Humanization: random velocity and timing

[jsmonzani]

Humanization would help making a sequence more alive. I'm proposing an implementation using the modulation matrix, maybe these could be simply two sequence parameters, like SWING.

• random velocity: it could be implemented in the matrix with VEL as a destination. Use case: set RAND as source, VEL as destination and set your amount. As VEL is also a modulation source, any modulation of VEL as destination should be computed before using it as a source.
• random timing: it could be a delay before triggering the step. Proposal: add NUDGE as a new destination and use RAND as source.

Even better if multiple notes on a same step have a different random nudge.

[kurogedelic]

1.6 is out: https://github.com/hugelton/Felucca/releases/tag/v1.6

Thanks for the detailed idea! 1.6 adds two MOD matrix destinations: VEL and TIME. RAND -> VEL gives every note its own velocity, including each note of a chord, since RAND is drawn per note. RAND -> TIME plays each sequencer step a little late, up to about 30 ms at full amount; S&H or the LFO give a slower drift instead. TIME can only delay, and it moves a whole step, so the notes of a chord shift together. Per-note offsets would need memory the FM-1 doesn't have left. One caveat: on older firmware, a slot set to VEL or TIME reads as DEPTH.

[jsmonzani]

Ooh thank you so much! I'm looking forward to try it!