I am working on a quite complex step sequencer that triggers both sounds and visual effects. Since I want my live sessions to be available for further editing I am used to record the audio with [writesf~].
Sequencing the audio works fine at first, but right when the video kicks in, it kind of slows down bringing along some little crackling noises. Here is the weird thing: even if the live session is distorted and slowed down, the audio file I record is perfect.
This confuses me a lot. How can it record data to file in such a perfect shape when the live audio is so currupted? It doesn't seem to be related to cpu usage, am I missing something?