Multiple Sound Output Channels
The Sound Manager has supported this starting with Sound Manager 3.0
which shipped with System 7.0. Inside Macintosh: Sound page 2-128 mentions
this feature but doesn't describe how it is used.
When you call SndNewChannel one of the parameters passed is a
long, which
specifies the initialization parameters for the channel. Normally, developers
just pass nil, or initMono or initStereo, but you can
also pass
kUseOptionalOutputDevice (which is defined to be -1).
The use of kUseOptionalOutputDevice will allow you to specify a
different
output component to have the sound played through. With kUseOptionalOutputDevice,
a developer is able to play sounds simultaneously on as many output devices as
are installed. For instance, a hardware developer who has a six-channel card
could make three stereo output components (one for each pair of output channels)
and the kUseOptionalOutputDevice selector would allow developers writing software
to this card to play three stereo sounds simultaneously.
The kUseOptionalOutputDevice selector works like this:
First, you have to find the component instance of the output component
you wish to use (this example finds the AIFF Writer sample output component):
//Find our output component's instance
outputDev.componentType = 'sdev';
outputDev.componentSubType = 'AIFW';
outputDev.componentManufacturer = 'appl';
outputDev.componentFlags = 0;
outputDev.componentFlagsMask = 0;
theAIFWComponent = FindNextComponent (0, &outputDev);
|
A generic routine to find all sound output components such as:
long FindAllsdevs (Component ** componentsArray) {
Component foundComponent,
aComponent;
ComponentDescription looking;
long numComponents,
i;
aComponent = 0;
looking.componentType = kSoundOutputDeviceType; // 'sdev'
looking.componentSubType = 0;
looking.componentManufacturer = 0;
looking.componentFlags = 0;
looking.componentFlagsMask = 0;
numComponents = CountComponents (&looking);
*componentsArray = (Component*)NewPtr (sizeof (Component) * numComponents);
if (componentsArray == nil) {
numComponents = 0; // won't be able to list them anyway
}
for (i = 0; i < numComponents; i++) {
foundComponent = FindNextComponent (aComponent, &looking);
(*componentsArray)[i] = foundComponent;
aComponent = foundComponent; // continue looking
}
return numComponents;
}
|
would be used to present the user with a list of sound output channels so that
they could output multi-channel (greater than 2 channel) sounds without the sound
having to be mixed down to two channels.
Now, all that is left to do is to make a new sound channel which will use the
selected ouput device:
err = SndNewChannel (&theOptionalOutputChan, kUseOptionalOutputDevice, (long)theAIFWComponent, nil);
|
This is all there is to it. Whenever a sound is played through theOptionalOutputChan
the sound will not go to the built-in hardware: instead, it will go to the AIFF Writer
output device (which could just as easily be the third and fourth channels of a multiple-channel
output card).
So, what does a developer of multiple-channel hardware have to do to allow
the use of the theOptionalOutputChan selector?
Not much. All that has to be done is to make an output component for each pair of
output channels the hardware supports. If your hardware has only two channels, one
output component is all that is needed. If your hardware has 20 output channels,
10 output components would be required.
I can almost hear developers now: "Ten output components?!? Are you crazy?!?"
No. Because of the reusability inherent to components, all that is really
required is ten 'thng' resources that register the same output component ten
times as if it was ten different output components.
The output component should be written in such a way that the actual hardware
channels it outputs to are abstracted: that way, one code base can talk to all
available channels. The output component just looks at itself at register time
to determine which output channels it controls and saves this information in its globals.
The additional code required is very minor: just enough code to keep track of which
channels to output on.
|
Note:
You can get the Speech Manager to talk through an optional output
channel as well. Call NewSpeechChannel as you normally would and then call
SetSpeechInfo with the soSoundOutput selector and the component
instance of
the output component: |
|  (Disk 1).iso/pc/technical documentation/macintosh technotes and q&as/technotes/tn/blueline.gif){kind=small}
 (Disk 1).iso/pc/technical documentation/macintosh technotes and q&as/technotes/tn/inputsource.gif)
 (Disk 1).iso/pc/technical documentation/macintosh technotes and q&as/technotes/tn/acrobatsmall.gif){kind=small}
Acrobat version of this Note