You're going to need two Performances that can Seamlessly Switch for both of these desires.

To the best of my knowledge, this is not possible with anything onboard the Montage/MODX.

it might be possible with an external controller that's rigged up to send macro controls, but I'm not sure that these ranges are exposed to MiDI/CC/OSC/Sysex/etc

Multiple Performances is an option. I did that before with previous generations. Today, I've found there's enough resources within a single Performance to cover me and I do have situations where I'm trying to achieve the same goals as you. YMMV.

(Mostly) what you can do with two Performances you can do within one Performance as long as 8 Parts is enough (at least without the MIDI loopback cable or similar "tricks" ). You would have two duplicates of Parts with different ranges but the same samples/envelopes/etc. Then your best bet would be to use a scene switch to set different combinations of the Parts as active using Keyboard Control. Keyboard control will approximate the SSS experience and will actually be "better" because your sounds could span several back-and-forths if you held keys down where SSS would cutoff a sound from the 1st SSS Performance's held notes when the 3rd Performance is switched to. This may not be a big deal to you -- I only mention it as an FYI.

Example:

Part 1 range C-2 to E2, Part 2 range F2-G8
Part 3 range C-2 to E3, Part 4 range F3-G8

Switching between Parts using Keyboard Control should also address the external needs since one Part could be Zone octave shifted while the other is not.

Example:

Part 5 Part Zone control - no octave shift
Part 6 Part Zone control - octave shift

Parts 1 and 3 are "identical" (mostly), Parts 2 and 4 are "identical" (mostly), Parts 5 and 6 are identical except for zone settings. Which sets of Parts are active at any given time is defined by the scene buttons.

Some of this you could do with ASSIGN buttons (only the Part1-Part4 stuff for ranges above) - but it wouldn't cover the external MIDI goals since ASSIGN is not a tool for stopping a given Part from sending MIDI information.

There are some tricks you can use, but you lose economy of resources.

1) Create 2 identical Parts, each with it's own required splits and note shift settings. You can use Scene buttons to activate/deactivate as necessary. You obviously burn a Part Slot doing this.

2) Similar to 1, but create 2 sets of identical elements within a Part, and set up a cross fade between each set's levels. This limits the sound creation to 4 elements (4+4=8), but you can use the note limits per Element, with a wider Part Limit to encompass both Element Sets. You can also use Element Tuning to fiddle the Note Shift (I prefer Note Shift for cleanliness, but Tuning is a potential workaround. This is less desirable/flexible than 1) but might be enough to get you moving forward. It also eats more polyphony (the muted Element set is still "playing").

Alternately there may be a way to switch or rotate Elements ON/OFF using XA Control... I don't recall exactly, but it is worth having a look.

Some of this you could do with ASSIGN buttons (only the Part1-Part4 stuff for ranges above) - but it wouldn't cover the external MIDI goals since ASSIGN is not a tool for stopping a given Part from sending MIDI information.

This involves XA Control.

Thanks for all the interesting ideas.

I'm only talking about 2-part sounds... one single-part sound to the left of the split point, and one single-part sound to the right.

The big limitation of Andrew's initial idea of multiple Performances (post #118849) and Jason's initial idea of using Scenes to select different combinations of 8 Parts within a single Performance (post #118856) comes into play with what I said in the second half of my OP, that at any time, I might want to change the sound of either my LH part or my RH part, on the fly. So the inherent problem with using combinations of Performances or combinations of Parts is, once I switch one of the sounds, my ability to continue to make these manipulations ends, because those alternate Performances or Parts will still invoke the old sound, and not my newly selected sound.

From what I can tell, the only viable approach would seem to be the possibility of doing these things with a sequence of sysex strings that I could send when desired from, for example, an iPad running Camelot Pro. And in fact the iPad would be there anyway, because that's the source of the external sounds.

[quotePost id=118872]

I'm only talking about 2-part sounds... one single-part sound to the left of the split point, and one single-part sound to the right.

That leaves room for SEVEN pairs of duplicate parts just waiting to be switched in.
[/quotePost]
A Performance can only have 8 keyboard-playable parts, so as I see it, there would be a total of 3 pairs of duplicate Parts to choose from, not 7 (4 pairs total... 1&2, 3&4, 5&6, 7&8... where each pair has the same sounds, but with different Key Ranges and/or Note Shifts). But that's okay, that would be enough. Or it would be, if switching into other sounds entirely wasn't an issue, which I'll get back to. (I guess you're envisioning some other approach here.)

[quotePost id=118872]

at any time, I might want to change the sound of either my LH part or my RH part, on the fly.

Won't you know in advance WHAT sounds you are going to want to use? Do you really need to be able to choose from more than 14 sounds to switch in?

Not trying to 'rain on your parade' but are you saying that during a performance you want to be able to switch to ANY of the parts in over 2200 performance without knowing in advance what those parts will be?
[/quotePost]
I don't need to be able to select from any of 2200, but I might want to be able to switch from among dozens. (Not that I even see a way to do 14, in the context of the task at hand.)

[quotePost id=118872]

once I switch one of the sounds, my ability to continue to make these manipulations ends, because those alternate Performances or Parts will still invoke the old sound, and not my newly selected sound.

Huh? Why would the old sound be invoked if you have replaced it with a new sound. Please explain that.[/quotePost]
Let's say I'm starting with my two sounds in Parts 1 and 2 (RH and LH sound respectively). I use a Scene change to switch to Parts 3 and 4 of the Performance, which have the exact same sounds, except with differently defined key ranges such that the split point between the two sounds is higher. At some point, I change the sound in Part 3. Later, I want to go back to the original split point. If I use a Scene change to switch back to Parts 1 and 2, I get back my original split point, which is what I want, but since my RH part has changed from Part 3 (which has a sound I have changed) back to Part 1 (which still has its original sound), that old sound would again be invoked with the scene change.

[quotePost id=118872]
I'm not aware of any "Param. with part" options when using SysEx.[/quotePost]
I don't think any such option is necessary.

If I use SysEx to change the key ranges of a part, once those key ranges have been changed, the keyboard doesn't "know" I used sysex to change those parameters as opposed to changing them manually, the parameters simply are what they are. If I bring a new sound into such a part, telling it to bring that new sound without its own parameters (using the existing parameters as they are instead), the new sound should respect the defined key range, regardless of the fact that I previously happen to have used sysex to define that key range.

[quotePost id=118872]
You can certainly use SysEx to swap in a part but I can't find any parameter that can be used as the equivalent of the manual button to either include or exclude part parameters such as key range.[/quotePost]
As in your post #118859, I was talking about using sysex not to swap in parts, but to change something like a key range parameter (or note shifts for internal/external sounds). Sounds would be swapped in the "normal," way, selecting a button and using the Category Search button on the MODX screen.

[quotePost id=118872]
Modx/Montage dont' have any real 'programming' functionality. If you want to write control sequences you need to use an external app/device.

If you are only using two parts for your performance then a 'viable' way might be:

1. Build a performance that uses all 16 parts and populate part 1 and part 2 with your starting pair.

2. Create SysEx control sequences that can build out parts 3 and 4 with the next pair of parts.

3. Create other sequences to build out parts 5/6, 7/8 and so on.

3. Use scenes to switch between pairs of parts - you have 8 paris of parts available and you have 8 scenes available.
[/quotePost]
If I'm going to use an external app to send sysex (as I mentioned in my example of using Camelot Pro), I think all I need to create is 4 sets of commands that can be sent on command. One to define one set of key ranges for Parts 1 and 2; another to define the alternate set of key ranges for Parts 1 and 2 (for the alternate split point between my LH and RH sounds); another to do the +12 note shift up on Part 2; and another to "un-do" that note shift (i.e. a -12 note shift down). Everything else (all my part selection) could be done as usual, by swapping sounds in and out of Parts 1 and 2 from the MODX's own screen.

But I had been hoping there would be a non-sysex, self-contained solution within the MODX! Apparently not, oh well.

It's actually very easy to do most of this on the Fantom-0. The only limitation here on the Fantom-0 is, when you want to swap new sounds into a part on the fly, they can only be internal sounds, whereas on the MODX, you can easily swap in either internal or external sounds, that's why I was hoping to be able to do something similar on the MODX. (I have both boards.)

[quotePost id=118875]I was just pointing out that there is more work to do than just using SysEx to change parts. You also need to do ALL of the setup work.[/quotePost]Right, but I was not envisioning ever using sysex to change parts. Only to change key range and note shift parameters.

[quotePost id=118875]One thing to remember is that 'note shfit' does NOT change the note you play - only the note you hear.

Middle C for Yamaha is note 60. Note shifting doesn't change that. If you note shift up an octave you are STILL playing note 60 - but you are hearing note 72.
[/quotePost]
Right. And as also mentioned in my OP, that change (so that what you're playing sounds an octave higher than where it is being played) would need to affect the sound in a part whether that sound were internal or external, which I thought might have been another complication to deal with, I'm not sure (I haven't dealt with that kind of thing in a while).

Still, without there being that "non-sysex, self-contained solution within the MODX," I doubt I'll pursue this further. But I appreciate everyone's input!

(As an aside, though, if I were to get an app like Camelot Pro involved, there may well be other ways to accomplish what I'm trying to do anyway, e.g. by setting Local Off and treating all internal sounds as external sounds, then using the app itself to manipulate the key ranges and transpositions as needed, instead of trying to invoke MODX' internal functions for these things.)

My response assumed goals 1 and 2 were paramount while the "Also ..." that came later was a nice-to-have. Therefore, I focused on the 2 goals and did not address the "Also ..." part which involves swapping out other sounds in these slots.

If that's also a requirement then you would probably want to save multiple "versions" of each so you have one Performance that's shifted and one that's not and then load into your Part the right User Performance/Part that already has the programming you need. This places burden on you when swapping out the sound -- but you're already doing a lot of work if "Param with part" was seen as a viable solution. I don't think any of that (meaning using "Param with part" ) is absolutely necessary if you have "the right" Performance/Part settings for your swap pool.

This may not solve all of the gotchas -- and maybe it's not a preferred workflow. However, that's what I can think of as an in-the-box solution.

For an out-of-the-box solution I would probably just bulk dump what I wanted into the Part I wanted to split shift or octave shift (including external gear). Most of this can be done in a simple "scripting" language such as Python or done in a portable Web (MIDI) local page using javascript. Leveraging a browser makes the interface (meaning GUI) easier to develop and such a page on a tablet would provide touch panel access to swap out what you wanted.

You could bulk any parameters you want. There wouldn't be a need to change some things you would like to keep the same. Maybe instead of bulk a series (a lot) of parameter changes. Which one is picked really depends on how you would need to carve out the items you didn't want to change.

All of that is a lot to do for the typical user. An extreme "DIY" approach.

[quotePost id=118877] I don't think any of that (meaning using "Param with part" ) is absolutely necessary if you have "the right" Performance/Part settings for your swap pool.[/quotePost]
My thought was that I could end up wanting to change sound X to sound Y at any time... i.e. I could be in a Scene with the original split point (key ranges), or I could be in a Scene with the altered split point. If I want to bring in the sound and have it respect whatever split point has already been defined, then I need to be cognizant of "Param with part". (Which is easy enough to do.) Though it occurs to me that it still might not respect any previously invoked Note Shift, hmmm. But regardless, as I said, I think the entire idea is just a step too far in using the board in ways other than intended, and not worth the effort.

At some point, I might try to do a video on the Roland, to better illustrate exactly what it is I'm looking to accomplish, and why it could be valuable.

If you modified CFX Concert Part Note Shift then saved this as "CFX Concert +12" then recalling "CFX Concert +12" instead of "CFX Concert" would not need to worry about if "Param with part" does what you want or not. All the programming you're wanting is already in the Performance you're swapping in. This applies to split locations too.

[quotePost id=118881]If you modified CFX Concert Part Note Shift then saved this as "CFX Concert +12" then recalling "CFX Concert +12" instead of "CFX Concert" would not need to worry about if "Param with part" does what you want or not. All the programming you're wanting is already in the Performance you're swapping in. This applies to split locations too.
[/quotePost]
Of course. But the idea was to be able to have a single button (or two) that you could hit, and every Performance would be able to behave the way you wanted. Not take the 50 sounds you might do this with, and create two or more different versions of them, and then make sure you navigate to and recall the correct version for the particular situation. 🙂

[quotePost id=118880]The good news is that the part key ranges are respected if you don't bring in the parms of the new part.

The bad news is that a new part means new elements/operators and who knows whether their key ranges even match the old part level ranges let alone any new one you create.

So the only way to guarantee that the key ranges are what you want at both the part and element/operator level is to set them yourself.

Which begs the question of how are you even going to know if the element/operator level ranges are right for what you are doing if you want to select from dozens of things.[/quotePost][quotePost id=118882]I think you have to have a certain level of fore knowledge about just which parts you are going to want to bring in at any given time or something is going to get seriously out of whack.
[/quotePost]
This part is SO much simpler than you think it is.

Experiment: Create a 2-Part split with two single-part sounds, say a string pad on the left and a piano on the right. Simple, right? Now deselect all the Param with part fields. Change the piano to some other single-part sound, say a synth brass sound. It will simply work. Now use note shift to raise the sounding octave of the left-hand strings 12 semitones. Still sounds fine. Change the upper limit of your left hand string sound and the lower limit of your right hand synth brass sound, such that you have a new split point that gives you more notes for your string sound (taking them away from the available notes for your synth brass sound). Still sounds fine. "New elements/operators and who knows whether their key ranges even match the old part level ranges" and "a certain level of fore knowledge about just which parts you are going to want to bring in at any given time" are really not factors at all.

Basically, what I described there is an example of what I'm talking about doing. Except imagine a single button that would raise the sounding octave of the left-hand strings 12 semitones instead of having go go to the edit screen and raise that value by 12, which is not something you probably want to do in the middle of a song. And similarly it would be great to have a single button that would select a pre-defined alternate useful split point to get some more space for your LH sound, to again spare you from having to go to the edit page and edit the top note of one part and the bottom note of another, which again, you're not going to want to do mid-song.

So kind of getting back to the first paragraph of this post, sure, you could configure a ton of Performances for all the possible combinations you could envision ever wanting to play. But what I was aiming for is that you could pick whatever sounds you'd want, at any time, and quickly get them in a choice of octaves or key ranges, without having to set up every possible combination in advance.

@Jason & @Bill and @Yamaha

Try Ableton Live's clip's based approach to dynamic editing of Parts. It's a DAW revelation, and why you hear/see so much talk about Ableton Live, and why it has inspired all DAW operation since.

This is sort of what it seems like AnotherScott has gotten addicted to within Roland's OS, as they have a cutdown version of a clip based approach in their Fantoms. It seems he's wanting some of that life of live, dynamic freedom of part swapping for/within with MODX/Montage - and I understand. Part of this low hanging fruit leads to adding some of that freedom to the Montage/MODX: https://www.yamahasynth.com/forum/low-hanging-fruit-fixes-improvements#reply-118827

It's highly addictive, and completely changes your views on live compositional experimentation. This is why Ableton has been such a huge success and whole ranges of hardware controllers are sold for taking it to a tangible experience, and why Roland has somewhat copied it into their Fantom range, and Bitwig, Logic. Ardour and others are now mirroring versions of this functionality.

It is particularly helpful with drums, bass and arpeggios, the "song making" backing parts that are usually static can easily, instantly, become dynamically responsive whilst leaving more than enough user fingers and cognitive space for lead/solo/melody playing and harmonising over the top.

Of course. But the idea was to be able to have a single button (or two) that you could hit, and every Performance would be able to behave the way you wanted. Not take the 50 sounds you might do this with, and create two or more different versions of them, and then make sure you navigate to and recall the correct version for the particular situation. 🙂

I expected the "work" involved during a gig to pick the right one might be pushed back. However, I see the amount of "work" to recall even a single Performance as a lot of overhead by itself - adding that you seemed OK with pressing the "Param with part" selection in addition to picking a Performance (even if it would be only one regardless of split). For me, even things that seem "OK" would be too much work.

As it stands you have 50 or more sounds you would have to pick from when swapping out even if you didn't create two versions of each. That's over my limit as it is and picking from 100 doesn't seem that much different. To me using [CATEGORY SEARCH] on the gig is above my cut-line.

Any buttons you press can't dynamically change splits nor can they affect external devices in terms of octave shifting. At least not using on-board capabilities.

... regarding Fantom - I haven't really gotten much use out of it so I'm nearly 0% acquainted. I'd be interested in learning about how this works for them just for my own edification. Are you swapping out Tones or Zones on the Fantom (or something else)? At any rate - learning the details of how another keyboard solves a problem and does this "better" (I only quote because this is subjective - not because I think one way or the other) would be good to articulate. Thanks in advance.

@Andrew

I get what you're saying - but it's about the sequencer block which is a different side of the tracks versus the sample/collection/folder hierarchy which I'm guessing is what is enabling this -- not clipped based sequencers or anything close to this. As such - I think what you say without the wider context of the discussion may be correct -- but not very applicable to this split/shift task. Regardless of applicability I agree that there are aspects of Roland's sequencer that are preferable.