Photon Voice suggested sampling rate for speech?

Hi! What is the suggested Photon Voice Sampling Rate when the room capacity is from 1-8 people and the audio is exclusively speech (but I'm using the WebRTCAudioDsp so can't use 24k)? I.e. is 48k overkill and should I be using 16k?

I will do further testing, but a view by Photon on this would be very helpful, especially as I work alone (so I can't quickly do voice chat tests with many people). Thanks!


  • JohnTube
    JohnTube mod
    edited October 2019
    Hi @Philipp,

    Use 16k, if you or your users have complaints about audio quality switch to 48k.
    You could also do this dynamically at runtime or even allow users to change voice quality settings from the app, you could replace values with words: Medium, High, etc.
  • Thanks for the insight!
  • Philipp
    edited October 2019
    [Moved followup question from here into distinct new thread for ease of answering]