Switching between poly/mono changes the character of the voice

[riban-bw]

Switching from polyphonic to monophonic mode changes how a voice sounds, which seems to be unexpected behaviour.

• Select Basic Saw preset.
• Change "Assign mode" from "Poly" to "Mono".

The sound gets louder and harsher (possibly due to clipping distortion).

I would not expect the character or volume of a patch to change when switching to monophonic.

I have v 2.3.1z running on a Zynthian V4 (RPi4). Same behaviour on a Zynthian custom build with RPi5.

[polluxsynth]

The key here is that when switching to Mono, the Voice Count parameter essentially becomes a unison count, governing how many voices are played in unison. I had a look at the Basic Saw patch, and Oscillator Spread parameter is set to 0, so there will be no detune among the voices, however, the phase relationships between the oscillators will be whatever it was when switching over to Mono. This probably accounts for the apparent 'harsh' sound, as in practice it's rather random what the phase relationships are between the oscillators, so you'll very likely end up with something that's not just a simple sawtooth.

If you set the Voice Count to 1, you'll get an ordinary monophonic voice, and that should sound ok. Alternatively, if you advance the Oscillator Spread parameter you'll get a nice slow detune between the voices, attempting to emulate a properly tuned analog synthesizer.