baconpaul
commented
3 years ago So some technical notes to myself
data structures in LFOModulationSource
bool osc_retrigger[n_oscs], lfo_retrigger[n_lfos]in surge voice scan those and call lfo->attack() for lfo or a newly defined osc->externalRetrigger() which we add to oscbase
then basically add it as an output layer in the Formula and test it
then go whack the mseg
then think about how to do step sequencer
clearly formula is the place to start as we hash out the bugs though
on and for now only string and twist would do anything in externalRetrigger (twist would hit the LPG if active; string would re-burst if in a burst mode)
mkruselj
Currently, there are
retrigger_FEGandretrigger_AEGmembers in LFOModulationSource, that allow the voice step sequencers (and, since #4201 was implemented, MSEGs) retriggering the filter and amplifier envelopes.Add an additional set of triggers (well, several vectors of triggers), that will allow triggering oscillators (currently, String and certain modes of Twist would require this, as per #4370), and also retriggering the LFO EGs.
UI-wise, there's lots to discuss, but for the time being we could have three-part instead of two-part split in the step sequencer trigger row:
This poses an issue where the default left click would enable all triggers, where in most cases we'd want just FEG and AEG to trigger simultaneously. Which leads us to the likely scenario that we would need to break off the triggers in step seq into separate lanes (so 3 lanes total), in order to free up the right click, so that we can use it in the additional trigger lane to pick the destination for the trigger (osc 1/2/3, any of the voice LFO EGs).
In MSEG and Formula this is rather easy - for MSEG it's in the RMB menu on the node, but we should also figure out a visual way to edit it.
This issue essentially subsumes #4370 so gonna close that one. It's also likely a much simpler way of gaining more flexibility regarding triggering stuff, compared to proposal in #4366.