feat: Add AudioOptions class

[karasusan]

• [REQUEST]: Configurable audio options (noise suppression, auto gain, etc.) #587
• [FEATURE]: Remove echo on iOS #651
• [BUG]: unity Webrtc AudioSource noise issue #903
• [BUG]: The sound is captured back to the microphone #964
• webrtc aec support UnityRenderStreaming#563

This PR adds the AudioOptions class to set audio options to tracks.
The options are defined by the AudioOptions class.

class AudioOptions
{
        public bool? echoCancellation;
        public bool? autoGainControl;
        public bool? noiseSuppression;
        public bool? highpassFilter;
}

This PR makes AudioStreamTrack constructor changing.

• AudioStreamTrack(AudioOptions options = null);
• AudioStreamTrack(AudioSource source, AudioOptions options = null)
• AudioStreamTrack(AudioListener listener, AudioOptions options = null)

[flamacore]

So hey, I've followed the issues that are mentioned in this PR and I also would like to give a try to the audio options.

Sadly I have problems with getting this specific pr to build as I'm not very well versed with CMake and the like. It throws an error about not finding presets and shuts down.

Anyhow, is there any progress on making this a reality? Any sort of easier way I can add this and experiment with it in my own project yet?

[karasusan]

@flamacore
Hi, if you can help us to test the audio option, I'll be glad to take you a package.

Here you are.
https://drive.google.com/file/d/1Qf4ZLTqNxT-hoLfgJj3JZIpFu3hEPLNI/view?usp=drive_link

Please install the tgz file following this manual.
https://docs.unity3d.com/Manual/upm-ui-tarball.html

[flamacore]

@karasusan would be delighted to. Thanks a lot for sharing the package. I'll give it a check soon and will inform if it works or not.

[MaximKurbanov]

I tested and issue #964 remains.
I tried this with android-android and windows-android:

  #...
    {
  #...
        var track = new AudioStreamTrack(sendAudioSource, GetAudioOptions());
        mediaStream.AddTrack(track);
    }

    private AudioOptions GetAudioOptions()
    {
        return new AudioOptions()
        {
            echoCancellation = true,
            noiseSuppression = true
        };
    }

[flamacore]

Can confirm the echo issue persists. Not sure if it's related to us enabling/disabling the options or if the internal echo cancellation works as intended. EDIT: I tried it multiple ways:

• iOS - iOS
• iOS - Android
• iOS - Windows
• Android Windows

[karasusan]

@MaximKurbanov
@flamacore

Thank you for checking. I'd like to clarify which this issue is a specific by mobile platforms or not.
For example, windows and windows, or windows or macOS.

It looks there are different implementations for mobile and desktop platforms.
https://stackoverflow.com/questions/62479789/webrtc-android-echo-cancellation

[MaximKurbanov]

@karasusan
I tested Windows-Mac, AEC does not work.

[karasusan]

@MaximKurbanov
I understood the cause of this issue.
Sorry but we need more time to fix it.

I can't share you many information about that, but at least I can say that the "audio processing" in native code is not processed.

The "audio processing" in webrtc is implemented here.
https://source.chromium.org/chromium/chromium/src/+/main:third_party/webrtc/modules/audio_processing/audio_processing_impl.cc

[pixlhero]

@karasusan
I tested the package from Oct 6th on Mac-Editor <-> iPad.
It also didn't cancel the echo.

This is a show-stopper for us. So we really, really need this fix in the next release (pre-8). 🙏

Also building the plugin doesn't work for us. So if 0272460 would fix anything (we're targeting iOS) it would be great if you could provide us with a .tgz build.

がんばってよ!

[pixlhero]

@karasusan I tested the package from Oct 6th on Mac-Editor <-> iPad. It also didn't cancel the echo.

This is a show-stopper for us. So we really, really need this fix in the next release (pre-8). 🙏

Also building the plugin doesn't work for us. So if 0272460 would fix anything (we're targeting iOS) it would be great if you could provide us with a .tgz build.

がんばってよ!

We found a workaround on iOS involving AVAudioSession. So for our project this issue is resolved.
https://gist.github.com/pixlhero/3d370dcdbcfaefa4df12138161307da4