hello! this is my first post - so please forgive and tell me about possible etiquette-violations…
i am working on a patch that involves 10channel audio output (via filestream to audioout with 5 instances in stereo). now i am doing an rms analysis for each stereo file (with 5 more filestreams playing the same material) and experience very low buffer rates, like 0.1 buffers per frame, which isn’t usefull to trigger anything.
any suggestions how i could improve that?
i reckon there might be a way to manually set the audiobuffer?
it’s rather large files (wav, 16bit, 44.1khz, up to ten minutes length sometimes. so between 20-160mb each file), could this maybe be the critical issue? but playback works really smooth, thats why i wonder… cpu- and disk-load look totally relaxed too.
hehe, thanks kalle!
so any advice for me?
i had a quick chat with tonfilm and reckon the most elegant way would be to route the signal of my multlichannel audioouts directly to the appropriate audioins and then connect the rms to this. unfortunately my soundcard (edirol fa-101) doesn’t have software to route it internally, and vvvv doesn’t seem to have capabilities for this as well…
ok, so for anybody experiencing similar problems. my workaround now is another patch in max/msp to play the sound and do the audio analysis, receiving control-cues from vvvv and sending back the analysis data, all via udp.
because all audio is running on directshow in vvvv, it becomes a bit limited and eventually creates unnecessary performance implications. but that’s actually fine, i knew that audio just isn’t vvvv’s strongest side and it doesn’t have to be.
