I thought this was straightforward like in the audio domain? [1] Now, not all noises of the same amplitude spectrum are created equal. For instance LFSR sequen

I thought this was straightforward like in the audio domain? [1] Now, not all noises of the same amplitude spectrum are created equal. For instance LFSR sequences are only ±1 yet they have a white spectrum. The difference lies in phase. I wonder if generating random phases, then putting it through an optimizer to find lowest crest factor could work. Oh, and can someone chime in whether void-and-cluster masks guarantee that all values are covered, just like in Bayer mask? [1] https://zynaddsubfx.sourceforge.io/doc/PADsynth/PADsynth.htm https://zynaddsubfx.sourceforge.io/doc/PADsynth/PADsynth.htm