Compression audio audit files

Hi Survey Solutions Team

I recently came across audio audit files in another application that were quite a bit smaller in size compared to what exports from Survey Solutions. Generally, one does not need high quality audio, as long as one can understand the interviewers in the audio it is fine. Is it possible to reduce the quality of the recorded audios? The benefits would be:

  • Faster synchronization between tablets and server
  • Faster export
  • Less data volume to handle if using audio recordings
  • Possibility of recording longer and more interviews. In this other application we had recorded all 1-1.5 hr long interviews. It has been very useful for field work supervision and making corrections.

Hello Andreas,

  • what is the “other application”? and
  • what file format and codec does it use?

Thank you, Sergiy

Hi Sergiy

It was a custom build platform in CSPro. The file format is .m4a, will check with the person who put it together on the codec and get back to you.

Thanks

Survey Solutions also uses *.m4a as mentioned here.

Since audio files in this format may be produced with different codecs and at different bitrates, your mystery developer has probably opted for a notably lower bitrate to obtain notably smaller files (with corresponding losses in quality).

Afaik, there is no parameter in Survey Solutions to adjust the bitrate by the end user.

Hi Sergiy

So, they said they used mpeg 4, and that one hour of audio is around 5.5 M. Yes, I guess they used a low bit rate. The quality is obviously not great, but they have found a good balance with the sound quality being good enough to hear the interviewer clearly, and the file size being small enough for data transfer in low connectivity settings.

MPEG-4 can be utilizing any of the codecs here:


clearly some of them may be more advantageous even for the same bitrates. One needs to also incorporate:
  • availability for projects (is it free? can it be distributed by free software?)
  • cost (if it is not free)?
  • dissemination (are there any restrictions for use in any countries? patent collisions or restrictions?)
  • packaging (whether it can be installed as part of the software or has to be acquired, installed, and configured by the end user)
  • cross-platform implementation (must work on Windows/Mac/Linux/Android)
  • cross-browser implementation (must work with Chrome, FireFox, Edge, etc).
  • license implications (does it require the using software to be open source? what type?)
    … and probably a dozen more factors…

So yes, it would be useful to know exactly what the combination of parameters (format, codec, bitrate, mono/stereo, etc) were used to achieve this. You can probably uncover this by inspecting your *.m4a files’ properties even without troubling the developer.

Dear @sergiy ,
It’s also a question of give users more power over application. CSPro audio object allow to save files dynamically with questions name, questionnaire ID etc. So, for sync facilitation, we split audio files on device in multiples small ones. In the server, we can use batch application to concat them.

It’s long ago we request true programming support on Survey Solutions.

Best,