1
Notes on status of CoreAudio Implementation of PortAudio
3
Document Last Updated December 9, 2005
5
There are currently two implementations of PortAudio for Mac Core Audio.
7
The original is in pa_mac_core_old.c, and the newer, default implementation
9
Only pa_mac_core.c is currently developed and supported as it uses apple's
10
current core audio technology. To select use the old implementation, replace
11
pa_mac_core.c with pa_mac_core_old.c (eg. "cp pa_mac_core_auhal.c
12
pa_mac_core.c"), then run configure and make as usual.
14
-------------------------------------------
16
Notes on Newer/Default AUHAL implementation:
20
Last Updated December 9, 2005
22
Principle of Operation:
24
This implementation uses AUHAL for audio I/O. To some extent, it also
25
operates at the "HAL" Layer, though this behavior can be limited by
26
platform specific flags (see pa_mac_core.h for details). The default
27
settings should be reasonable: they don't change the SR of the device and
28
don't cause interruptions if other devices are using the device.
30
Major Software Elements Used: Apple's HAL AUs provide output SR
31
conversion transparently, however, only on output, so this
32
implementation uses AudioConverters to convert the sample rate on input.
33
A PortAudio ring buffer is used to buffer input when sample rate
34
conversion is required or when separate audio units are used for duplex
35
IO. Finally, a PortAudio buffer processor is used to convert formats and
36
provide additional buffers if needed. Internally, interleaved floating
37
point data streams are used exclusively - the audio unit converts from
38
the audio hardware's native format to interleaved float PCM and
39
PortAudio's Buffer processor is used for conversion to user formats.
41
Simplex Input: Simplex input uses a single callback. If sample rate
42
conversion is required, a ring buffer and AudioConverter are used as
45
Simplex output: Simplex output uses a single callback. No ring buffer or
46
audio converter is used because AUHAL does its own output SR conversion.
48
Duplex, one device (no SR conversion): When one device is used, a single
49
callback is used. This achieves very low latency.
51
Duplex, separate devices or SR conversion: When SR conversion is
52
required, data must be buffered before it is converted and data is not
53
always available at the same times on input and output, so SR conversion
54
requires the same treatment as separate devices. The input callback
55
reads data and puts it in the ring buffer. The output callback reads the
56
data off the ring buffer, into an audio converter and finally to the
59
Platform Specific Options:
61
By using the flags in pa_mac_core.h, the user may specify several options.
62
For example, the user can specify the sample-rate conversion quality, and
63
the extent to which PA will attempt to "play nice" and to what extent it
64
will interrupt other apps to improve performance. For example, if 44100 Hz
65
sample rate is requested but the device is set at 48000 Hz, PA can either
66
change the device for optimal playback ("Pro" mode), which may interrupt
67
other programs playing back audio, or simple use a sample-rate coversion,
68
which allows for friendlier sharing of the device ("Play Nice" mode).
70
Additionally, the user may define a "channel mapping" by calling
71
paSetupMacCoreChannelMap() on their stream info structure before opening
72
the stream with it. See below for creating a channel map.
76
- Buffering: No buffering beyond that provided by core audio is provided
77
except where absolutely needed for the implementation to work. This may cause
78
issues with large framesPerBuffer settings and it also means that no additional
79
latency will be provided even if a large latency setting is selected.
81
- Latency: Latency settings are generally ignored. They may be used as a
82
hint for buffer size in paHostFramesPerBufferUnspecified, or the value may
83
be used in cases where additional buffering is needed, such as doing input and
84
output on seperate devices. Latency settings are always automatically bound
85
to "safe" values, however, so setting extreme values here should not be
88
- Buffer Size: paHostFramesPerBufferUnspecified and specific host buffer sizes
89
are supported. paHostFramesPerBufferUnspecified works best in "pro" mode,
90
where the buffer size and sample rate of the audio device is most likely
91
to match the expected values. In the case of paHostFramesPerBuffer, an
92
appropriate framesPerBuffer value will be used that guarantees minimum
93
requested latency if that's possible.
95
- Timing info. It reports on stream time, but I'm probably doing something
96
wrong since patest_sine_time often reports negative latency numbers. Also,
97
there are currently issues with some devices whehn plugging/unplugging
100
- xrun detection: The only xrun detection performed is when reading
101
and writing the ring buffer. There is probably more that can be done.
103
- abort/stop issues: stopping a stream is always a complete operation,
104
but latency should be low enough to make the lack of a separate abort
105
unnecessary. Apple clarifies its AudioOutputUnitStop() call here:
106
http://lists.apple.com/archives/coreaudio-api/2005/Dec/msg00055.html
108
- blocking interface: should work fine.
110
- multichannel: It has been tested successfully on multichannel hardware
111
from MOTU: traveler and 896HD. Also Presonus firepod and others. It is
112
believed to work with all Core Audio devices, including virtual devices
115
- sample rate conversion quality: By default, SR conversion is the maximum
116
available. This can be tweaked using flags pa_mac_core.h. Note that the AU
117
render quyality property is used to set the sample rate conversion quality
118
as "documented" here:
119
http://lists.apple.com/archives/coreaudio-api/2004/Jan/msg00141.html
121
- x86/Universal Binary: Universal binaries can be build.
125
Creating a channel map:
127
How to create the map array - Text taken From AUHAL.rtfd :
129
Clients can tell the AUHAL units which channels of the device they are interested in. For example, the client may be processing stereo data, but outputting to a six-channel device. This is done by using the kAudioOutputUnitProperty_ChannelMap property. To use this property:
132
Create an array of SInt32 that is the size of the number of channels of the device (Get the Format of the AUHAL's output Element == 0)
133
Initialize each of the array's values to -1 (-1 indicates that that channel is NOT to be presented in the conversion.)
135
Next, for each channel of your app's output, set:
136
channelMapArray[deviceOutputChannel] = desiredAppOutputChannel.
138
For example: we have a 6 channel output device and our application has a stereo source it wants to provide to the device. Suppose we want that stereo source to go to the 3rd and 4th channels of the device. The channel map would look like this: { -1, -1, 0, 1, -1, -1 }
140
Where the formats are:
141
Input Element == 0: 2 channels (- client format - settable)
142
Output Element == 0: 6 channels (- device format - NOT settable)
144
So channel 2 (zero-based) of the device will take the first channel of output and channel 3 will take the second channel of output. (This translates to the 3rd and 4th plugs of the 6 output plugs of the device of course!)
147
Create an array of SInt32 that is the size of the number of channels of the format you require for input. Get (or Set in this case as needed) the AUHAL's output Element == 1.
149
Next, for each channel of input you require, set:
150
channelMapArray[desiredAppInputChannel] = deviceOutputChannel;
152
For example: we have a 6 channel input device from which we wish to receive stereo input from the 3rd and 4th channels. The channel map looks like this: { 2, 3 }
154
Where the formats are:
155
Input Element == 0: 2 channels (- device format - NOT settable)
156
Output Element == 0: 6 channels (- client format - settable)
160
----------------------------------------
162
Notes on Original implementation:
164
by Phil Burk and Darren Gibbs
166
Last updated March 20, 2002
170
Output with very low latency, <10 msec.
171
Half duplex input or output.
172
Full duplex on the same CoreAudio device.
173
The paFLoat32, paInt16, paInt8, paUInt8 sample formats.
177
KNOWN BUGS OR LIMITATIONS
179
We do not yet support simultaneous input and output on different
180
devices. Note that some CoreAudio devices like the Roland UH30 look
181
like one device but are actually two different CoreAudio devices. The
182
Built-In audio is typically one CoreAudio device.
188
CoreAudio devices can support both input and output. But the sample
189
rates supported may be different. So we have map one or two PortAudio
190
device to each CoreAudio device depending on whether it supports
191
input, output or both.
193
When we query devices, we first get a list of CoreAudio devices. Then
194
we scan the list and add a PortAudio device for each CoreAudio device
195
that supports input. Then we make a scan for output devices.