I'm creating a Mac OS X CoreAudio command-line program with proprietary rendering of some alphanumeric terminal input into a live audio signal, by means of AudioUnits, trying to stay as simple as possible. All works fine up to matching output sample rate.
As a starting point I'm using the Chapter 07 tutorial code of Addisson Wesley's "Learning Core Audio", CH07_AUGraphSineWave.
I initialize the AudioComponent "by the book":
void CreateAndConnectOutputUnit (MyGenerator *generator)
{
AudioComponentDescription theoutput = {0};
theoutput.componentType = kAudioUnitType_Output;
theoutput.componentSubType = kAudioUnitSubType_DefaultOutput;
theoutput.componentManufacturer = kAudioUnitManufacturer_Apple;
AudioComponent comp = AudioComponentFindNext (NULL, &theoutput);
if (comp == NULL) {
printf ("can't get output unit");
exit (-1);
}
CheckError (AudioComponentInstanceNew(comp, &generator->outputUnit),
"Couldn't open component for outputUnit");
AURenderCallbackStruct input;
input.inputProc = MyRenderProc;
input.inputProcRefCon = generator;
CheckError(AudioUnitSetProperty(generator->outputUnit,
kAudioUnitProperty_SetRenderCallback,
kAudioUnitScope_Input,
0,
&input,
sizeof(input)),
"AudioUnitSetProperty failed");
CheckError (AudioUnitInitialize(generator->outputUnit),
"Couldn't initialize output unit");
}
My main problem is in my not knowing how to retreive the output hardware sample rate for the rendering AURenderCallbackStruct since it does play a vital part in the signal generating process. I can't afford having the sample rate hard-coded into the rendering callback, although knowing it's the easiest way, since rate mismatch causes the signal being played at wrong pitch.
Is there a way of getting the default output's sample rate on such a low-level API? Is there a way of matching it somehow, without getting overly complicated? Have I missed something?
Thanks in advance.
Regards,
Tom