avfoundation - 使用 CallKit 的 VOIP 应用程序中的音频短路

Question

我正在使用 SpeakerBox应用程序作为我的 VOIP 应用程序的基础。我已经设法让一切正常工作，但我似乎无法摆脱从麦克风到设备扬声器的音频“短路”。

换句话说，当我打电话时，我可以在扬声器中听到自己的声音以及对方的声音。我该如何更改？

AVAudioSession 设置:

    AVAudioSession *sessionInstance = [AVAudioSession sharedInstance];

    NSError *error = nil;
    [sessionInstance setCategory:AVAudioSessionCategoryPlayAndRecord error:&error];
    XThrowIfError((OSStatus)error.code, "couldn't set session's audio category");

    [sessionInstance setMode:AVAudioSessionModeVoiceChat error:&error];
    XThrowIfError((OSStatus)error.code, "couldn't set session's audio mode");

    NSTimeInterval bufferDuration = .005;
    [sessionInstance setPreferredIOBufferDuration:bufferDuration error:&error];
    XThrowIfError((OSStatus)error.code, "couldn't set session's I/O buffer duration");

    [sessionInstance setPreferredSampleRate:44100 error:&error];
    XThrowIfError((OSStatus)error.code, "couldn't set session's preferred sample rate");

IO 单元的设置:

- (void)setupIOUnit
{
try {
    // Create a new instance of Apple Voice Processing IO

    AudioComponentDescription desc;
    desc.componentType = kAudioUnitType_Output;
    desc.componentSubType = kAudioUnitSubType_VoiceProcessingIO;
    desc.componentManufacturer = kAudioUnitManufacturer_Apple;
    desc.componentFlags = 0;
    desc.componentFlagsMask = 0;

    AudioComponent comp = AudioComponentFindNext(NULL, &desc);
    XThrowIfError(AudioComponentInstanceNew(comp, &_rioUnit), "couldn't create a new instance of Apple Voice Processing IO");

    //  Enable input and output on Apple Voice Processing IO
    //  Input is enabled on the input scope of the input element
    //  Output is enabled on the output scope of the output element

    UInt32 one = 1;
    XThrowIfError(AudioUnitSetProperty(_rioUnit, kAudioOutputUnitProperty_EnableIO, kAudioUnitScope_Input, 1, &one, sizeof(one)), "could not enable input on Apple Voice Processing IO");
    XThrowIfError(AudioUnitSetProperty(_rioUnit, kAudioOutputUnitProperty_EnableIO, kAudioUnitScope_Output, 0, &one, sizeof(one)), "could not enable output on Apple Voice Processing IO");

    // Explicitly set the input and output client formats
    // sample rate = 44100, num channels = 1, format = 32 bit floating point

    CAStreamBasicDescription ioFormat = CAStreamBasicDescription(44100, 1, CAStreamBasicDescription::kPCMFormatFloat32, false);
    XThrowIfError(AudioUnitSetProperty(_rioUnit, kAudioUnitProperty_StreamFormat, kAudioUnitScope_Output, 1, &ioFormat, sizeof(ioFormat)), "couldn't set the input client format on Apple Voice Processing IO");
    XThrowIfError(AudioUnitSetProperty(_rioUnit, kAudioUnitProperty_StreamFormat, kAudioUnitScope_Input, 0, &ioFormat, sizeof(ioFormat)), "couldn't set the output client format on Apple Voice Processing IO");

    // Set the MaximumFramesPerSlice property. This property is used to describe to an audio unit the maximum number
    // of samples it will be asked to produce on any single given call to AudioUnitRender
    UInt32 maxFramesPerSlice = 4096;
    XThrowIfError(AudioUnitSetProperty(_rioUnit, kAudioUnitProperty_MaximumFramesPerSlice, kAudioUnitScope_Global, 0, &maxFramesPerSlice, sizeof(UInt32)), "couldn't set max frames per slice on Apple Voice Processing IO");

    // Get the property value back from Apple Voice Processing IO. We are going to use this value to allocate buffers accordingly
    UInt32 propSize = sizeof(UInt32);
    XThrowIfError(AudioUnitGetProperty(_rioUnit, kAudioUnitProperty_MaximumFramesPerSlice, kAudioUnitScope_Global, 0, &maxFramesPerSlice, &propSize), "couldn't get max frames per slice on Apple Voice Processing IO");

    // We need references to certain data in the render callback
    // This simple struct is used to hold that information

    cd.rioUnit = _rioUnit;
    cd.muteAudio = &_muteAudio;
    cd.audioChainIsBeingReconstructed = &_audioChainIsBeingReconstructed;

    // Set the render callback on Apple Voice Processing IO
    AURenderCallbackStruct renderCallback;
    renderCallback.inputProc = performRender;
    renderCallback.inputProcRefCon = NULL;
    XThrowIfError(AudioUnitSetProperty(_rioUnit, kAudioUnitProperty_SetRenderCallback, kAudioUnitScope_Input, 0, &renderCallback, sizeof(renderCallback)), "couldn't set render callback on Apple Voice Processing IO");

    // Initialize the Apple Voice Processing IO instance
    XThrowIfError(AudioUnitInitialize(_rioUnit), "couldn't initialize Apple Voice Processing IO instance");
}

catch (CAXException &e) {
    NSLog(@"Error returned from setupIOUnit: %d: %s", (int)e.mError, e.mOperation);
}
catch (...) {
    NSLog(@"Unknown error returned from setupIOUnit");
}

return;
}

启动 IOUnit:

NSError *error = nil;
[[AVAudioSession sharedInstance] setActive:YES error:&error];
if (nil != error) NSLog(@"AVAudioSession set active (TRUE) failed with error: %@", error);

OSStatus err = AudioOutputUnitStart(_rioUnit);
if (err) NSLog(@"couldn't start Apple Voice Processing IO: %d", (int)err);
return err;

停止 IOUnit

NSError *error = nil;
[[AVAudioSession sharedInstance] setActive:NO withOptions:AVAudioSessionSetActiveOptionNotifyOthersOnDeactivation error:&error];
if (nil != error) NSLog(@"AVAudioSession set active (FALSE) failed with error: %@", error);

OSStatus err = AudioOutputUnitStop(_rioUnit);
if (err) NSLog(@"couldn't stop Apple Voice Processing IO: %d", (int)err);
return err;

我使用 PJSIP 作为我的 SIP 栈，并且有一个 Asterisk 服务器。问题必须出在客户端，因为我们也有一个没有这个问题的基于 Android 的 PJSIP 实现。

Answer 1

我在使用 WebRTC 时遇到了同样的问题。我最终得出的结论是，您不应该在 AudioController.mm 中设置 IOUnit，而是将其留给 PJSIP(在我的例子中是 WebRTC)。

快速修复如下: 注释掉AudioController.mm的setupAudioChain中的[self setupIOUnit];和didActivate audioSession中的startAudio()的 ProviderDelegate.swift。