Hi Falk,
I think that You can use the same voices in Montage as You do in Your MoXF.
The different is only the thing that now You have one mode for all voices, performances and mixes. I'm not sure but I think all Your mixes will work correct with the Montage. And what I heard in YouTube that the CFX authomaticly uses the new sample based CFX.
I hope I'm right about this.
All the best
Rainer

Excellent questions and are at the core of understanding what's new and different about Montage

1) For projects already produced with MOXF8, I could relatively easily transfer them to Montage since all XF (and therefore also MOXF) voices are included in Montage as "single part performances". So I would have to create a 16 part Montage performance containing the "MOXF voices" as parts (using the respective single part performances in Montage), assign effects, volumes, etc. as I did in my MOFX mixing for that project, and then everything would basically sound the same as before? Plus, as a benefit, I would have 8 more (double) insertion FX slots available? Is this basically correct?

Essentially correct. But not only do the 8 additional PARTS now have Dual Insertion Effects, there is a separate Dual Insertion Effect available for the A/D Input PART.
But essentially what you state is spot-on. But it will sound BETTER than it did before! 🙂

2) Now lets say I would like to replace a MOXF piano in this project with the new Montage CFX piano. The piano part is contained in a single MIDI track, say channel 1. Now if I understand it right, many of the new performances are multi-part (including the CFX). So how could I reallocate an existing MIDI track to the CFX piano? Would I have to duplicate it a couple of times, creating several tracks, and then assign to each of them different MIDI channels matching the CFX parts? This is not clear to me...

The principal CFX piano Performance is called "CFX Concert" - it occupies four PARTS. You can trigger it with a single MIDI channel of data - that is not a problem (the Montage has a special MIDI I/O mode called SINGLE designed to use with multiple PART single instrument PERFORMANCES).

The issue you will discover coming from MOXF or Motif XF is that if this piano takes up four PART slots. When re-voicing your data, you can run out of PART... Not MIDI Channels, PART slots. If you had used all 16 PARTS in your MOXF sequence, how can you select a FOUR PART "CFX Concert" piano?

3) More gererally, what is the best way to record multi part performances to a DAW? How will the result look like, will you have different tracks, each with a different MIDI channel in the DAW for each part of the recorded performance? So if a piano consists of (say) 5 parts, there will be 5 tracks, using MIDI channels 1-5, potentially cotaining different note / velocity / controller data, which (when played together) gives you the piano sound?

When working with Montage, rendering audio becomes an important part of the workflow.

Because it breaks open the architecture and breaks the MOXF/Motif XF mold where One PART, occupied One MIDI channel and used One Track.

Now your Piano could be 4 PARTS, your Strings could be 7 PARTS, your Brass could 5 PARTS... that 16 PARTS and you have only just created 3 tracks. Montage sounds can be HUGE in comparison to the MOXF/Motif XF... HUGE.. not a marketing word as used here, but HUGE!!! If the sound you choose is multiple PARTS choose it because you are using what it is capable of doing. What's new about Montage will not be contained by the same "one Part, one Channel, one Track" paradigm of its predecessors.

I've re-voiced several of my MOXF/Motif XF sequences and can tell you, I wind up rerecording many of the PARTS because - now I have more I can do with them. But AUDIO becomes a portion of your new workflow! If you are fine with how it sounds as all single Parts - you can easily just do that. But when you want to add one of the new Montage HUGE sounds, you will need to think this through. Weighing the benefits on a per composition basis.

So rendering AUDIO - committing your MIDI recording as an AUDIO WAVE becomes a necessity. Once it is audio, you release your hardware for new duties.

The Direct Performance Recorder greatly eases the task of recording what you do on the Montage as MIDI data. You can then drag and drop it as a .MID file into Cubase (or your favorite DAW) work on correcting any mistakes and then take advantage of the 32 AUDIO BUS Output capability to render your data as AUDIO.

Thank you BM that was helpful. I think I got the overall idea of Montage now:

- on the one hand, you can use it as a 16-times multitimbral instrument, featuring 1 part "conventional voices", no seamless switching, the "old way"

- on the other hand, you can bundle all resources of the instrument, effectively turning it into a monotimbral synth featuring next-gen-up-to-8-part "voices" (I know, its called performance now), allowing seamless switching (thats why you cannot bundle 16 parts), the "new way"

- and, everything whats in between those two extremes

The "old way" is more or less Motif style, plus some enhancements (more effects, FM-X). The "new way" is obviously interesting for live, but also for music production because of the superior quality of the multi-part voices (aka performances). The price you have to pay is rendering tracks to audio (just as you would do with a monotimbral synth) because you are running out of parts...

Will have some follow-up questions on the workflow for the latter - since I am a rather lousy keyboard player, I have to edit & tweak my (MIDI) recordings in a DAW before rendering anything to audio. What you suggested, recording to the internal performance recorder, then transferring to the DAW (do I still need a USB stick for this?), then fitting into a project, tweaking, then rendering to audio, sounds a bit cumbersome to me. Will there be a way to directly record those multi part performances to the DAW?

Will have some follow-up questions on the workflow for the latter - since I am a rather lousy keyboard player, I have to edit & tweak my (MIDI) recordings in a DAW before rendering anything to audio. What you suggested, recording to the internal performance recorder, then transferring to the DAW (do I still need a USB stick for this?), then fitting into a project, tweaking, then rendering to audio, sounds a bit cumbersome to me. Will there be a way to directly record those multi part performances to the DAW?

Yes, of course, you can record multi Part Performances to your DAW. You have a choice. Recognize that what you say is "a bit cumbersome" for you does not necessarily apply to the next person... They may find setting up to record the multiple Parts to their DAW what is a bit cumbersome... The good news is, you'll each have a choice of methods to do what you need.

Because of the nature of some of the Performances in Montage (the use of multiple Pars), rendering audio will be a part of your workflow. Allowing you to Freeze (most pro DAWs have a way to render 'read only' versions of your data), while you are in the stage of production where you are still editing note data in your playing. However, once you get it like you like you can 'commit' it to an actual audio track. Thus the reason for 32 audio bus outputs via your USB connection. (Keep you MIDI data as your extra level of "undo" in case you later want to redo a musical part).