I don't get your point, do you mean that audio processing is irrelevant since an assembly language low pass filter is spending most time in read/write and not processing?
Originally Posted by gens
I might have misunderstood, but I understand if you dislike hardware audio "processing" because they do things you shouldn't do to audio streams, but that's not really what true audio and audio processing in games is about..
The thing is that for one audio stream source, the best is to do *nothing* to it, but in a game, where you have up to several hundred audio streams that are triggered simultaneously, getting this represented correctly with respect to the player position and direction in a surround perspective, you kind of have to. And this probably requires quite a bit more processing time.
So I think there is definitely benefits to do this in hardware.