Using RemoteIO audio unit

By MICHAEL TYSON | Published: NOVEMBER 4, 2008

I’ve had nasty old time trying to get some audio stuff going on the iPhone, no thanks to Apple’s lack of documentation. I have a feeling it’s all still quite hush-hush, so no details here, but if you’re an iPhone developer interested in getting RemoteIO/IO Remote/whatever it’s called working on the iPhone… Do I have good news for you.

Read on here, at the Developer forums

Drop me a line if you find this helpful.

Update: I’m told the new NDA is pretty much all-good with blog postings. So, read on for the goods.

Update 2: Thanks to Joel Reymont, we now have an explanation for the “CrashIfClientProvidedBogusAudioBufferList” iPhone simulator bug: The simulator doesn’t like mono audio. Thanks, Joel!

Update 3: Happily, Apple have now created some excellent documentation on Remote IO, with some good sample projects. I recommend using that as a resource, now that it’s there, as that will continue to be updated.

So, we need to obtain an instance of the RemoteIO audio unit, configure it, and hook it up to a recording callback, which is used to notify you that there is data ready to be grabbed, and where you pull the data from the audio unit.

Overview

1. Identify the audio component (kAudioUnitType_Output/ kAudioUnitSubType_RemoteIO/ kAudioUnitManufacturerApple)
2. Use AudioComponentFindNext(NULL, &descriptionOfAudioComponent) to obtain the AudioComponent, which is like the factory with which you obtain the audio unit
   
3. Use AudioComponentInstanceNew(ourComponent, &audioUnit) to make an instance of the audio unit
   
4. Enable IO for recording and possibly playback with AudioUnitSetProperty
   
5. Describe the audio format in an AudioStreamBasicDescription structure, and apply the format using AudioUnitSetProperty
   
6. Provide a callback for recording, and possibly playback, again using AudioUnitSetProperty
   
7. Allocate some buffers
   
8. Initialise the audio unit
   
9. Start the audio unit
   
10. Rejoice
    

Here’s my code: I’m using both recording and playback. Use what applies to you!

Initialisation

Initialisation looks like this. We have a member variable of type AudioComponentInstance which will contain our audio unit.

The audio format described below uses SInt16 for samples (i.e. signed, 16 bits per sample)

#define kOutputBus 0
#define kInputBus 1
// ...
 
OSStatus status;
AudioComponentInstance audioUnit;
 
// Describe audio component
AudioComponentDescription desc;
desc.componentType = kAudioUnitType_Output;
desc.componentSubType = kAudioUnitSubType_RemoteIO;
desc.componentFlags = 0;
desc.componentFlagsMask = 0;
desc.componentManufacturer = kAudioUnitManufacturer_Apple;
 
// Get component
AudioComponent inputComponent = AudioComponentFindNext(NULL, &desc);
 
// Get audio units
status = AudioComponentInstanceNew(inputComponent, &audioUnit);
checkStatus(status);
 
// Enable IO for recording
UInt32 flag = 1;
status = AudioUnitSetProperty(audioUnit, 
                              kAudioOutputUnitProperty_EnableIO, 
                              kAudioUnitScope_Input, 
                              kInputBus,
                              &flag, 
                              sizeof(flag));
checkStatus(status);
 
// Enable IO for playback
status = AudioUnitSetProperty(audioUnit, 
                              kAudioOutputUnitProperty_EnableIO, 
                              kAudioUnitScope_Output, 
                              kOutputBus,
                              &flag, 
                              sizeof(flag));
checkStatus(status);
 
// Describe format
audioFormat.mSampleRate = 44100.00;
audioFormat.mFormatID = kAudioFormatLinearPCM;
audioFormat.mFormatFlags = kAudioFormatFlagIsSignedInteger | kAudioFormatFlagIsPacked;
audioFormat.mFramesPerPacket = 1;
audioFormat.mChannelsPerFrame = 1;
audioFormat.mBitsPerChannel = 16;
audioFormat.mBytesPerPacket = 2;
audioFormat.mBytesPerFrame = 2;
 
// Apply format
status = AudioUnitSetProperty(audioUnit, 
                              kAudioUnitProperty_StreamFormat, 
                              kAudioUnitScope_Output, 
                              kInputBus, 
                              &audioFormat, 
                              sizeof(audioFormat));
checkStatus(status);
status = AudioUnitSetProperty(audioUnit, 
                              kAudioUnitProperty_StreamFormat, 
                              kAudioUnitScope_Input, 
                              kOutputBus, 
                              &audioFormat, 
                              sizeof(audioFormat));
checkStatus(status);
 
 
// Set input callback
AURenderCallbackStruct callbackStruct;
callbackStruct.inputProc = recordingCallback;
callbackStruct.inputProcRefCon = self;
status = AudioUnitSetProperty(audioUnit, 
                              kAudioOutputUnitProperty_SetInputCallback, 
                              kAudioUnitScope_Global, 
                              kInputBus, 
                              &callbackStruct, 
                              sizeof(callbackStruct));
checkStatus(status);
 
// Set output callback
callbackStruct.inputProc = playbackCallback;
callbackStruct.inputProcRefCon = self;
status = AudioUnitSetProperty(audioUnit, 
                              kAudioUnitProperty_SetRenderCallback, 
                              kAudioUnitScope_Global, 
                              kOutputBus,
                              &callbackStruct, 
                              sizeof(callbackStruct));
checkStatus(status);
 
// Disable buffer allocation for the recorder (optional - do this if we want to pass in our own)
flag = 0;
status = AudioUnitSetProperty(audioUnit, 
                              kAudioUnitProperty_ShouldAllocateBuffer,
                              kAudioUnitScope_Output, 
                              kInputBus,
                              &flag, 
                              sizeof(flag));
 
// TODO: Allocate our own buffers if we want
 
// Initialise
status = AudioUnitInitialize(audioUnit);
checkStatus(status);

Then, when you’re ready to start:

OSStatus status = AudioOutputUnitStart(audioUnit);checkStatus(status);

And to stop:

OSStatus status = AudioOutputUnitStop(audioUnit);checkStatus(status);

Then, when we’re finished:

AudioUnitUninitialize(audioUnit);

And now for our callbacks.

RECORDING 

static OSStatus recordingCallback(void *inRefCon, 
                                  AudioUnitRenderActionFlags *ioActionFlags, 
                                  const AudioTimeStamp *inTimeStamp, 
                                  UInt32 inBusNumber, 
                                  UInt32 inNumberFrames, 
                                  AudioBufferList *ioData) {
 
    // TODO: Use inRefCon to access our interface object to do stuff
    // Then, use inNumberFrames to figure out how much data is available, and make
    // that much space available in buffers in an AudioBufferList.
 
    AudioBufferList *bufferList; // <- Fill this up with buffers (you will want to malloc it, as it's a dynamic-length list)
 
    // Then:
    // Obtain recorded samples
 
    OSStatus status;
 
    status = AudioUnitRender([audioInterface audioUnit], 
                             ioActionFlags, 
                             inTimeStamp, 
                             inBusNumber, 
                             inNumberFrames, 
                             bufferList);
    checkStatus(status);
 
    // Now, we have the samples we just read sitting in buffers in bufferList
    DoStuffWithTheRecordedAudio(bufferList);
    return noErr;
}

PLAYBACK

static OSStatus playbackCallback(void *inRefCon, 
                                  AudioUnitRenderActionFlags *ioActionFlags, 
                                  const AudioTimeStamp *inTimeStamp, 
                                  UInt32 inBusNumber, 
                                  UInt32 inNumberFrames, 
                                  AudioBufferList *ioData) {    
    // Notes: ioData contains buffers (may be more than one!)
    // Fill them up as much as you can. Remember to set the size value in each buffer to match how
    // much data is in the buffer.
    return noErr;
}

Finally, rejoice with me in this discovery ;)

Resources that helped

• http://pastie.org/pastes/219616
• http://developer.apple.com/samplecode/CAPlayThrough/listing8.html
• http://listas.apesol.org/pipermail/svn-libsdl.org/2008-July/000797.html

No thanks at all to Apple for their lack of accessible documentation on this topic – They really have a long way to go here! Also boo to them with their lack of search engine, and refusal to open up their docs to Google. It’s a jungle out there!

Update: You can adjust the latency of RemoteIO (and, in fact, any other audio framework) by setting the kAudioSessionProperty_PreferredHardwareIOBufferDuration property:

float aBufferLength = 0.005; // In secondsAudioSessionSetProperty(kAudioSessionProperty_PreferredHardwareIOBufferDuration,                         sizeof(aBufferLength), &aBufferLength);

This adjusts the length of buffers that’re passed to you – if buffer length was originally, say, 1024 samples, then halving the number of samples halves the amount of time taken to process them.