Odd, I thought, the signal was clean when bypass was enabled - I’ll give it another go, perhaps I need to test with a cleaner sound source, and perhaps look at it in a scope.
I’ll also compare it to what I’m getting with Axoloti.
One small issue, as I don’t have a ‘real clouds’ , I don’t have any benchmark for the wet signal - but I assume bypass is fully clean, so perhaps if we get that then wet will improve too.
EDIT: I think I’ve an idea where the issue might lie.
first its important to know, clouds is limited to 16bit resolution, so you will always get some bit reduction. (from 32->16)
however, I think during this reduction, I’ve allowed for too much headroom, at the expense of resolution.
this afternoon, I’m going to try reducing this significantly. this should improve the sound considerably, but will mean that controlling the input gain will be even more important… I’ll probably add a input gain control, to the patch to help this.
(interestingly, its similar in Axoloti, and no one has mentioned it… so if this helps I can improve in in Axoloti too )
EDIT: ok, tested… doesnt appear to be distortion (at least not with bypass/mix=0)
but Ive (finally) now found a bigger issue, which was causing big issues (due to organelle audio buffers being much larger than MI/axoloti), so Im testing that… the next release of clds will be much better