ofSoundStream: different device for input and output

Hi, Is it possible to set different sound device for input and output?
I’m using Mac OSX but it would be nice if there’s a cross-platform solution.
Thank you in advance!

1 Like

yes you need to use 2 ofSoundStrem, one per device and set one of them to only input and the other to only output


I would suggest you to use a bind of either Ableton live, of if you want open source stuff pure data or supercollider, to handle sound within open frameworks. there are a couple of available addons. I wasn’t able to get the Ableton one set and running, but I was able to do so with pd and supercollider. and they are stable, they opensouce, and specially supercollider can be a really nice tool, to get routing throughout complex output routes. it’s not by chance that most wave field synthesis systems use supercollider

1 Like

Thanks for your answer @arturo but I don’t understand…
Do you mean something like this?

ofSoundStream inputStream, outputStream;


I tried several things but I couldn’t get it to work.
Could you explain a bit more please?

Thanks for your advice!
I am actually using pure data with OF (ofxPd)
And I couldn’t figure out setting the different sound device for input and output.

yeah something like that, it depends on the sñpecifics of your setup but:


try ofxSupercollier :slight_smile:
but the Arturo answer might be as well what you are looking for :slight_smile:

with pd I guess if you declare for instance dac~ 3 4 and bind the outputs of pd to openframeworks, that I may perfectly work

It worked perfectly.
Thank you so much @arturo :slight_smile:

hey @cuinjune and others…

I set this two soundStreams, one for input and one for the output. Both are working fine.
Input: to feed the input buffer to my “fft analyser”
Output: running a noise generator taken from the OF\examples\sound\audioOutputExample

void ofApp::audioOut(ofSoundBuffer & buffer){
        // ---------------------- noise --------------
	for (size_t i = 0; i < buffer.getNumFrames(); i++){
		lAudio[i] = buffer[i*buffer.getNumChannels()    ] = ofRandom(0, 1);
		rAudio[i] = buffer[i*buffer.getNumChannels() + 1] = ofRandom(0, 1);

and how can you “patch” the input to output?
(to listen the incoming sound signal, not the generated noise)

I want to replace this noise generator, and allow me to listen to the input sound. any idea?
I use the live input to process some FFT and I want to send the “signal” to the out soundStream.

I am using Windows10/VS2017.

I think you just take a copy of what is passed to audioIn, then copy that to the buffer passed to audioOut.

You probably want to protect the buffer you’re using to copy the data with a mutex so audioIn/Out aren’t trying to read/write it at the same time.

1 Like