Blog

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. 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.

Drop me a line if you find this helpful.

TPAudioController iOS engine now available

Icon medium

If you’d rather not worry about this stuff, I’ve released TPAudioController, the iOS audio engine library that takes all the work out of creating an audio app.

TPAudioController is a flexible, easy-to-use and complete audio engine for iOS, built upon Core Audio. It handles all setup and management of the low-level Remote IO audio unit system, with support for capturing input via the microphone, automatic mixing of multiple audio signals with per-channel volume and pan controls, and capturing audio system output for session recording.

Read more, and download the sample app or evaluation framework here.

Update: 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: 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.

Update: Tom Zicarelli has created a very extensive sample app that demonstrates the use of AUGraph, with all sorts of goodies.

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)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107

#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:

1
2

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

And to stop:

1
2

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

Then, when we’re finished:

1

AudioUnitUninitialize(audioUnit);

And now for our callbacks.

Recording

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

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

1
2
3
4
5
6
7
8
9
10
11

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

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 seconds
AudioSessionSetProperty(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.

Related posts

Tagged , , , . Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.

192 Comments

  1. Michael
    Posted January 1, 2012 at 3:50 pm | Permalink

    Hi Michael.

    after i have put together all this code , how would i get the actual audio data ? where exactly does it saved ?

    and when all this code is done ? i have to put it all into 1 method and then call it ? or should i only call -start method ?

    what operation should be taken to get the real time data ?

    i have spent days to understand it but i couldnt . how exactly i

    Reply
    • Michael Tyson
      Posted January 8, 2012 at 12:02 pm | Permalink

      Hi, Michael – it sounds like another tutorial might be in order. Please stay tuned, I’ll put one together over the next week or two and post it on the blog.

      Reply
1 … 3 4 5

3 Trackbacks

  1. By [Time code]; on January 17, 2009 at 7:12 am

    [...] complexity and move on to Audio Toolbox (or perhaps even Core Audio… a DevForums thread and a blog by developer Michael Tyson report extremely low latency by using the RemoteIO audio unit [...]

  2. By iPhone Core Audio Part 3 – Audio Callback « Tim Bolstad on March 18, 2010 at 3:35 pm

    [...] I also wouldn’t have gotten anywhere on VocaForm without Michael Tyson’s post on using the remoteIO AU. [...]

  3. By Decibel metering from an iPhone audio unit | Politepix on June 18, 2010 at 11:05 am

    [...] where you can do a lot of useful stuff, and then at the lowest level there are two types of Audio Unit: Remote I/O (or remoteio) and the Voice Processing Audio Unit [...]

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*

You may use Markdown (surround code in `back-ticks`), or these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong> <pre lang="" line="" escaped="" highlight="">

Subscribe without commenting