Feature Request : MPE support

[Stollmeister]

Great news Urs!

About the pitch bend, I suspect that having the sound be less than perfect when bent 48 semitones is fine. Such extreme bends will be very rare. For me, it is simply neat to not having to change the Roli from 48 note pitchbend to 24 just for Zebra. It does not mean I will actually bend that far. (My Roli is only two octaves so I couldn't even if I wanted to.) I bet Zebra 3 won't sound worse than pitchbent samplers.

[thetechnobear]

Yes, and I reiterate: We'll do +/- 48 semi pitchbend and CC74. We'll do this in subsequent updates. If you spot an update without +/-48 semi pitchbend in any of our traditional public beta phases, remind us. About CC74, we'll do this with a general refactor of our standard MIDI controllers. Not sure if it's the next cycle of updates or teh one thereafter.

great news ... are there likely to be any updates in the near future?

What I'm sceptic about is the whole thing about a master channel, number of channel restrictions, some kind of zones and a whole bunch of mode switches. Looking through different drafts, I found contradicting information and I think these might be moving targets.

fair enough, definitely is the PB and CC74 that 99% of what we need for MPE

Note that some of this stuff means sound quality compromise. In Zebra 3 for instance we will have modules that will either use excessive CPU if pitched beyond +24 semitones while playing, or possibly exhibit a lot of aliasing if pitched below -24 semitones. There are processes and algorithms that "prefer" not to be pitched in such extreme ways, but are absolutely fine within boundaries that have been considered reasonable before MPE.

I wouldn't worry about that too much... its not very common to be doing a 4 octave slide

the main reason, Id like 48 semis, is so that the 'calibration' is the consistent with other MPE synths. so when I switch VSTs I don't have to reconfigure the controller

thanks for supporting us
Mark

[ulph]

I'm about to try a Linnstrument on loaned time.

Any gotchas i need to be aware of in case any of you have experience with that unit with diva/zebra? Pitch bend i know of.

[pdxindy]

Yes, and I reiterate: We'll do +/- 48 semi pitchbend and CC74. We'll do this in subsequent updates. If you spot an update without +/-48 semi pitchbend in any of our traditional public beta phases, remind us. About CC74, we'll do this with a general refactor of our standard MIDI controllers. Not sure if it's the next cycle of updates or teh one thereafter.

What I'm sceptic about is the whole thing about a master channel, number of channel restrictions, some kind of zones and a whole bunch of mode switches. Looking through different drafts, I found contradicting information and I think these might be moving targets.

Note that some of this stuff means sound quality compromise. In Zebra 3 for instance we will have modules that will either use excessive CPU if pitched beyond +24 semitones while playing, or possibly exhibit a lot of aliasing if pitched below -24 semitones. There are processes and algorithms that "prefer" not to be pitched in such extreme ways, but are absolutely fine within boundaries that have been considered reasonable before MPE.

I think the +/-48 and CC74 are what most people want (and would be satisfied with).

Also, as some people pointed out... it is not that people want to be able to bend notes 48 semitones, it is that they want a consistent experience across plugins/controllers and with editing data in the DAW.

My single biggest wish for my u-he synths is getting those two things sorted for MPE support. I would happily pay for it too.

[hyperscientist]

Also, if they can't finish it they should not call it standard and put their users on our case.

This is how standard gov organisations often work. For example The Web Platform (HTML, CSS etc.) - for the new spec to become a standard it has to be implemented by minimum of two web browsers already. This is how viability of ideas is being put to real world test.

Specification may have many states along the road: initially it's just notes somewhere on some wiki or github, then working drafts (browser vendors already work on implementation at this stage), then it's a candidate recommendation, proposed rec and finally a recommendation.

Often a spec is changed to reflect real world browser vendors implementations or even (!) incorrect usage in the wild (by users / web developers).

Companies that rush to implement drafts should really be applauded for their willingness to work for betterment of the whole community.

[ulph]

I noticed a rather significant slew to aftertouch on Diva and Bazille (haven't tried the others yet)

I'm trying to make patches that do more or less without velocity/adsr and rely more on my finger pressure.

I get better (less smoothed) response when using expression instead of pressure (configurable on my linnstrument), but then the onset/release is so quick there's a click instead.

Is the slew on pressure by design?

I can probably find a workaround by summing control, esp for Bazille... I need something with in-between those extremes.

[EvilDragon]

Slew is adjustable in Preferences.

[ulph]

Yes but this is not the slew found in settings. There is a noticeble ramp in if slamming down on my Linnstrument when using pressure. Almost a quarter of a second. Not at all when using expression instead (linnstrument can be set up like that so its fine).

Looking at the midi data it looks good, so i am suspecting uhe aftertouch handling being designed like this.

Will investigate a bit more tonight.

[Urs]

Slew limiting on AT is generally too in our stuff and doesn't respond to MIDI preferences.

Over the past few days I've fixed this, and I also added +/-36/48 semi pitchbend to all remaining synths. So when we post a Diva beta in a few days it'll be pretty much "there". Only thing missing is CC74 which should be easily routed to 2 or 11.

Cheers,

- Urs

[ulph]

Urs, neat!

Ironically, some slew seems to be useful though to distinguish from any envelope attack and the pressure. (assuming they both control filter and/or volume). Maybe an option to have it settable? (or we just use the LAG modifier etc)

EDIT:
cleaned up post

[ulph]

As I don't see this mentioned anywhere;

Zebra (HZ) has a problem when it comes to pitch bend and MPE?

As far I tried, pitch bend seemed global on Zebra (HZ), meaning I bend one note will bend all.

[ulph]

For anyone curious of what it can sound like, here is a clip of me playing some Diva on a Linnstrument.



Diva works well, although ACE/Bazille are growing on me as there are more options to design patches that both respond well to pressure and velocity.

[Urs]

As I don't see this mentioned anywhere;

Zebra (HZ) has a problem when it comes to pitch bend and MPE?

As far I tried, pitch bend seemed global on Zebra (HZ), meaning I bend one note will bend all.

Yes, Zebra's voice handling could not be ported to Multichannel MIDI. That only worked for the later plug-ins. Hence we're working on Zebra 3 with all new voice processing.

[Chris-S]

Not sure how to assign channel pressure or release velocity (Roli Lift) yet though.

Do u-he synths already support release velocity?

[EvilDragon]

Nope, unfortunately


Which seems weird, it should be pretty straightforward to implement, since velocity is already there. Release velocity is just the reverse of that