Creating Calibration Files

[rex.b...@dcineco.com]

Hi Julian,

I've created an Excel tool to help with calibrating AudioTool platforms. Here's how it works:

Plug a source of pink noise into the device's mic input. This should be a good, flat source. A test CD on a player with a line output should work or you could use a PC program like Audacity if you have a USB sound adaptor. I would not trust the headphone output of any phone or PC. I used an old NTI MR1 which is pretty good. You will need an appropriate adaptor or cable to connect to the mic terminals on the 3.5mm jack. I've seen them on eBay. I made mine by sacrificing an old headset and soldering the cable to an XLR Female. Set the level to give a similar reading as you get from your measurement mic. Set Weighting Flat, Averaging On and Use Single Calibration. Once it has settled, Store the spectrum. This should be the frequency response of the mic input of your device. Copy the at file to a PC.

Open up the AudioTool Calibrate Excel and click the Load Pink AT File button. Select the file you just stored and wait a few seconds. The tool will analyse and plot the 1/3rd octave response of the mic input. The response figures are relative to that at 1kHz.

Now you can add frequency response factors for your measurement mic for each band under Mic dB. This is often provided as a plot or table with the mic or from the spec-sheet. If the plot shows the mic response as -3dB at 100Hz, for example, enter -3.0 in the Mic dB field on the 100Hz row. You can also make adjustments if you want to tailor the overall response by setting offsets in Trim dB. The tool shows plots of the initial pink response, the mic response, the calculated calibration curve and the theoretical final result.

Click the Make CAL File button and chose an appropriate name and path and the tool will create the cal file. Copy this to your device and load it into AudioTool with the Load Cal button. You will probably now need to recalibrate the single-point adjustment for SPL using a mic calibrator or known accurate reference meter. In my case, all the correction was at the low end and it had almost no effect on the SPL calibration. 

Just to be clear: This is not a substitute for proper calibration with reference sources and instruments. It is a technique that may help to improve the accuracy of an otherwise uncalibrated setup. It should also be noted that, left to its own devices, the tool will attempt to level bands that have little or no response at all. The resulting cal file could give strange results in AudioTool. I recommend using the Trim dB factors to curtail crazy corrections and generally smooth the calibration curve to a sensible shape. As with sound systems, it is often better to accept a few ripples and some of the natural roll-off at the top and bottom ends.

Anyway, here it is. I hope it helps somebody.

Rex 

[Julian Bunn]

Hi Rex,

Fantastic! I will add this to my Web pages (with credit) and link it from the AudioTool online manual, if you have no objections?

Julian

--
You received this message because you are subscribed to the Google Groups "AudioTool Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussio...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[rex.b...@dcineco.com]

Hi Julian,

No objections as long as folk understand it is an approximate calibration. 

Meanwhile, I had another idea. The attached version (V 1.1) now allows an iMM6 calibration text file to be loaded and applied to the mix. I don't have an actual mic but I tested it with a text file someone posted earlier and it seems about right.

Rex

To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussion-group+unsub...@googlegroups.com.

[rex.b...@dcineco.com]

Hi Julian,

I added calibration pages for MicW i436 and NTI miniSPL mics based on the manufacturer's specified response data. V 1.2 attached. These are mics that I have and I trust the published data. I am happy to add other mic profiles if the response data is available. Once you have calibrated the mic input of your device, the tool makes it easy to create cal files for any mics you may want to use.

Rex

[rex.b...@dcineco.com]

Hi Julian,

I found an error in the calculation and display of the corrected result. It didn't affect the cal file, though. I also tidied-up the management of Mic calibration factors. V 1.3 attached.

Rex

On Friday, October 11, 2013 6:35:45 PM UTC+1, Julian wrote:

To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussion-group+unsub...@googlegroups.com.

[rex.b...@dcineco.com]

Hi Julian,

I found another error when calibrating my Nexus7 - which only has 10Hz FFT bins so doesn't have a 25Hz value. V 1.3.1 attached uses an average of the adjacent bins instead.

Rex 

On Friday, October 11, 2013 6:35:45 PM UTC+1, Julian wrote:

To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussion-group+unsub...@googlegroups.com.

[Julian Bunn]

Hi Rex,

Thanks for the update. This is now on the website:

https://sites.google.com/site/bofinit/audiotool-calibration-method

and I also put a link to it from the AudioTool online manual.

Thanks again!

Julian

On Sun, Oct 13, 2013 at 8:37 AM, <rex.b...@dcineco.com> wrote:

Hi Julian,

I found another error when calibrating my Nexus7 - which only has 10Hz FFT bins so doesn't have a 25Hz value. V 1.3.1 attached uses an average of the adjacent bins instead.

Rex 

On Friday, October 11, 2013 6:35:45 PM UTC+1, Julian wrote:

Hi Rex,

Fantastic! I will add this to my Web pages (with credit) and link it from the AudioTool online manual, if you have no objections?

Julian

On Fri, Oct 11, 2013 at 9:13 AM, <rex.b...@dcineco.com> wrote:

Hi Julian,

I've created an Excel tool to help with calibrating AudioTool platforms. Here's how it works:

Plug a source of pink noise into the device's mic input. This should be a good, flat source. A test CD on a player with a line output should work or you could use a PC program like Audacity if you have a USB sound adaptor. I would not trust the headphone output of any phone or PC. I used an old NTI MR1 which is pretty good. You will need an appropriate adaptor or cable to connect to the mic terminals on the 3.5mm jack. I've seen them on eBay. I made mine by sacrificing an old headset and soldering the cable to an XLR Female. Set the level to give a similar reading as you get from your measurement mic. Set Weighting Flat, Averaging On and Use Single Calibration. Once it has settled, Store the spectrum. This should be the frequency response of the mic input of your device. Copy the at file to a PC.

Open up the AudioTool Calibrate Excel and click the Load Pink AT File button. Select the file you just stored and wait a few seconds. The tool will analyse and plot the 1/3rd octave response of the mic input. The response figures are relative to that at 1kHz.

Now you can add frequency response factors for your measurement mic for each band under Mic dB. This is often provided as a plot or table with the mic or from the spec-sheet. If the plot shows the mic response as -3dB at 100Hz, for example, enter -3.0 in the Mic dB field on the 100Hz row. You can also make adjustments if you want to tailor the overall response by setting offsets in Trim dB. The tool shows plots of the initial pink response, the mic response, the calculated calibration curve and the theoretical final result.

Click the Make CAL File button and chose an appropriate name and path and the tool will create the cal file. Copy this to your device and load it into AudioTool with the Load Cal button. You will probably now need to recalibrate the single-point adjustment for SPL using a mic calibrator or known accurate reference meter. In my case, all the correction was at the low end and it had almost no effect on the SPL calibration. 

Just to be clear: This is not a substitute for proper calibration with reference sources and instruments. It is a technique that may help to improve the accuracy of an otherwise uncalibrated setup. It should also be noted that, left to its own devices, the tool will attempt to level bands that have little or no response at all. The resulting cal file could give strange results in AudioTool. I recommend using the Trim dB factors to curtail crazy corrections and generally smooth the calibration curve to a sensible shape. As with sound systems, it is often better to accept a few ripples and some of the natural roll-off at the top and bottom ends.

Anyway, here it is. I hope it helps somebody.

Rex 

--
You received this message because you are subscribed to the Google Groups "AudioTool Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussion-group+unsub...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "AudioTool Discussion Group" group.

To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussio...@googlegroups.com.

[rex.b...@dcineco.com]

Hi Julian,

Let's hope it helps a few people to get more accurate results. It may also startle folk when they see how bad the LF response of their phone's mic input is. I was shocked at what I found on my Nexus7.

Rex

To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussion-group+unsubsc...@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

[Andre Grobler]

Very handy... if I understand correctly you could use this to also just calibrate the mic input regardless of the mic used. Only purpose it has is to get a slightly better response with any mic... not that you know anything absolute. It just takes one more nonlinearity away?

[rex.b...@dcineco.com]

Yes you could leave the mic calibration factors set to zero. In most cases, the response is largely down to the mic input anyway - particularly at the LF end of the spectrum. Most measurement mics are within a couple of dBs of flat across the range.

Once you have the response reasonably flat, you can set the absolute calibration using the cal factor. You can do this with a piston mic calibrator or by reference to a known accurate instrument. 

Rex

[Andre Grobler]

Can see what u mean by funky inputs. The load cal button does not appear even though there i selected 1/3 octave calibration... it isnt in menu key either. Any ideas latest android...

[Andre Grobler]

Aha... the button name is not what it is, but what it will be... sigh no wonder all my home built speakers have the same silly fr... don't worry all will be sorted. I think

[Andre Grobler]

New question

Is there a surefire way to undo any calibration whatsoever, it seems I finalised the first (wrong) curve - which just corrects for the mic input response - no mic attached... and now i need to load a correct curve, however it is not possible to create a correct pink noise file, if you already have the wrong calibration curve in there or do I misunderstand someting?

[Julian Bunn]

I'm not sure I understand the question. You need to load a second curve that adds to the currently loaded curve?

Julian

On Mon, Nov 11, 2013 at 12:59 AM, Andre Grobler <andregro...@gmail.com> wrote:

New question

Is there a surefire way to undo any calibration whatsoever, it seems I finalised the first (wrong) curve - which just corrects for the mic input response - no mic attached... and now i need to load a correct curve, however it is not possible to create a correct pink noise file, if you already have the wrong calibration curve in there or do I misunderstand someting?

--

You received this message because you are subscribed to the Google Groups "AudioTool Discussion Group" group.

To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussio...@googlegroups.com.

[Andre Grobler]

Sorry about lack of clarity. I made a pink noise file with wrong settings, and made a cal file using it. Loaded that cal file which is wrong. Now I cant make a new correct pink noise file without that original wrong calibration affecting it. However I just unistalled and removed audiotool folder and redid it. Seems to have worked. Wouls be nicw if there is an option to delete the existing calibration before doing anew one

[rex.b...@dcineco.com]

You should set AudioTool to use Single Calibration before capturing the pink response (see first post). Then it doesn't matter what cal file you had loaded. Its OK to have several different cal files. AudioTool uses the last one you loaded. I have one for each mic I use. BTW: I also saved a cal file with no adjustments called Flat for when I want to compare results.

[rex.b...@dcineco.com]

Hi Julian,

After seeing AudioTool in action on my HTC One, a colleague installed iAudioTool on his iPhone. Now he wants to calibrate it... From observation, iAudioTools appears to have only octave calibration factors. Is this true for all versions? Could you post a flat cal file for iAudioTool so I can see what I need to add to the spreadsheet? Is the format of the at file the same as for full-strength AudioTool?

  

[Julian Bunn]

iAudioTool for iOS currently only supports Octave calibration (and single value). 

A flat cal file has all zeroes ... if you save a cal file and look at it in a text editor, you can simply set all the values to zero and use it like that.

The .at file format is the same between iOS and Android versions.

Julian

On Tue, Nov 12, 2013 at 5:23 AM, <rex.b...@dcineco.com> wrote:

Hi Julian,

After seeing AudioTool in action on my HTC One, a colleague installed iAudioTool on his iPhone. Now he wants to calibrate it... From observation, iAudioTools appears to have only octave calibration factors. Is this true for all versions? Could you post a flat cal file for iAudioTool so I can see what I need to add to the spreadsheet? Is the format of the at file the same as for full-strength AudioTool?

  

[rex.b...@dcineco.com]

Thanks, Julian.

I'm happy to hear that iAudioTool uses the same at file format. My question about the flat octave cal file was really about the frequency bands. I don't have an iPhone so I can't save a file and look at it. Can I pick my own octave frequencies or do they need to match the ones that iAudioTool uses? I would have guessed 20, 40, 80, 160 ... 20K but when I set my Android AudioTool to octave mode the centre frequencies are 31, 63, 125...16K. What should I use for iAudioTool?

Rex 

To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussion-group+unsub...@googlegroups.com.

[Julian Bunn]

Hi Rex,

Sorry, I misunderstood :-)

The octave band centre frequencies used in iAudioTool are at the standard ISO values:

31.5

63

125

250

500

1000

2000

4000

8000

16000

Hope this helps.

Julian

To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussio...@googlegroups.com.

[rex.b...@dcineco.com]

Many thanks, Julian.

Rex

To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussion-group+unsubsc...@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

[Andre Grobler]

Excellent Thanks!

[hpau...@googlemail.com]

Hi,

Is IT just me or does someone else get the message that the Makro is broken when clicking on Load File? I tried different ms Office Versions and libre Office under linux. Same Problem. Even on another Computer. I also tried different Versions of the exel File.

[rex.b...@dcineco.com]

I have updated the Excel tool for preparing AudioTool calibration files. The new version uses the 1/3rd Octave data that Julian has helpfully appended to the captured spectrum (at) files. This allows it to work with files captured with FFT resolutions of 4096, 8192 or even 16384 (fine). It is also much faster!

As well as processing the captured pink response, the tool can load calibration files for Dayton iMM6 and miniDSP UMIK-1 measurement microphones. It can also be used for manual entry of microphone response factors.

The process is essentially the same as that described in the first post except that microphone responses are managed on dedicated worksheets. Once the microphone response has been loaded or entered, it can be sent to the Main sheet by clicking the Send to Main button. In this way, cal files for several different mics can be prepared for the same pink response. 

I think this tool should be able to cope with at files produced by recent versions of AudioTool. If you do have any problems, please email me with a copy of your at file.

Rex

[Julian Bunn]

Excellent work, Rex! Many thanks for making this available to AudioTool users!

Julian

--

You received this message because you are subscribed to the Google Groups "AudioTool Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussion-group+unsub...@googlegroups.com.

To post to this group, send email to audiotool-discussion-group@googlegroups.com.
Visit this group at https://groups.google.com/group/audiotool-discussion-group.
For more options, visit https://groups.google.com/d/optout.

[rex.b...@dcineco.com]

Thank you, Julian. AudioTool is one of the most useful apps on my phone. I'm happy to help other users make it as accurate as possible.

Rex