Thank you all for you feedback. I’m gonna try to reply to stuff, but I’m really new to these forums so things might go silly.
I actually intentionally made the decision against using the sample rate node since I wanted the same knob position to have the same effect, regardless of the sample rate that the user has their document running at. But I realize that having an “effective sample rate” over the nyquist rate will really break things, so I should probably bring the max down to 20kHz or around there.
I want to thank you and @J301 for pointing this out, but since I only run the iOS version (for now. I can’t wait for A4 and to get it on multiple platforms), I didn’t know there was a better control. I’ll try swapping it out to make the patch easier to control.
This was me having a brain fart and thinking a phasor went from 0…1 instead of 0…2pi. I tried having an expr node with x>0.5 as the condition, but for some strange reason it didn’t work right. Removing the unit delay would absolutely make this better.
I chose to go with 16 bits as a “max” level since it is used (with integer, not float) samples for most commonly shared sound. And as to fractional bits, I’m trying to use them in the more information-theoretic context where “3.7 bits” implies that there are 2^3.7, or almost 13 distinct values for the output. I’m off by almost a bit in my calculation since I have all of my math being done as zero plus choices being either above or below zero (so with “3” bits, there would be (2^4)-1 possible outputs.
Thank you all for the feedback, I’m going to try to improve this patch over the next few days and want to post an updated version of this.