I think I found out. Blakes great Video at Chuck Levins said 8 parts playable from the keyboard. But if you can take some of the existing multipart sounds and now combine them into one part (due to 128 elements available per part) this can effectively give you more parts under keyboard control. Still just 24 semitones part shift - not sure that's never implemented. But the new Montage M looks amazing.

Well this is an interesting question. For example, I have converted some TX816 performances (comprising of 8 parts each) into Montage format. I'm thinking that, how would I build e.g. two of the TX816 performances in the Montage M so that I could play both performances simultaneously using the keyboard? I mean, how would I combine them into one part?

(due to 128 elements available per part)

I'd be surprised if it supports 128 elements for each of 16 parts.

More likely it supports up to 128 elements in any given part with a max of 128 elements for ALL parts combined.
Similar to the motion sequence lane thing.

Won't know for sure until my shipment arrives and I can verify.

I have converted some TX816 performances (comprising of 8 parts each) into Montage format. I'm thinking that, how would I build e.g. two of the TX816 performances in the Montage M so that I could play both performances simultaneously using the keyboard? I mean, how would I combine them into one part?

You would only be able to do that if you converted your TX816 performances to AWM2 that uses elements. If your converted perfs are FM-X you won't be able to.

Elements, and the new 128 limit suppoer, only applies to AWM2 parts.

about the 128 voice polyphony per factory Awm2, AND 128 voice polyphony per user...nobody on you-tube demos including
Blake, has explained what should be necessary, so as it was with the Montage, it is 128 stereo/mono voices. This is important because most other synths including the Motif featured this, so if it was stated the a synth had 128 voice poly, then any stereo wav reduced poly to 64. It has been this way for the longest time. To simply say the M has 400 voice poly requires a little explanation. You have to make sure that you are using AWM2 FAC, AWM USER, FMX, and ANX together and keep track that you are allocating the necessary distribution, and that is not easy. Using 128 elements would absolutely use 128 notes of polyphony. Now I could be wrong to some degree, so please explain if you have something official.

about the 128 elements per part or performance, well how to view and edit that. The real evidence of a good design is ease of use and I doubt that Yamaha has rethought how to make an elegant and sensible navigation system. Montage was a mess. No you tube mention of the sequencer, but technical details on Yamaha state there is. What do we know about the sequencing abilities?

128 elements per Part doesn't mean if 128 elements were used you'd eat up all polyphony.

Think of a Part where you load in 32 variations of the same sound with different "character" differences for each. Put these 32 in a cycle group so it sounds like every keypress you get a slightly different sound. Layer into the same Part cycle groups with "odd" numbers so the other cycles would all repeat out of phase. You could end up with an interesting evolving sound and could put 128 elements to good use while only spending 3-4 units of polyphony.

I'm sort of surprised that XA Control still shows all the same old stuff. So many elements it seems like you'd want more control options (like all of the A.SW states that are missing) and even more interval based gestures or other things borrowed from S.Art

Montage classic saw significant growth in the early stages so maybe there's a longer game here.

Regarding the original question - I don't know. Read the docs and watched lots of videos and the live stream. The most definitive source in my opinion said 8 parts but the conflicting messages need to be sorted out. Add it to the Montage M FAQ.

I would sort this out on your own by going to a music store (when available) and checking out for yourself. I know it may be a bit, but you're going to want to test drive before purchasing anyways.

Did anyone get answers regarding the AWM2 Polyphony 128 User + 128 Preset..?

Is the polyphony based on the number of waveforms from User vs Preset memory?
Or is the polyphony based on the number of Performances that are stored in User vs Preset memory, regardless of where the waveforms are stored..!?

If it's based on the waveforms location (User vs Preset), I using 95% Preset waveforms, so I wouldn't see much increase at all on the Montage M vs the Montage classic.
If it's based on the Performance save location, same deal. 100% of my Performances that I use, especially for Live and most things, are all in User memory. So it's still basically the same AWM2 and FM-X polyphony on the Montage M (the way I save and use Performances, as well as waveforms), as it is on the Montage classic...unless I am missing something..!?

It's waveform location based. Not Performance storage bank based.

Yamaha will probably release some compelling content as a library so you'll get to use the user polyphony for that content.

So effectively for my purposes, the Montage M would not have much more polyphony than the Montage does the way I utilize waveforms, aside from a bit of additional polyphony from the AN-X engine.

I find it a bit odd that the MODX+ has the same FM-X polyphony as the new Montage M, and for the way I utilize AWM2 preset waveforms at 95% (5% User memory), the Montage and MODX+ would give me almost the same polyphony as the Montage M as well... Not that any of those synths are hard up for polyphony and have polyphony issues 😉

The "best" way to utilize this is likely to get your wallet out and purchase professionally crafted libraries that are specifically for Montage M. It's a wait-and-see what comes through in terms of that content.

[quotePost id=123660]

It's waveform location based.

If I'm not mistaken Bad Mister explained in the past that it had to do with the fact that preset waveforms are HIGHLY compressed but user waveforms were not. That there was on-board code to decompress the preset waveforms but no on-board to do the same compression for the user waveforms.
[/quotePost]
It would be nice if you could copy some of your favorite sounds from the AWM2 factory section into the AWM2 user section (even if it meant creating uncompressed versions of those waves to put there), which would give you a way to extend your polyphony among factory-supplied AWM2 sounds. Maybe?

I agree about the move for doubling polyphony. Yamaha has concerns about their internal waveforms "leaking out" but I'm sure this could be done in an internal/secure way that wouldn't allow for these waveforms to end up in save files. Flags in the all/user/library files could just mark the preset source for the data and also indicate it's been copied there (library slot or user bank). And so the process of restoring the save file would be to take all of the custom stuff and source it from the file itself while the pointed-to "user space presets" would be copied internally and never make their way out.

Yeah I agree they should have added more polyphony for AWM2 on both User and Preset waveforms (if they had to divide them up like they did). Even if they gave each AWM2 192 each instead of 128, just to differentiate from the MODX+ a bit more. But I'm not concerned about polyphony on my Montage so it's never been an issue for me to begin with. Having extra Elements per PART would definitely come in handy, but I'll make due like I always do on the Montage...there are many options to turn the 8 KDB CTRL PARTs into 16 or 32 on the Montage classic!

Regarding creating uncompressed versions of those preset waveforms and have them in User memory, you can always use SampleRobot! Find the waveforms you are looking for, create an Init AWM2 PART, shut down all effects, add the waveforms in, and sample them, then save as a library file. 😉

Then grab the Performances you want and save them custom to User memory, and change each Element waveform to point to the new library file (or Library Import them into User memory if you don't want them in a loaded library)... 🙂