Per voice paraphonic sequencing

[MickH]

Is it possible to sequence each of the different voices independently? I'm trying to work it out but the manual only seems to suggest that a single track controlling all voices simultaneously is possible.

What I thought would be really exciting to do would be to create independent tracks for each voice of different lengths and thereby create constantly shifting three track polymetric sequences...

[creativespiral]

Hey Mick -

Yes, you can sequencer Per Oscillator.

Just put the synth into Paraphonic Mode, then in Sequence Menu, choose Sequencer Params > Paraphonic ON.   It will change the first three lanes from Note, Duration, Ratcheting  into  Osc1Note, Osc2Note, Osc3Note.   

You can assign holds and rests.  Also, you can change the end step to be different than the last step (16), by holding down the step that you want to be last and hitting the RESET button in the seq area...  to create more complex poly-rhythms. 

[MickH]

Oh man, thanks a bunch for your reply, couldn't work that out there. Great for making some spacey loops!

[LoboLives]

I still think it's slightly a shame that each oscillator can't have it's own envelope. It would essentially make paraphonic sequencing multitimbral meaning you could have three totally different sequences going.

[creativespiral]

@LoboLIves - That would be cool, but even if the envelopes are digitally generated (not sure if they are on this board?),  that would mean you'd need 3x the VCF circuitry and/or 3x the VCAs to articulate the notes...  would probably add significant electronics cost and design complexity.   

The Pro3 is definitely a mono-synth at heart, but the paraphonic capabilities are really quite good / well implemented... with the expected limitations of being "paraphonic" and not poly.    You can achieve some pretty cool sounding paraphonic pads/strings/plucks/etc  and do some complex sequencing in paraphonic mode. 

There is one request that I would love to see implemented as a feature in a future OS:   Make it so the arpeggiator can be used in conjunction with the sequencer in "Trigger-mode" (aka: key step mode).    That would make it easier to create complex Arp Sequences that can be played up and down the keybed in any key/mode.

[LoboLives]

@LoboLIves - That would be cool, but even if the envelopes are digitally generated (not sure if they are on this board?),  that would mean you'd need 3x the VCF circuitry and/or 3x the VCAs to articulate the notes...  would probably add significant electronics cost and design complexity.   

The Pro3 is definitely a mono-synth at heart, but the paraphonic capabilities are really quite good / well implemented... with the expected limitations of being "paraphonic" and not poly.    You can achieve some pretty cool sounding paraphonic pads/strings/plucks/etc  and do some complex sequencing in paraphonic mode. 

There is one request that I would love to see implemented as a feature in a future OS:   Make it so the arpeggiator can be used in conjunction with the sequencer in "Trigger-mode" (aka: key step mode).    That would make it easier to create complex Arp Sequences that can be played up and down the keybed in any key/mode.

I hear ya. I know the Pro 3 isn't built for that request. I just think with three filters and three oscillators I was just hoping with the announcement of the Pro 3 we'd get a 3 part multitimbral synth. I guess it would be too complicated. 

[sockmonkey]

I still think it's slightly a shame that each oscillator can't have it's own envelope. It would essentially make paraphonic sequencing multitimbral meaning you could have three totally different sequences going.

I just made a patch the other day which sort of does this. The limitation I encountered was that there's no way to decouple the VCA envelope from the VCA. Ideally, you could just open the VCA (mod DC -> mythical "VCA level" destination) and then use the VCA envelope and the 2 aux envelopes for the 3 oscillators. I ended up using the filter envelope for OSC 1 instead, but it's a waste of an envelope in this case. In any case, it wasn't a very useful paraphonic patch, but I did have 3 distinct timbre/envelope combinations responding to incoming notes.

[Kosmikos]

And how do you trigger the correct envelope in this case?

[sockmonkey]

And how do you trigger the correct envelope in this case?

I don't. All 3 envelopes are triggered, but in paraphonic mode, only one oscillator is sent through the mixer to the VCA. I guess there's a paraphonic pre-mixer of sorts in the overall signal flow (with the VCA envelope hard-wired to the osc levels). In a perfect world, there would be modulation sources for "Paraphonic Trigger 1-3" so that you could scale other values depending on which voice is sounding. And/or maybe "Paraphonic Level 1-3" as a destination? One can dream.
 
In any case, I was able to achieve the effect I was going for, more or less. A little clunky, but approximately doable.

[Kosmikos]

I still need to wrap my head around this tbh. I’ve had trouble with the velocities in paraphonic sequencing, it seems to affect all voices, but since you can set them on 3 tracks the result ends up a bit random when two notes or more are being triggered on the same step.
It almost seems like the sequencer has been designed for 3 separate envelopes, but the rest of the synth hasn’t.

[sockmonkey]

I still need to wrap my head around this tbh. I’ve had trouble with the velocities in paraphonic sequencing, it seems to affect all voices, but since you can set them on 3 tracks the result ends up a bit random when two notes or more are being triggered on the same step.
It almost seems like the sequencer has been designed for 3 separate envelopes, but the rest of the synth hasn’t.

Once I get back to my Pro 3 (probably tomorrow), I can post the sysex for the patch. I don't know that it's all that helpful, or if it's "best practice", but it's what I could figure out in the time available.

[MickH]

Can't you assign a separate envelope to each oscillator volume into the amplifier? It's not perfect but will go some way towards having unique per voice envelope response

[creativespiral]

I just made a patch the other day which sort of does this. The limitation I encountered was that there's no way to decouple the VCA envelope from the VCA. Ideally, you could just open the VCA (mod DC -> mythical "VCA level" destination) and then use the VCA envelope and the 2 aux envelopes for the 3 oscillators. I ended up using the filter envelope for OSC 1 instead, but it's a waste of an envelope in this case. In any case, it wasn't a very useful paraphonic patch, but I did have 3 distinct timbre/envelope combinations responding to incoming notes.

I don't. All 3 envelopes are triggered, but in paraphonic mode, only one oscillator is sent through the mixer to the VCA. I guess there's a paraphonic pre-mixer of sorts in the overall signal flow (with the VCA envelope hard-wired to the osc levels). In a perfect world, there would be modulation sources for "Paraphonic Trigger 1-3" so that you could scale other values depending on which voice is sounding. And/or maybe "Paraphonic Level 1-3" as a destination? One can dream.
In any case, I was able to achieve the effect I was going for, more or less. A little clunky, but approximately doable.

@sockmonkey - this is an interesting tactic!  Yeah, you can target Osc1Level, Osc2Level, Osc3Level as mod destinations.   (technically each of the oscillators does have a sort of VCA circuit used in the mixer section, pre filter... separate from the main VCA circuit)   

I see what you are saying about "decoupling" the VCA envelope from the main VCA level... what is interesting is that there already is a separate "VCA level" control in the Amp Env soft menu... on other DSI/Seq instruments, I believe that control was used just to allow the VCA to operate in a sort of "drone mode"... being always on, separate from the Amp Env... Poly Evolver for instance.   I'm actually not sure what the point of having it is, if not for that use, as the Amp Env Amount provides redundant control over the max level.   

Perhaps this is something that could be a feature request, to allow the Amp Env to be decoupled from the hard-assigned main VCA, and instead just use that VCA Level control to turn the main VCA into a drone/always on mode, or at some assigned level.   Then you could use all three oscillators with
Aux1 -> Osc1
Aux2 -> Osc2
AmpEnv -> Osc3 

As-is currently, using this method, you can get decent control of two oscillators with separate "faux VCA env control" - set the Amp Env fully open / 100% sustain, and use the two Aux Envelopes to control two Osc Levels.   What is happening is that the Main VCA still controls / allows all three oscillator levels through (after they have been mixed down and filtered), but you just apply the two Aux envelopes at the mix stage (pre-filter).   

It's a different type of sound, since the Faux Amp Envelope Shaping is happening pre-filter, but could be useful. 

I'm gonna play around with this some more... I think it might work well for drum sequencing or drum / bass line.   Maybe one osc for kick/snare, one for hi-hats, and a bass line with the fully open / non articulated one.
 

[sockmonkey]

Can't you assign a separate envelope to each oscillator volume into the amplifier? It's not perfect but will go some way towards having unique per voice envelope response

That's basically what I was describing above, and it kinda-sorta works.

[Kosmikos]

As all envelopes still gets re-triggered at the same time, this might work fine for sequencing short sounds and could give an interesting feel like compression to longer ones.

[Kosmikos]

I’ve been giving it another try, and it appears that each Osc has its own envelope in paraphonic mode after all, so there should be no need to re-assign envelopes to the osc levels (I actually didn’t get anything good with this anyway).
This seems to combine to the main amp VCA envelope and uses the same values.
This can be heard by setting the amp envelope to something like A:90 D:0 S:80 R:90. Then play/release a note, wait to reach half of the release, and play it again. You then can ear the double decay.
Also with this ADSR setting you can ear that a second note played while a first is already playing will not be as loud as the same note on its own.
Velocity seems to apply only on the main VCA envelope, so it really screws up the paraphonic output if activated. I’m really not sure why each paraphonic track has its own velocity.

[creativespiral]

I just did a deep dive into this and found several new things in the paraphonic sequencer that I didn't notice before...  and maybe found a bug / unintended behavior:

Switch to Paraphonic Mode, and Seq Mode to Paraphonic: ON. 

This changes the behavior of [Track Select] + [1] [2] [3]...  instead of being the three pages of a Single-Mode sequencer, that key-combo now brings you to the First Page of the Osc1, Osc2, Osc3 note+velocity page.   But you can actually access the "sub menus / secondary pages" of all three of them by using the soft menu (.0 note/velocity, .1 duration, .2 ratcheting).  (I never tried using soft menu to dive into extra pages in paraphonic mode ie: 1.1, 1.2, 2.1, 2.2, and 3.1, 3.2)   So, while the behavior of selecting duration/ratcheting is different, it appears there is still "multiple tracks of data" for each of the three oscillators. 

I'm wondering if this is an intended capability?  Maybe @pym or @cmdb can chime.   I just assumed it was not the case since I figured that each of those first three tracks were just being converted to a different usage in paraphonic mode (and wasn't conceptualizing how that would work with a single VCA)... but maybe those first three tracks can hold all the data of the 3 variables?   Actually 4 variables if you consider notenum/velocity as their own.   NoteNum, Velocity, Duration, Ratcheting for each of three oscillators?

Most likely, the data is being applied to the pre-filter mixing stage?.. combining the velocity/duration/ratcheting data into pulses/curves that are applied directly to the Osc mix levels?  That's actually a really cool capability, if that is the intended behavior. 

It does seem this functionality partially works, with some quirks and maybe a bug... but it could present some really interesting possibilities.   Here's a few things I've noticed:

1. All three tracks/oscillators do have a ratcheting control page (1.3, 2.3, 3.3), which is individually respected, per osc... very cool!   This definitely makes me think a pulse/curve is being dynamically generated and sent to the pre-filter mixer vca for each osc (ie: osc volume)  Taking velocity, duration and ratchets and combining them into equivalent pwm pulses.   

2. For the ratcheting to work, the duration needs to be set low... for two ratchets, the duration needs to be 50% or less... For four ratchets, it needs to be 25% or less...  (actually, the thresholds are slightly higher than those percentages listed for some reason, but those are safe numbers to use... divide 1/Ratchets)

3. Possible bugs:
a. Track 1.1 seems to control the duration of all three of them.
b. Track 2.0 and 3.0 Velocity seems to be tied together sharing the same variable, and controls the velocity of all three.


The other method discussed above of decoupling the VCA envelope from the main VCA level (as a feature request) would actually be a totally different approach to accomplishing similar behavior of paraphonic sequencing (without ratcheting, but with high resolution ADSR envelope curves per oscillator level, before the filter stage)   

[sockmonkey]

I still need to wrap my head around this tbh. I’ve had trouble with the velocities in paraphonic sequencing, it seems to affect all voices, but since you can set them on 3 tracks the result ends up a bit random when two notes or more are being triggered on the same step.
It almost seems like the sequencer has been designed for 3 separate envelopes, but the rest of the synth hasn’t.

Once I get back to my Pro 3 (probably tomorrow), I can post the sysex for the patch. I don't know that it's all that helpful, or if it's "best practice", but it's what I could figure out in the time available.

Here's that sysex. I don't claim that it's great voice programming, but it's a reasonable illustration of what I was talking about. @creativespiral has done a deep dive here, in particular with the paraphonic sequencer (nice finds!), but if you turn on paraphonic mode with the attached patch, you can hear the individual effects of the 3 mixer envelopes.

[blewis]

Kind of shows us how clever the per oscillator envelopes are on the Pro 2. When in paraphonic mode, there are separate envelopes calculated for each oscillator. Easier to do since they are all digital, but a cool implementation. I seem to remember it was a last second addition for Pro 2.

[Pym]

Yes, each oscillator has its own envelope

I'll look into the paraphonic velocity

I’ve been giving it another try, and it appears that each Osc has its own envelope in paraphonic mode after all, so there should be no need to re-assign envelopes to the osc levels (I actually didn’t get anything good with this anyway).
This seems to combine to the main amp VCA envelope and uses the same values.
This can be heard by setting the amp envelope to something like A:90 D:0 S:80 R:90. Then play/release a note, wait to reach half of the release, and play it again. You then can ear the double decay.
Also with this ADSR setting you can ear that a second note played while a first is already playing will not be as loud as the same note on its own.
Velocity seems to apply only on the main VCA envelope, so it really screws up the paraphonic output if activated. I’m really not sure why each paraphonic track has its own velocity.