Unless I'm totally misunderstanding the Mix Audio Channels node, or perhaps it's not working on M1.

What I'm wanting to do is mix pre-recorded audio together. Tried all manner of messing reconstructing the audio frames using the timestamp from a single source of audio frame rate, but no luck.

Maybe just busted on M1?

The result of putting two audio channels together is something that sounds like it's literally alternating between frames of audio, which slows audio to half speed and makes all kinds of glitching.

testAudio.mp4

Comments

Keith...first off, let me

George_Toledo's picture
Submitted by

Keith...first off, let me state that I can be super dense until I wrap my head around something, so apologies for that in advance. Apologies if any of this increases the noise floor.

What is the Fire Periodically supposed to do in these graphs, when attached to the Set Time input ports? Increment the timestamp? Could it be that Set Time needs to occur at slight offset, prior to the Play event? Or does this work fine until the Mix Audio Channels is introduced?

If you make a duplicate file and load two distinct files instead of having the players share the same file loaded, does it clear up issues?

Anyway, don't feel compelled to spend time answering all of this...just giving some initial thoughts on how I might shuffle some stuff around to try to dig deeper into what the bug might be, or a temporary workaround.

Fire periodically is just

keithlang's picture
Submitted by

Fire periodically is just emulating my app, which is basically firing off audio clips. This could be any audio sources—for example a single long audio file that's triggered by Fire on Start.

Distinct files make no difference (tried that). Given that I can reproduce a Vuo crash here with a log (with this setup), I'm thinking it's something beyond my control at this point.

can confirm this behavior. it

krezrock's picture
Submitted by

can confirm this behavior. it also occurs if you send audio straight to the send live audio node.

if you add a second send live audio node. it sounds correct. but it's probably not ideal to run two of those nodes.

most unfortunate, as having isolated instruments / sources is ideal.