Microphones for HTC U11 Phone

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

I recently got an HTC U11 which, like many new phones, no longer has an analogue 3.5mm jack socket. The good news is that it works very well with a miniDSP UMIK-1 USB mic. I connect this using a short USB-Mini/USB-Micro OTG cable and a USB-Micro/USB-C adaptor. When connected, the phone switches to the UMIK-1 automatically and AudioTool sees it immediately. I used the downloaded 90 degree file from the miniDSP website to create a 1/3rd octave cal file and just required a +2dB Offset to give the correct SPL reading for my mic calibrator.

The HTC U11 also ships with an adaptor to allow legacy analogue headsets to be connected to the USB-C port. Out of interest, I checked to see if this would also permit the use of analogue measurement mics like MicW i436 and Dayton iMM6. The answer is "yes, but...". When connected through the adaptor, AudioTool sees the mic immediately. So far, so good. Then I ran through my procedure to evaluate the response of the mic input using a pink-noise source. It's very good between 80Hz and 16kHz but is 6dB down at 20kHz and 32dB down at 20Hz. Whilst this can be corrected using my spreadsheet, I don't know how much I would trust any measurements at the LF end of the spectrum.

My conclusion is that the best mic to use with an HTC U11 is a miniDSP UMIK-1.

[Julian Bunn]

Fascinating report: many thanks, Rex!

Does it seem that the HTC U11 has some sort of aggressive filtering on the analogue headset interface then?

--
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]

On Monday, 6 November 2017 22:04:45 UTC, Julian wrote:

Fascinating report: many thanks, Rex!

Does it seem that the HTC U11 has some sort of aggressive filtering on the analogue headset interface then?

Yes it's more than 20dB/octave at the bottom end. I suppose it makes sense for its intended use with speech and voice commands. When I get chance, I'll try a different external converter to see if that has a more useful response.

BTW: I was sifting through some capture (at) files from old phones. Some have Length: 4096 and some Length: 8192. Both types were captured using normal (not Fine) frequency resolution. What determines the length? I only ask because they need different handling in my spreadsheet.

[Julian Bunn]

The sample length is determined by a few factors:

1) the minumum size that the Android OS reports is the minimum that can be used,

2) at least 4096 samples

3) a power of 2

4) whether the user has chosen "Fine"

So, if the OS reports the minimum is 3000, for example, the buffer size will be set to 4096. As a counter example, if the OS reports the minimum is 5000, the buffer size will be set to 8192.

When the user chooses "Fine" frequency, it sets the sample length to 16384. This yields a typical frequency resolution of around 2Hz at a sample rate of 44100.

Julian

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

Thanks for the explanation, Julian. I found that I had two versions of the spreadsheet - one that worked with 4096 and the other for 8192. I shall try to come-up with one that handles both sizes.

Best regards, Rex 

 

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

Further testing reveals that the HTC U11 applies digital filtering to apparently all external A/D Mic inputs - including the UMIK-1. The LF roll-off is practically a brick-wall filter that is -3dB at 63Hz, -10dB at 40Hz and -32dB at 20Hz. I have had to revert to my old calibration technique to compensate. I had hoped that, by using a USB Mic, that was no longer necessary. Oh well...  

 

I wonder if similar filtering is used on other phones that have omitted an analogue Mic input. Has anyone else run a response test for their digital Mic input?

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

As previously mentioned, I could see that the HTC U11 was applying LF and HF filtering to USB mic inputs - including the miniDSP UMIK-1. What I needed to know was exactly how much roll-off was being applied. This is not just my typical Engineer's OCD showing. I use AudioTool to check and adjust cinema auditorium sound systems and I have to know what is happening at the extremes of the audio band. Without easy access to a full-spectrum SPL reference source, I tried a different approach.

I took a USB external sound interface (mine was UGreen USB 2.0 External Stereo Sound Adapter) and connected it to the U11 using a USB-A/USB-C OTG adapter. I connected a known flat pink-noise source to the mic input and adjusted the level to give the SPL reading I normally use for room checks (85dBC). I set AudioTool to Fine resolution, Flat response, Averaging on and stored the Spectrum file. 

I then connected the UGreen adaptor, with the same pink source, to a PC and used Room EQ Wizard's RTA function to capture its response. Using an Excel worksheet, I subtracted the response of the adaptor from the stored Spectrum file to show the difference due to the U11's filtering. I used this as the Pink response with my Excel tool to generate a calibration file for the combined response of the U11 with UMIK-1. I show the three response plots below. The result is that I now get sensible readings that I can trust - and it only took about ten minutes.

Am I the only AudioTool user who cares about this level of accuracy? Possibly but, if you are using a UMIK-1 to measure below 100Hz or above 10kHz, you may want to check whether your phone is telling you the truth!

  

[Julian Bunn]

Very useful, Rex! Thanks for posting :-)

Julian

--

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

In the somewhat unlikely event that anyone else wants to perform this type of calibration, I have attached the Excel tool I use to process the data. You need to have the following files available: An AudioTool at Spectrum from pink-noise through an external USB sound adaptor - set Use Single Point Calibration, Fine response, Weighting Flat and Averaging on; A RoomEQWizard response of the same adaptor and pink-noise on a PC - set Weighting Flat (Z) and Average 32 - export the measurement to a txt file using tab separators. 

Open the Excel tool and click Load Pink AT File to load the AudioTool response file. Now click Load REW File to load the REW response file. Clicking Make Adjusted AT File will generate a corrected at file. This file is the EQ/filtering that your phone applies to USB mic inputs. It can be used as the Pink AT File in my AudioTool Calibrate 2.0.0 tool to create the correct calibration for your USB measurement mic.

I show below the results for my HTC U11 phone with a UGreen USB 2.0 External Stereo Sound Adapter. As you can see, the phone applies a lot of roll-off at the LF end - but not as much as is shown by the uncorrected Pink response.