used to obtain accurate THD N measurements Otherwise a chirp measurement will likely be preferred The log frequency spacing option will cause the low frequency portion of the chirp signal to last longer giving greater low frequency resolution than what can be achieved with a linear chirp signal For most cases a log frequency spacing will likely be preferred Copyright 2012 2013 McIntosh Applied Engineering LLC 29 WAV file measurement A WAV file measurement
36. ed and so they appear first on the tab bar However before an acquisition can be performed the controls in the Setup tab must be properly set for the desired measurement Ares will save the last settings in these controls to the ARES cfg file and recall them the next time Ares runs The Acquisition controls are dynamic in that as the user selects different features the controls that are displayed will change to reflect the required settings for the various features For instance if the right input channel is disabled in the Setup section the Maximum peak to base input level for the right input channel will disappear in the sound card calibration section The benefit of dynamic controls is that the user only needs to enter the information requested on the screen and the screen won t be cluttered with unused or disabled input controls The downside of dynamic controls is that there isn t just one display layout to get used to which can lead to some confusion Please be patient Chances are you ll only use a few different setup configurations and you ll quickly get acquainted with the control interface In general to perform an acquisition you will follow the steps given below 1 Use the Setup controls to choose the record and playback sound card interfaces 2 Use the Acquisition soundcard calibration controls to specify calibration values 3 Use the Acquisition stepped sine or chirp or WAV file measurement controls to perform a measurem
37. ed with an ear simulator and corrected back to ERP The voltage or digital level is the signal sent to the telephony system to produce the downlink audio For normal telephony bandwidth the data should at least cover 176 Hz to 4590 Hz For wideband telephony bandwidth the data should at least cover 88 Hz to 9000 Hz Data set w Receive Send xfer fn data set 1 Transfer Function to use for calc Left Right transfer Function Correction Type 3 3 B amp K HATS Sumulator DRP to ERP C wide band audio Calculate RLR RLR 0 Data set comment Append RLR to data set comment The RLR calculator uses the specification given in ITU P 79 To compute RLR the transfer function H should be the pressure response from an artificial ear s Ear Reference Point Pgrp over the signal from the telephone network Dyner Usually the signal from the network will be a digital value but a voltage signal is also possible H Pirr D NET The drawing below shows a rough sketch of what a configuration might look like for an RLR measurement Copyright 2012 2013 McIntosh Applied Engineering LLC 36 PERP q 2 wired or wireless voltage from connection to DUT artificial ear TA _ a m pur device under test such as telephone or headset HATS or ear artificial ear simulator For a typical measurement Dyer will likely originate from the PC s playback soundcard and Perp will be recorded by one of the recordin
38. en Oa 2 of FL orp T Hl Finally an artificial ear typically measures the DRP drum reference point pressure and not the ERP ear reference point pressure So a correction from DRP to ERP must be applied Let this correction be C C Porp P DRP If the transfer functions H2 and Hore actually measured Pprp instead of Perp C must be applied to the measurement so the final estimate for Hsrmg becomes H2 H2 aie AE H1 H STMR CRTICAL NOTE It is vital that the acoustic path from the MRP microphone to the artificial ear not change between Hon and Hog or H2on and H2o Moving the DUT even a fraction of a millimeter to turn it on or off could alter the acoustic path and ruin the measurements It is advised that some means to turn the device on or off without actually touching it be devised Perhaps a hard wired switch that s several inches away could be used to disable the battery The full STMR Calculator dialog box is shown below Copyright 2012 2013 McIntosh Applied Engineering LLC 45 STMR Calculator STMR Side Tone Maksing Rating is computed according to the ITU P 79 specification using a transfer Function From an MRP mic to the unit generated ERP audio If the unit doesnt allow an MRP microphone to be in place an output to MRP mic transfer Function can be used to predict the MRP signal IF the unit doesnt occlude the ear a unit turned off MRP mic to ear mic transfer Function can be used to subtrac
39. ent Copyright 2012 2013 McIntosh Applied Engineering LLC 4 Use the Plotting controls to compare multiple measurements 5 Use the Files Clipboard controls to export or input different measurements Each of these sections will be discussed in detail below Setup Controls The Setup tab controls are where the input and output sound cards are selected as well as the behavior of the left and right channels for those sound cards The controls are shown below Acquisition Setup Plotting Files Clipboard Sound cards for measurement Recording device M Audio Transit USB lt select sound card for input signals leftchan mic NEXUS ch 2 v right chan voltage measuremen F a eas T behavior of left amp right channels C release sound card resource Play Record pulse from soundcards Playback device M Audio Transit USB T q elect sound card for output signals Output channels left amp right v lt Select which channel the output is played onto Output units voltage q __________ select units for output RS232 COM port for B amp K NEXUS COMI Select COM port for NEXUS control Note the NEXUS must be properly with the microphone calibration Also set baud to 2400 and echo on Recording device All of the sound cards that are available for recording will be displayed in the recording sound card list box Ares doesn t differentiate between sound cards with stereo or mono in
40. g sound card inputs If the left channel is used then the Transfer function to use for calc in the dialog box would be set to Left Output transfer function Data set w Receive Send xfer fn will be the data set that contains the desired transfer function H The RLR specification requires that the ERP Ear Reference Point pressure be used but the ear simulators are designed to provide a DRP Drum Reference Point pressure so a DRP to ERP correction is required The Correction that s chosen depends on the type of artificial ear used A list of all of the ear simulator corrections supported by Ares is shown below no correction Type 1 BRE 4185 Sealed Ear Sumulator DRP to ERP Type 2 IEC 11 Ear Sumulator DRF to ERP Type 3 2 BRE 4195 Low Leak Ear Sumulator DAF to ERP Type 3 2 B amp K 4195 High Leak Ear Sumulator DRP to ERP Type 3 3 BRE HATS Sumulator DRF to ERP Select the type of ear simulator used for the measurement and press the Calculate RLR button to compute the RLR value If the frequency range of the data wasn t wide enough for the calculation a warning dialog box will be given but an RLR value will be calculated from what frequency data was available The RLR value can be saved in the data set s comment field The data set s comment field is available for editing at the bottom of the dialog box To make it easy to add the RLR value press the Append RLR to data set comment button and the RLR value will be appended to the commen
41. ght click on graph select Zoom draw a rectangular zoom box around the data to better view the 1kHz waveform and verify that it s not clipping or distorting MARS MARS1 module 1 Frequency Response Measurement Mm Eile Edit view NewModule Modules Tools Timedata Window Help LOG md 4 3 2 Acquisition Setup Plotting Files Clipboard Sound Card Data Left channel Maximum peak to base input level left channel 0 340624 O apply 1kHz filter before RMS calc Maximum peak to base output level recorded data 0 1 normalized levels Peak to base output level for cal 0 3 F output level 0 1 DFS v Play output calibration tone Right channel max output 1 a Peak to base output level for cal output level 0 1 0 4 Cal output by recording output via loopback time s Play output calibration tone S 0 3 je Cal output by recording output via loopback 0 716 0 718 0 72 0 722 0 724 0 726 0 728 0 73 time s CAP NUM SCRL If the calibration of the left and right channels results in identical calibration values for both the left and right input channels the sound card is probably configured as a mono input When this occurs look for a setting for using the sound card in shared mode and set it to a 2 channel
42. he RMS channel noise on the Recording device sound card When no signal is present this will measure the idle channel noise on the Recording device sound card channels This will alternatively play a Way File out of the Playback device sound card during the measurement IF 4 proper Wi File is chosen the maximum SPL output of the audio device can be computed For safety testing This tool doesn t use existing Frequency response datasets but performs a new time domain RMS signal measurement Play Wit File out Playback device od length of time to record channel seconds C Apply no filtering 2 Apply A weighting Filter O Apply weighting filter Apply psophometric weighting Filter Perform channel noise measurement Left channel noise O 000155718 Vrms P wk 76 1532 BY P wt l Right channel noise 0000163165 Pa P wt 75 7475 dBFa P wt 18 2319 dBP SPL The amount of time data that s recorded and analyzed is determined by the length of time to record channel seconds field If the Play WAV file out Playback device option is selected you will be able to browse for a WAV file to play out of the Playback soundcard device The WAV file will be played out both the left and right channels with an unmodified amplitude The amount of time the playback runs is equal to the amount of time entered into the length of time to record channel seconds field If the WAV file is longer than this time it will be truncated
43. ink signal to the headset 1 50 Vpeak to base maximum output level for the CBT for the uplink signal from the headset 1 06 Vpeak to base maximum input level for the recording analog sound card 1 98 Vpeak to base maximum output level for the playback analog sound card 1 90 Vpeak to base If the recording sound card input is configured as a digital signal and connected to the output of the CBT its effective digital calibration level will be 1 98 max input voltage 1 87 1 06 max output voltage from CBT 7 Copyright 2012 2013 McIntosh Applied Engineering LLC 14 So the DFS input calibration will be 1 87 meaning a full scale input on a recording channel will equal a 1 87 DFS value If the playback sound card output is configured as a digital signal and connected to the input of the CBT its effective digital calibration level will be 1 90 max output voltage 197 1 50 max input voltage So the DFS output calibration will be 1 27 meaning a full scale output on the playback channel will equal a 1 27 DFS value on the CBT The screen captures below show how the analog sound cards could be configured as digital devices and the above digital calibration values manually entered into the soundcard calibration section of the Acquisition controls Wager et tee Tt x Acquisition Setup Plotting Files Clip Acquisition Setup Plotting Files Clipboard Type of measurement to perform sound cards for measurement soundcard c
44. left inputfoutput C right inputfoutput Channel magnitude Other graphs magnitude SNR CI THD N Cl time data C time envelope left input channel right input channel output channel Plotting Options C dB magnitude log frequency 2000 3000 Frequency Hz Copyright 2012 2013 McIntosh Applied Engineering LLC 2 As will be seen with the adj output for L const using meas L O xfer fn and adj output for R const using meas R O xfer fn options the transfer function H1 can be improved through iteration to achieve a L or R constant value with less deviation from the target value Usually one measurement will achieve a tolerance of about 1 dB about the target value If a tighter tolerance is required try using a adj output for L const using meas L O xfer fn or adj output for R const using meas R O xfer fn with several iterations for the H1 measurement adj output for L const using meas L O xfer fn and adj output for R const using meas R O xfer fn options These options should be read as adjust the output to achieve a constant left channel value using a measured left input output transfer function or adjust the output to achieve a constant right channel value using a measured right input output transfer function These options are similar to the adj output for L const using L O from data set options described earlier however the transfer function that s used to compute the desired output is
45. lt to achieve with simple test signals such as pure tones and chirps It may be that only WAV file measurement using a real speech signal will Copyright 2012 2013 McIntosh Applied Engineering LLC 40 produce a good estimate for H It is recommended that you try both the WAV and chirp measurement methods to see which method achieves the most consistent estimate for H Some echo cancellers may use a voice gating technique that will simply turn off its uplink signal Unet or replace it with comfort noise This type of behavior can typically be seen on the recorded time domain data in which case measuring a consistent value for H may not be possible Since the WAV measurement technique only uses power spectrums and not correlated noise comfort noise will be included in the uplink signal s power spectrum level and may lead to misleading uplink levels Also when using WAV type measurements it s important for the ambient acoustic environment to be quiet as any ambient uplink noise will also contribute to the measured uplink power and be considered as an echo signal Stepped sine and chirp measurements will largely be immune to such noise The units of Unet and Dret Should both be digital DFS or analog voltage for old telephone systems so the units of H is essentially dimensionless The ITU P 342 specification recommends that TCLw be measured with a downlink signal level of 15 dBm0 According to the ETSI Digital Enhanced Cordless Telec
46. measured when the Perform measurement button is pressed So there will be at least two measurements made The first measurement is made to measure the input output transfer function H1 H1 is then used to compute the output necessary to achieve the desired RMS target level Then this adjusted output is used to perform a second measurement which should achieve the desired RMS target level In this case the RMS target units will be the units of the associated input channel adj output for L const using meas L70 xferfn RMS target 0 02 of iterations 7 J lt initial output 0 7 There is an option to perform more than one measurement to estimate H1 This is controlled by the of iterations parameter One iteration means that only one measurement is used to estimate H before the final measurement is made Increasing this to 2 means that two estimates for H1 are made each one improving the estimate for the output that will achieve a constant value for the input spectrum So the total number of measurements made will be the of iterations plus one The screen capture below shows the left input channel magnitude for measurements with three different values for the of iterations 1 2 and 3 As can be seen the target value 0 002 for the left channel improves as the of iterations is increased Copyright 2012 2013 McIntosh Applied Engineering LLC 23 T ee module 1 Frequency Response Measurement Im File Edit View Ne
47. measurement to perform module 1 Frequency Response Measure soundcard calibration v Maximum peak to base input level left channel 1 y v right channel 1 y v Input voltage calibration measurement Measure RMS levels of record device inputs C apply 1kHz filter before RMS calc Maximum peak to base output level max output 1 ly v Peak to base output level for cal output level 1 DFS v Play output calibration tone Cal output by recording output via loopback Calibration for voltage inputs BME MO module 1 Frequency Response Measu im File Edit view New Module Modules 1 Ge led 3 ie O Acquisition Setup Plotting Files Clipboard Type of measurement to perform Tools Time data soundcard calibration v Maximum peak to base input level left channel 1 IDES v right channel 1 DFS v 1kHz peak to base digital cal sine level digital drive 1 DFS v Digital input calibration measurement Measure RMS levels of record device inputs C apply 1kHz filter before RMS calc Maximum peak to base output level max output 1 DFS a Peak to base output level for cal output level 1 DFS v Play output calibration tone Cal output by recording output via loopback When the inputs are voltages or microphones calibration is provided specifying the maximum peak to b
48. ment Transfer function graphs magnitude C phase C real amp imag left inputfright input C left inputfoutput C right input output m Z Q m 5 mn m amp aa Other graphs C magnitude SNR CI THD N C time data Cl time envelope C left input channel C right input channel C output channel Plotting Options 2000 3000 C dB magnitude log frequency Frequency Hz Ready Once a measurement is successfully completed the measured results will be copied into a data set buffer with the given data set name and comment By default the name will be data set 1 and the comment will be blank The data sets can be plotted in the Plotting section or saved to individual files in the Files Clipboard section The data for the last successful measurement can be plotted in the Acquisition section using the plotting options in the lower left corner of the screen The screen capture below shows the magnitude of the left and right input channels plotted for the last measurement Notice that their levels vary substantially For this measurement the output level was held constant at an RMS level of 0 7 V for all frequencies It is sometimes desirable to vary the output voltage to obtain a uniform input voltage level This can be done using the drop down list box that allows you to choose how the output should be controlled Copyright 2012 2013 McIntosh Applied Engineering LLC 18
49. nk output transfer function list boxes to select H 4 if this is a wide band measurement check wide band audio 5 press the Calculate SLR button The result will appear in the SLR edit field Frequently H cannot be measured directly because either the MRP microphone cannot be in place while the DUT is in place 1 e for a large phone that covers the HATS mouth area or because the uplink signal is digital and it s not possible to place the analog Pyrp signal and the digital Unger signal on the inputs of a sound card simultaneously sound cards are either fully analog or digital not mixed In these cases an intermediate transfer function is measured from the voltage input to the artificial mouth V am speaker to the MRP microphone Pyrp Call this transfer function H1 Then the transfer function from the artificial mouth voltage Vy to the uplink network signal DFSner or Vyner is measured with the DUT in place on the HATS Call this transfer function H2 U ver Vam Purp LV ay U ver P MRP Copyright 2012 2013 McIntosh Applied Engineering LLC 39 To compute SLR using this approach follow these steps 1 measure H1 by placing the MRP mic signal on one of the recording soundcard inputs H1 is then the MRP mic over the output voltage transfer function 2 measure H2 by selecting a sound card with the Unger network signal on one of its input channels H2 is this uplink signal over the output voltage transfer function 3 check
50. o provide the user some feedback as to the quality of the uplink audio Near or Far End A telephone conversation consists of a near end talker and a far end talker The near end consists of the phone and user that you re involved with B amp K Briel amp Kj r world leading acoustic testing equipment manufacturer HATS Head And Torso Simulator such as a B amp K Type 4128 Copyright 2012 2013 McIntosh Applied Engineering LLC 35 ERP Ear Reference Point a theoretical location at the outer surface of an ear s pinna DRP Drum Reference Point the ear drum location in an ear RLR Calculator The RLR calculator allows for the Receiver Loudness Rating to be computed from a transfer function measurement It consists of a dialog box where you specify the transfer function that s to be used for the calculation the type of correction to be applied for various artificial ears typically used in such measurements and whether or not the calculation should be done for a narrow band or wide band telephony network As of this writing almost all telephony networks are narrow band Internet communications services such as Skype are wideband RLR Calculator Compute RLR Receive Loudness Rating based on Receive Send sensitivity transfer Function as specfied in the ITU P 79 specification This transfer Function will be either Paj For an analog system or Pa DFS for a digital system The pressure is that produced by the downlink audio meaur
51. observe the input signal levels to ensure good SNR levels and that no clipping is occurring The module will place the text CLIPPING in the graph legend if clipping occurred on the inputs for a specific measurement Copyright 2012 2013 McIntosh Applied Engineering LLC The module will not produce a clipped sine wave output It will automatically reduce the output to prevent a clipped sine wave out of the sound card However Ares will allow the output of a WAV file to be clipped but will warn the user if this has occurred during a measurement Interface The interface for this module consists of a series of tabbed controls on the left side of the Ares window and some module specific menu items wes ee module 1 Frequency Response Meas em Mm File Edit View NewModule Modules Tools Time data JIGiad Acquisition Setup Plotting Files Clipboard tabbed controls odule specific menu items Type of measurement to perform soundcard calibration v Maximum peak to base input level left channel 1 v L right channel 1 y v Input voltage calibration measurement i Measure RMS levels of record device inputs C apply 1kHz filter before RMS calc Maximum peak to base output level max output 1 v Peak to base output level for cal output level 1 DFS Play output calibration tone Cal output by recording output via loopback The Acquisition controls will most likely be the most frequently us
52. ommunication DECT TBR10 requirements document which references the ITU G 711 specification 3 14 dBm0 corresponds to a full scale sine wave Ares uses a digital reference where a full scale digital sine wave is 3 01 dB DFS So 3 01 dB DFS and 3 14 dBm0 are equivalent levels To convert dBm0 units to dB DFS units subtract 6 15 dB from the dBm0 value dB DFS dBm0 6 15 So the 15 dBm0 recommended value is equivalent to 21 15 dB DFS For stepped sine or chirp measurements it s straightforward to specify the RMS output level as 15 dBm0O using the units list box fixed level ka RMS output 15 data set name data set 1 For WAV file measurements use the Calculate WAV file metrics button to determine what the dBm0 level for the WAV file is As shown below the larry english 44p1khz wav file has an RMS level of 7 5 dBm0 To bring this down to the suggested 15 dBm0 level the WAV gain should be set to 15 7 5 7 5 dB or a linear value of 0 42 Copyright 2012 2013 McIntosh Applied Engineering LLC 4 J Jason cpp cpp32_2010 trunk MARS GUI MARS MARS source_material FR manual materialilarry english 44p1khz way sample rate 44 1 kHz Only the left channel data is used for a WAY file measurement Left channel metrics w o gain headroom 0 dB RMS level 0 207836 DFS 0 1 normalized 13 6456 dB DFS 0 1 normalized 7 49558 dBm0O Left channel metrics with gain headroom 0 000265076 dB RMS
53. on Fdrp Prop ar transfer Function Left Right transfer Function Ww correction Type 3 3 BRE HATS Surnulator DRF to ERP wf wide band audio Note that use output voltage to MRP pressure reference transfer function is unchecked use unit turned off transfer function to correct for direct MRP to ear audio path is unchecked Hon 1S Specified using the Data set with unit turned on and transfer function list boxes the HATS type 3 3 ear is selected in the correction list box Finally it s assumed that we re using narrow band data so wide band audio is unchecked Copyright 2012 2013 McIntosh Applied Engineering LLC 46 Pressing Calculate STMR will compute the STMR value Example 2 H ft is non zero and MRP microphone is in place Example 2 will be assumed to be similar to Example 1 except that this time the Hoff path is non zero In this case the use unit turned off transfer function to correct for direct MRP to ear audio path option is checked and the H2o transfer function is specified in the Data set w unit turned off and its associated transfer function list box Assume that the H2o data resides in a data set with the name Pdrp Pmrp off and the H2 data resides in a data set with the name Pdrp Pmrp on Since the MRP and DRP microphone were both analog voltages they could both be recorded simultaneously with the left and right inputs The DRP microphone was put onto the left channel and th
54. on will be used to convert the desired voltage level to a sound card output level that will achieve the desired voltage fixed level wt RAMS output 0 7 If Ares determines that the RMS output level is too large to achieve a dialog box will alert the user that the output is too large and offer to reduce the level to an achievable value If the output is specified to be a digital value the controls will ask for a digital RMS value rather than an RMS voltage Note that a full scale sine wave will have a value of 0 7071 RMS DFS or 3 dB DFS Copyright 2012 2013 McIntosh Applied Engineering LLC 19 adj output for L const using L O from data set and adj output for R const using R O from data set options These options should be read as adjust the output to achieve a constant left channel value using a left input output transfer function from a data set or adjust the output to achieve a constant right channel value using a right input output transfer function from a data set When one of these options is selected you will be asked to specify a left output or right output transfer function from an existing data set to use in computing an output voltage as function of frequency that will achieve a desired left or right input value The controls that appear for this option are shown below adj output for L const using L70 from dataset RMS target 0 7 Pa s The RMS target is the desired value to be achieved on the left or
55. ophone To calibrate a microphone insert it into a calibration device that generats a known SPL level and record the voltage from the device using the Measure RMS levels of record device inputs button Divide the reported voltage by the known pressure in Pascals to come up with the volts Pa calibration value If the calibration source is a 1 kHz signal the kHz filter before RMS calc option can be selected to apply a 1 kHz band pass filter to the recorded data improving the accuracy of the calibration A B amp K Type 4231 is a typical calibrator that generats a 1kHz tone at 94 dB SPL or 1 Pa If the Type 4231 is used to calibrate a microphone and a value of 50 mV is measured the microphone s sensitivity will be 50 mV 1 Pa 50 mV Pa Note that before this technique can be used the sound card inputs must be calibrated with accurate Maximum peak to base input levels specified Measure RMS levels of record device inputs and kHz filter before RMS calc This button will cause the RMS level of the sound card inputs to be measured and reported to a dialog box The reported values include the DFS units read directly from the normalized sound card sample values along with the voltages obtained by applying the channel s voltage calibration values to the DFS values Normalized RMS values measured on sound card inputs Left channel 0 00258623 RMS DFS 0 000880972 RMS using 0 34064 DFS cal Factor Right channel 0 00261116 RMS DFS
56. or WAV file measurements the magnitude plot shows the RMS level in the octave frequency band Usually a magnitude spectrum of a wide band signal such as a WAV file will be plotted as a power spectrum in units of V NHz but Ares plots the magnitude as a linear Volts or Pascals or DFS depending on the appropriate units The total RMS level of the WAV file will be the square root of the sum of the individual voltage values squared that is if VG is the voltage of the WAV magnitude curve at frequency 1 then RMSof WAVfilesignal Y V0 i allfreq where 1 is an index that ranges over all frequency values SNR The SNR signal to noise ratio is displayed in units of dB and is computed according to the formula SNR 20 Leedamena power total power SNR is only available for stepped sine and chirp measurements It is not computed for WAV file measurements THD N THD N total harmonic distortion plus noise uses the formula THD N q total power fundamental power 100 total power Only the THD N from the stepped sine measurement should be used The chirp measurement attempts to return a THD N estimate but it presently isn t a good estimate THD N is not computed for WAV file measurements Copyright 2012 2013 McIntosh Applied Engineering LLC 27 time data and time envelope The time data graphs show the normalized 0 1 digital waveforms sent to the playback sound card device for the output channel or read from the rec
57. ording sound card device for the left and right input channels These represent the exact time domain signals that Ares sent or recorded and so it is important to verify that they generally look like good clean signals that don t have excessive noise or are clipping the sound card inputs The time envelope graphs plot a maximum absolute value of the time domain signal taken over small windowed time sections Since the time data graphs plot EVERY data point in the time domain signals Ares can become very slow when it draws them The envelope graphs will have a fraction of the data points as the full time data graphs and so the envelope graphs plot much quicker These graphs are shown for the left input channel for a stepped sine measurement Note that each of the stepped sine frequencies is clearly visible but there is a small amount of noise on the signals The input level is well scaled reaching a peak value of 0 2 of the 1 0 maximum digital value GL module 1 Frequency Response Measurement Im File Edit view NewModule Modules Tools Time data Window Help DE A ha Oe Acquisition Setup Plotting Files Clipboard left chan Type of measurement to perform stepped sine measurement x Frequency parameters start stop points C log frequency spacing adj output using freq vs amplitude data file Time data normalized freq amp txt data set name data set 1 data set comment 10 0 b Perf
58. orm measurement Time s Transfer function graphs C magnitude C phase C real amp imag left inputfright input C left input output C right input output Other graphs C magnitude SNR CI THD N time data time envelope left input channel C right input channel C output channel Time data envelope max value Plotting Options Cl dB magnitude log frequency 10 Time s The time data played out and recorded with the sound card inputs can be exported to a WAV file using the Time data menu as shown below This allows the waveforms to be examined more closely with a WAV dedicated file editor Response Measurement Tools Window Help Save recorded time data From last acqusition to Way File Save played time data From last acqgsition to Way File Copyright 2012 2013 McIntosh Applied Engineering LLC 28 Plotting Options There are two plotting options dB magnitude and log frequency These will only be applied to the appropriate graphs and the legends will indicate that the value is in dB Chirp Measurement The chirp measurement plays a swept sine signal out of the playback device This measurement is much quicker than a stepped sine measurement and will likely be the preferred measurement for systems that don t require a WAV file measurement WAV file measurements will likely use real speech signals that are required for digital telephony systems The parameters for the chirp m
59. ould typically connect the test microphone signal Ir to the left input channel and the reference microphone signal Ip to the right input channel and then plot the left input right input transfer function However if the signal from the test microphone is digital and the reference microphone is an analog voltage then the two microphone signals can t be measured simultaneously because sound cards do not support mixing analog and digital inputs In this case one can measure the transfer function H1 of the reference microphone over the output signal H1 Ta 2a O V which will have units of Pa V Then select one of the adj output for L const using L O from data set or adj output for R const using R O from data set options so the H1 transfer function can be selected to adjust the output voltage and set the RMS Target value to be 1 Pa This will drive the speaker at a level that will achieve a 1 Pa pressure at the reference microphone even without the reference microphone being measured We will use this to measure the test microphone so that its magnitude response will be relative to this 1 Pa pressure reference level When the test microphone signal Ir is measured and the magnitude of the input channel is displayed the curve will be relative to a 1 Pa reference microphone level which is essentially what a I7 Ip transfer function would look like Note that the Ir magnitude plot will have units of DFS and not DFS Pa however the curve will
60. pectrum analyzer feature It measures the response of a system to a known input Three types of measurement methods are provided Stepped Sine Chirp WAV File Stepped Sine is the slowest but provides the most accurate THD N measurement Chirp is the quickest and will be preferred for most measurements WAV File doesn t provide phase measurements but is useful for measuring the response of telephone systems using real speech Each is discussed in more detail in their own sections The module supports three types of units voltage Pascals digital or digital full scale DFS To simplify making acoustic measurements with microphones the module supports using an RS232 interface toa B amp K NEXUS amplifier Ares will read the amplifier s Pa V sensitivity through the RS232 connection and automatically apply it to the input signals to obtain calibrated Pascal measurements Note that this requires that the microphone sensitivity be programmed into the NEXUS The block diagram below shows an example of using a NEXUS to amplify the left input channel but it can also be used for the right input channel as well B amp K NEXUS Recording device amplifier sound card etapi right input Playback device linear sound card OVtpet system left or right Unlike the Flow Impedance Measurement Module the Frequency Response Measurement Module will not automatically adjust the NEXUS gains to optimize the input signal level The user must
61. perly calibrated for the microphone being used mic manual cal value As with the mic NEXUS ch 1 4 option the input will be treated as a voltage input for calibration but the V Pa conversion value for this channel is manually entered by the user and a NEXUS amplifier is not required The calibration value is entered in soundcard calibration selection in the Acquisition section digital level 0 1 The input will be treated as a digital signal whose levels are normalized to have a maximum value of 1 and a minimum value of 1 The 0 1 part of the descriptor is meant to indicate this normalization Many communications networks deal with digital levels rather than voltages For a digital signal the units will be specified as DFS for Digital Full Scale In DFS units a full scale sine wave will have an RMS level of 0 707 DFSrms or 3dB DFS If one input is configured as digital both inputs must be configured as digital A mixture of voltage and digital inputs for the left and right input channels isn t supported A digital input level calibration will be presented in the soundcard calibration selection in the Acquisition section If there are differences in digital levels between the PC and the system being measured this digital calibration value can be used to compensate for it release sound card resource It has been a challenge to get the Ares acquisition software to work across all versions of Windows and sound card drive
62. puts A mono recording sound card can be used for single input channel acquisition but a stereo input device is needed for a left right transfer function It s the author s experience that most laptops will have a built in sound card that only has a mono input channel The input options for each left amp right channel are shown in the list box below mic NEXUS ch 2 v not used voltage measurement mic NEXUS ch 1 mic NEXUS ch 2 mic NEXUS ch 3 mic NEXUS ch 4 mic manual cal value digital level 0 1 not used No data will be recorded from this channel voltage measurement Treats the channel as a simple AC voltage input This will require that the input channel be calibrated The calibration is given in terms of the maximum peak to base sine wave input that can be recorded on the Copyright 2012 2013 McIntosh Applied Engineering LLC channel before the input is clipped See the soundcard calibration selection in the Acquisition section for more details mic NEXUS ch 1 4 The input for this channel is to be from a NEXUS amplifier that s connected to a RS232 port These inputs are still treated as voltage inputs for calibration purposes but the output sensitivity in V Pa will be read from the specified NEXUS channel to provide the voltage to Pascal conversion for the channel That conversion will be applied to the signal and the data will be reported with units of Pa To use this option the NEXUS must be connected and pro
63. r DRF to ERP wide band audio Channel Noise Measurement The Channel Noise Measurement tool will perform a measurement of the weighted RMS power on the input channels of the recording soundcard device This is different from the other tools in that it will record new data and not process previously acquired transfer function data Four weighting options are provided none just compute a straight RMS value from the sound card inputs A standard audio SPL A weighting typically used for sounds under 60 dB SPL C standard audio SPL C weighting typically used for sounds over 80 dB SPL P psophometric weighting typically used for telephony noise spectra Copyright 2012 2013 McIntosh Applied Engineering LLC 48 When the Perform channel noise measurement button 1s pressed the measurement will be performed the time domain signal filtered through the appropriate weighting filter an RMS level will be computed and the results reported in the Left channel noise and Right channel noise display area The Channel Noise Measurement dialog box interface is shown below The results are presented in the correct units that the input channels have been assigned in the Setup control section In the example below the left channel was setup to be a voltage input while the right channel was setup to be a microphone input So the left channel units are voltage and dBV while the right channel units are Pascals and SPL Channel Noise Measurement Measures t
64. rement and checking the input levels may be the best course of action Graphing options The graphing options appear in the lower left corner of the Ares window Transfer function graphs magnitude C phase real amp imag left inputfright input C left input output C right inputfoutput Other graphs magnitude SNAR THD N time data time envelope C left input channel C right input channel C output channel Plotting Options _ dB magnitude J log frequency The options are divided into three sections Transfer function graphs Other graphs and Plotting options Transfer function graphs The Transfer function graphs options allow the user to specify how and which transfer functions are to be plotted The options are magnitude phase and real amp imaginary parts Once the type of graph is selected which of the three transfer functions are to appear in the graphs can then be selected The tree transfer functions are left input right input left input output right input output Copyright 2012 2013 McIntosh Applied Engineering LLC 26 Other Graphs The Other Graphs options allow the user to select what quantity is to be plotted for the three signals left input right input and output These options are magnitude SNR THD N time data and a time envelope magnitude For stepped sine and chirp measurements the magnitude graph shows the RMS level of the sine wave at the specific frequency F
65. rs On some systems not releasing the sound card resources after they have been used has made some sound card drivers usable An option has been provided to allow the user to select whether the sound card resources are released or not The user will have to experiment with this setting and their sound cards to see 1f it improves performance with a given sound card device If the problem persists contact McIntosh Applied Engineering MAE for assistance Play Record pulse from soundcards This button can be used to test the sound card configuration for proper operation It plays a short tone burst out of the sound card both the left and right channels while simultaneously recording the input signals The signal consists of three sections of a 1kHz tone Copyright 2012 2013 McIntosh Applied Engineering LLC 100ms 1kHz tone at 1 40 of the maximum output level 100ms 1kHz tone at 1 4 of the maximum output level 100ms 1kHz tone at 1 40 of the maximum output level The output of this sequence must be measurable by the sound card input channels This can be done by connecting the output to a speaker and measuring the speaker pulse with microphones connected to the left and right inputs or by simply connecting the left and right inputs directly to the sound card output If the sound cards are working properly the result should be something like that seen below There will be two graphs one for the left channel and one for the right channel
66. sound card level 0 1 RIGHT Time seconds The blue curve is the output signal and the green curve is the input signal There are two very important aspects of the recorded signal that must be observed to verify that the sound cards are working properly First the input signal green curve must have some portion of its loudest central 100ms segment overlap the output signal s blue curve loudest central part In the graph above the input green signal has the last 50 ms overlap with the beginning of the output blue curve from time index 0 1 to 0 15 Ares tried to start output playback and input signal s recording at the Same time but as is seen in the above graph the sound card didn t start capturing the input signal until 50 ms after the output signal began playing out If there is too much delay between the start of the playback and the recording the data acquisition routines won t properly synchronize the signals and there will be a problem By ensuring that the central section of the input green curve overlaps with the central section of the output blue curve there will be less than a 100ms delay between playback and recording which the data acquisition software will properly handle The second important property is that the length of the central section of the recorded signal should be the same length as the central section of the playback signal With some USB sound cards it has been observed that they will lose
67. t field text SLR Calculator The SLR Calculator allows for the Send Loudness Rating to be computed from transfer function measurements It consists of a dialog box where you specify the transfer function that s to be used for the calculation It requires that you specify whether or not the measurement uses an intermediate MRP mic over artificial mouth voltage transfer function and whether or not the calculation should be done for a narrow band or wide band telephony network As of this Copyright 2012 2013 McIntosh Applied Engineering LLC 37 writing almost all telephony networks are narrow band Internet communications services such as Skype are wideband SLR Calculator Computes SLR Send Loudness Rating based on Receive Send sensitivity transfer Function as specified in the ITU P 79 specification This transfer Function will be either Pa For an analog system or DFS Pa for a digital system The voltage or digital level is that produced by the uplink audio meaured from the device The pressure is that produced at the mouth reference position MRP measured from an artificial mouth For most phones the MRP pressure cannot be measured with the phone in place Instead an output voltage to MPR mic transfer Function is measured First and then applied to a voltage to uplink audio transfer Function to finally achieve a MPR pressure to uplink audio transfer Function which is then used to comptue SLR For normal telephony bandwidth
68. t the bottom of the dialog box It can be used to record the TCLw value Pressing the Append TCLw to data set comment button makes it easy to add the value to the comment field Copyright 2012 2013 McIntosh Applied Engineering LLC 42 STMR Calculator The STMR Calculator computes the Side Tone Masking Rating from measured transfer functions Many phones play the user s voice back into the phone s earpiece speaker so the user can essentially hear his or her own voice A side tone is this feedback audio STMR is a measure of the loudness of this side tone audio The side tone audio gives the talker on the near end of the conversation an idea of what the listener on the far end is hearing For land line phones it has been used as a means of informing the near end talker that the line is still up With cell phones and Bluetooth headsets the side tone signal is a good indication of how much wind noise is present on the uplink microphone signal and will encourage a user to find a less windy location to carry on the conversation STMR is computed from the transfer function of the pressure Pyyrp at an MRP microphone to the pressure Pgrp at the ERP ear reference point of an artificial ear while the artificial mouth 1s driven as sketched below Pmrp NN gt lt puT Device Under Test excitation signal VAM into artificial mouth Let s call the transfer function HsTMR H Porp STMR P MRP
69. t the direct audio path from the unit turned on MRP to ear mic transfer Function thus correcting for the direct audio path For normal telephony bandwidth the data should at least cover 176 Hz to 4590 Hz For wideband telephony bandwidth the data should at least cover 88 Hz to 9000 Hz C luse output voltage to MRP pressure refrence transfer Function C use unit turned off transfer Function to correct for direct MRP to ear audio path Data set w unit turned on Pdrp Pmrp transfer function Left Right transfer function correction Type 3 3 B amp K HATS Sumulator DRP to ERP C wide band audio Calculate STMR STMR o l Data set comment Append STMR to data set comment Example 1 Hoff 0 and MRP microphone is in place Assume we re using a HATS and the DUT fully occludes the ear so Hog O0 Also assume we can measure the Pyrp pressure with the DUT in place Then the H1 transfer function isn t needed In this case we only need to specify the Hon transfer function Pprp Pmrep but we ll need to specify the correction factor C Assuming that the Pprp Parp was measured with the left and right input channels and stored in a data set named Pdrp Pmrp then the STMR Calculator controls would be set to the following Juse output voltage to MRP pressure refrence transfer Function use unit turned off transfer Function to correct For direct MRP to ear audio path Data set w unit turned
70. the necessary FFT spectrums so the measurement results will still be accurate Calculate WAV file metrics Pressing this button will cause the WAV file metrics to be computed and displayed in a dialog box An example is shown below These metrics are computed without and with the WAV gain applied to the WAV file data JiJasonitempilarry english 44p1khz way sample rate 44 1 kHz Only the left channel data is used For a WG File measurement Left channel metrics wo gain head room 0 dB RMS level 0 207836 DFS 0 1 normalized 13 6456 dB DFS 0 1 normalized 7 49558 dBrmO Left channel metrics with gain head room 0 915209 dB RMS level 0 187043 DFS 0 1 normalized 14 5611 dB DFS 0 1 normalized 6 41115 dBm0 Above text copied to clipboard headroom this is the maximum gain that can be applied to the WAV file without clipping RMS level the RMS level of the WAV file data start and stop frequencies The WAV file measurement uses a series of short time FFT s to process the output and input signals The start and stop frequencies specify the lower and upper bound of the FFT results that are kept The frequencies outside of this range are simply discarded Octave bands The short time FFT power spectrums are summed into octave bands to produce the spectrum data 1 3 R10 to 1 24 R80 bands are supported Copyright 2012 2013 McIntosh Applied Engineering LLC 31 T FFT frame This parameter specifies the
71. tible with external microphones providing both phantom power and a considerable amount of gain In this case the large gain will likely cause the Maximum peak to base input level to be very small perhaps less than 100 mV Since most signals from an amplified microphone or voltage across a speaker will be much higher than 100 mV the gain on the sound card inputs may need to be reduced The user will need to adjust the input gains to achieve a desired maximum input level Typically a maximum input level between 1 and 10 V is desirable for most applications However if non amplified microphones are being measured perhaps a high input gain is exactly what s needed If the maximum peak to base input level is less than 100 mV the MAE200 ACVC1 A cannot be used to calibrate the input voltages Since every sound card driver interface is different due to various drivers and Windows versions the user must figure out how to adjust the input gains for their particular sound card However in general you want to go to the Control Panel select Sound select Recording right click on the sound card being used for recording and then go through the controls to adjust the microphone gain The dialog boxes for doing this in Windows 7 are shown below Copyright 2012 2013 McIntosh Applied Engineering LLC 13 Sound Playback Recording Sounds Communications Select a recording device below to modify its settings microphone gain
72. used to control the B amp K NEXUS amplifier If calibrated microphone levels are required it s extremely convenient to simply have Ares read the calibration levels directly from a B amp K NEXUS amplifier While it s not necessary to use a NEXUS amplifier using a NEXUS is highly recommended It allows for the microphone gain to be changed without having to manually change the microphone s sensitivity in the Ares interface preventing the accidental loss of proper calibration during measurements Acquisition Controls The Acquisition tab is where most of the measurements are performed It consists of several sets of controls selected with the Type of measurement to perform list box The list box is shown below soundcard calibration soundcard calibration stepped sine measurement chirp measurement WAY file measurement soundcard calibration The sound card input and output calibration values are specified in this section The controls change slightly depending on whether the inputs or outputs are specified to be analog voltages Copyright 2012 2013 McIntosh Applied Engineering LLC 11 microphone inputs or digital levels The screen capture below and to the left is for when input and outputs are both voltages and the capture on the right is for when they are both digital i Ea File Edit view New Module Modules Tools Timedata MeN A IETT O Acquisition Setup Plotting Files Clipboard Type of
73. wModule Modules Tools Time data Window Help LF bel AGO Acquisition Setup Plotting Files Clipboard left chan RMS V 1 iteration left chan RMS V 2 iterations left chan RMS V 3 iterations Transfer function graphs C magnitude C phase C real amp imag left inputfright input C left inputfoutput C right inputfoutput Other graphs magnitude SNR CI THD N left input channel C right input channel C output channel Plotting Options C dB magnitude log frequency Data sets 1 iteration 2 iterations Channel magnitude plot data for selected data set 3000 Frequency Hz The initial output parameter determines the constant output level that s used for the first measurement After the first measurement the output will vary with frequency to achieve the RMS target value adj output using freq vs amplitude data file This option takes the output drive level from a text data file For this option a single extra control is created that allows you to browse for the data file that s to be used adjust using frequency amplitude data file w freg amp t t C Per browse button The format of the data file is to be two numerical text columns separated by spaces The first column is frequency and the second column is the RMS output level for the corresponding frequency An example of the contents of a data file is given below It consists of three lines

Manual for the Ares Frequency Response Measurement Module

Contents

Download Pdf Manuals

Related Search

Related Contents