Zebra 3.0 - out now

[Andreya_Autumn]

We don't have Morph between 16 different curves in our MSEGs.

[exmatproton]

I don't think it's necessary for all curves to be the same length in this new mode. We don't force that in Surge... It's simply, MSEG timeline units match your DAW's timeline (depending on baseline MSEG rate of course).

Indeed. As long as the endpoint is fixed, there shouldn't be an issue with other points in between the start and endpoint

[Andreya_Autumn]

Say you have one curve with a 4 beat loop and another with a 3 beat loop. You put the morph in the middle. What should happen? As it stands right now, the loop becomes 3.5 beats long and the shape morphs between the two.

You could just slam the morph over instantly, avoiding the probably undesirable in-between state. But the point is, presently, the phase of the second loop with respect to the timeline *depends on when you did the morph*. Consequently, if Urs et al add a timeline sync mode without doing what I said above, it still wouldn't avoid the "whence-do-you-begin-playback" issues. Because the correct state of the MSEG would depend on prior playback and automation.

Locking the loop points lock to the start/end and forcing all the curves to same length fixes that.

[Andreya_Autumn]

There is one other possibility, but it requires a completely different logic for the curve morph. If each curve has its own playhead which runs constantly, and morphing crossfade between the outputs of the different curves, that would work too. So morphing from the 4-beat to the 3-beat loop would then give you a 12-beat pattern as the full 4:3 polymeter plays out. That would invalidate the morph points thing though. I doubt U-He would want to go that path. Locking the lengths to be the same seems like a more in-universe strategy to me.

[Urs]

So I think that all we need to do is adding a trigger mode where we currently have poly and single. That mode would “pretend “ at the moment of Note On that the MSEG with its current settings and modulations would have been playing from song start.

Whatever happens after, is up to the sound designer. If they want all or individual voices to speed up (and thus unlock from sync), then that is possible. If they want to keep in sync, they need to make sure that loop points are aligned when morphing.

As a nice touch, the Attack phase can be used to move the pattern back or forth for individual voices (or globally).

[IvyBirds]

You are basically talking about LFOs, where half of the functionality of the MSEG is discarded.

How do you want to sync an MSEG in Zebra 3 where the loop length is morphed from 2 quarters to 5 quarters? You can only properly do it by saying "this MSEG is not an MSEG, it is solely an LFO without any attack phase, without any release phase, without any loop points, and all of its snapshots that can be morphed between are actually of same length, regardless of what the user has drawn into it".

Urs you should check out the Shapemaster LFO/MSEG for VCV Rack by Mindmeld

https://community.vcvrack.com/t/introdu ... ster/13109

https://library.vcvrack.com/MindMeldModular/ShapeMaster

As an example of what is being discussed here. It's one of my favorite VCV Rack modules and would be cool to have something similar in Zebra 3

[pdxindy]

No, don't be like that. Don't be an apologist for Urs, some criticisms are valid and you shouldn't be shutting them down. "But look, it works for me!" is not really a good answer to valid criticism.

As soon as Andreya_Autumn posted about the playback problem, I agreed it was a good point. I also said in a subsequent post that just because I do this stuff in a DAW that happens to work better for it, doesn't mean everyone can and that it should be addressed in the synth.

It's okay that we have a differing points of view. The discussion, with its disagreements, is leading towards a positive outcome. It's going well.

[pdxindy]

So I think that all we need to do is adding a trigger mode where we currently have poly and single. That mode would “pretend “ at the moment of Note On that the MSEG with its current settings and modulations would have been playing from song start.

Whatever happens after, is up to the sound designer. If they want all or individual voices to speed up (and thus unlock from sync), then that is possible. If they want to keep in sync, they need to make sure that loop points are aligned when morphing.

As a nice touch, the Attack phase can be used to move the pattern back or forth for individual voices (or globally).

Yeah, that covers it. It doesn't need to stay sync'ed to the beat, just to project start so that playback is consistent wherever you start playback from.

[Andreya_Autumn]

I guess that works too yeah.

[pdxindy]

I guess that works too yeah.

Also has the benefit of not needing new tools or redesigned GUI elements.

[Andreya_Autumn]

Yup.

[exmatproton]

very true

[ptnah]

still inescapably tied to note-on, right?

[exmatproton]

still inescapably tied to note-on, right?

which is better then what we have now..but yeah..
It will keep the function of morphing and changing end-points though, so that is neat, i think

[Urs]

still inescapably tied to note-on, right?

Yes, but with the position it would have had, had the MSEG played from the beginning of the song.

(Just like the corresponding LFO mode, which "breaks free" if the LFO tempo is adjusted while playing. That's an important feature because otherwise, if you're 100 bars into the song and you change the LFO rate, and you still want to stay in perfect sync, the LFO will instantly jump like dozens of cycled forward or backward, which is glitchy)