First, it's worth looking at a quick primer to MODX: https://www.yamahasynth.com/modx-category/mastering-modx-navigation-tips

In the MOXF, a Performance consisted of up to 4 parts. The sounds that occupied parts were called Voices.

You're right, there are no voices. Before, you could edit a voice and all Performances which used that same (user) Voice would inherit the changes made to the underlying voice. There are no voices in MODX. Instead, the Performance holds its own information and is not affected by changes in other Performances (no sharing).

Inside a Performance is a PART. Each PART is roughly equivalent to a voice. One big difference is that each PART is unique. Even if you end up combining an "S700" piano PART with an "Ensemble String" part to create your own 2-PART performance - the "S700" AND "Ensemble String" PARTs inside this new Performance are not connected to the "S700" and "Ensemble String" presets inside MODX. They are their own copy and can be modified/altered without affecting any other Performance.

Some Performances are made up of multiple Parts. These can be seen as traditional Performances in old architecture. Some Performances are made up of only one PART. You can treat these like Voices. You can even, if you want, treat a multi-PART Performance like an "uber-voice" -- say you really like a 4-PART piano and do not want to use a 1-PART version. You could merge this 4-PART performance with some other 1-PART string to create a layered sound.

The way I typically construct Performances is search for single-PART Performances as the basis for combining what, before, were voices. [CATEGORY] (the search button) can filter out everything except for "Single Part" Performances. I would merge together multiple single-PART Performances to create my new user Performance.

As mentioned, you can also merge together a Performance that has more than one PART. Since typically I would use single-PART - if I did want a multi-PART Performance combined, I would first start with the multi-PART Performance (maybe delete PARTs that do not apply to my new Performance - maybe keep all of them) - then add single-PART performances to that multi-PART performance as a starting building block.

This generation, you can have up to 8 PARTs under direct local keyboard control.

In the MODX, a Performance consists of up to 16 parts. What are the sounds that occupy those parts called?

PARTs

So for example, if you're looking at the parts of a Performance, and you call up the category search screen and select a category to see a list of sounds you can choose for that part, are those sounds called something? And to what started me on this question, where can I find a list of those sounds? I downloaded the data list, and there is a lists of all the Performances, Waveforms, sample Live Sets, etc., but apparently no list of what the available sounds are that can be brought into Performances, i.e. no equivalent to the MOXF's Voice list. Am I missing something?

The Performance list is the closest you can get. It would help if the list showed the number of PARTs each Performance holds. You would then be able to recognize which PARTs in the documentation are single-PART Performances and roughly equivalent in usage to Voices. For now, use [CATEGORY] (the button) and the built-in feature to restrict (filter) the list by Single-Part Performances.

You'll also notice that each Performance in the list (when not filtering by type) is a different color. Blue or Green. The Green Performances are single-PART and the Blue Performances are multi-PART.

In the MODX, a Performance consists of up to 16 parts. What are the sounds that occupy those parts called?

PARTS.

Here’s what you’re missing.

What was VOICE Mode is no more. In the old MOXF system, a Performance was made up of four Parts, each Part referenced (pointed to) a Voice location back in Voice Mode. A Performance Part was a set of offset parameters applied to the Voice. For example, if you placed “Full Concert Grand” into a Performance and you Note Limited it to C3-G8 as if making a split... the Performance referenced the original Voice but limited access to just C3 and above. If your Bass sound was in the User Voice Bank at USER 1(A01), in Part 2, likewise it referenced that User location but offset it so it only played C-2 thru B2... Performances were just a set of offsets to the Voice data

If by accident you loaded a different User Bank whatever occupied User 1(A01) would serendipitously now be apart of your split Performance. Remember? This will never happen on the MODX

Well, the MODX has so much more memory than the MOXF did, every program has its own unique memory. No more pointing to or referencing a location. No more serendipity programming.

Every program in MODX is a Performance. A Performance can be 1 or more Parts, up to 16.

All 1300+ MOXF/Motif XF Voices are in the MODX as Single Part Performances - they form the backbone of the MODX sound set.
You’ll find the “Full Concert Grand” as a Single Part Performance... it contains 15 empty Part slots... you can merge it with other programs... you’ll see it occupies Slot 1, and there are 15 empty slots for you to add other Parts.

The Performance Mode in the MOXF was limited to four Parts... all were on the same MIDI channel. The equivalent thing in the MODX is an eight Part (KBD CTRL) Performance... but each Part can have its own MIDI Channel.

Eight Parts can be linked under what is now referred to as KBD CTRL (Keyboard Control). You can play these directly or place them under control of one of the eight available Arpeggiators. The MODX can transmit on as many as eight MIDI channels simultaneously (you’ll want your Drum Arps On their own channel, after all!

What’s on tap in the MODX is tons (and tons) more memory. Enough memory so each Part in a Performance exists as a unique entity.
Single Part Performances are listed in green in the Category Search... Multi Part Performances are listed in blue.

You can setup a 16 Part Performance, one Part on each MIDI Channel for sequencing purposes... or for when you need rapid access to individual instrument sounds, with no interruption in the sound. You can move between sounds within the same Performance seamlessly.

The MODX, now at version 1.10.x has 2,143 Factory Preset Performances... more to come shortly.
The User Bank is initially empty... it contains 640 locations. As you create, Load, or import programs to it, the Performance folder will keep track showing you how many of the 640 you have occupied. It’s basically five Banks of 128 Performance locations.

The User Bank area can also includes room for 256 User Arps, 256 Motion Sequences, 32 User Curves, 2,048 User Waveforms, 2,048 Live Set slots, 8 Micro Tunings, and 128 Audition Phrases.

Once you filled this you can take the User Bank and SAVE AS... a Library. Each Library is like your own ROM Bank that stays in Flash until you Delete it or initialize your MODX.
You can have eight Libraries + the User on board.

Yes,
2,143 Factory Preset Performances
5,760 User + Library Performances

Yes, each Performance could hold 16 old style Voices (I don’t even want to know how many that is, because you don’t actually use the memory like that.. so it’s a meaningless spec. It sort of like, you can build your sounds with whatever you need to make it work.

Any Part, from any Performance can be copied and used in any Part of any other Performance. And each has its own integrity.

"Full Concert Grand" was a three way velocity switching piano... built with 8 Elements
With MODX you can break the 8 Element limit... using multiple Parts linked under KBD CTRL to build your sound. For example, the “CFX Concert" Piano is a four Part (18 Element, 9-way Velocity switching) Piano.

Building these big sounds means more nuance, more articulations and variations within a sound. Take a brass ensemble... if you are the keyboard player on the gig, and need to cover the Horn section stuff, with multiple Parts in one Performance you can have sforzando swells on the MW, rips and fall offs on the Assign Switches, separate sax section, separate brass section, a separate section of high trumpets that you bring in with the Super Knob... etc., etc. all in that one Performance!

You don’t have to switch to separate programs to have access to all these different articulations- they are all under your control within the one Performance.

Can you layer eight sounds? ... yes but that would be the worst way to use the tech... it takes no imagination to just layer sound on sound on sound.... however, you can determine, with the massive Control matrix, exactly when each component should play! Extremely programmable.

You’re gonna love it!

Great info, thanks to both of you!

So am I understanding correctly, then, that when you are editing a Part in a Performance, and you call up Category Search and pick (for example) piano, what it is actually showing you are all the factory-supplied single-part Performances that are piano sounds? IOW, the sounds that come up when browsing for a Part are not any kind of separate sound library, but rather a filtered subset of the same Performances you see when browsing all Performances? And you are then essentially placing other Performances inside the Performance you are editing?

There are two ways to modify the PART - neither of them is automatically a filtered list.

When you click on an empty slot by pressing the [+], you will see the Performance Merge function. This lists all Performances regardless if they are Multi-PART or Single-PART. You have to change the "Attribute" to Single-PART or only click on Green Performances to get to Single-PARTs. In Performance merge if the Performance is made up of 2 PARTs - then you can merge in both PARTs. The "source" will default to "All" - which means pull in all of the PARTs of a Multi-PART Performance. You can also tell Performance Merge to only pull in a single PART (even for Multi-PART Performances). For the most part, this will only pull in some fraction of what makes the instrument/Performance sound correctly - so you may end up with notes that do not sound (some range of the keyboard missing, some velocities, etc). Or - maybe your own custom user Performances you happen to know that PART 2 of some other Performance stands on its own and you want to copy just that PART to a new Performance - here is where selecting the source will make sense as well.

Another way to edit/modify PARTs in your Performance is to click on an existing non-empty PART. When you do this, the left-hand screen will show a menu and the top selection will be "Category Search". If you click on that - you will see a different menu which lists all of the Performances. Unlike Performance Merge - this feature only ever will replace a single PART with another single PART. The list still shows all single and multi-PART Performances. You still have to filter the list if you only want to see single-PART Performances. And, like Performance Merge, you can select the source PART to copy (for Multi-PART Performances) so you can pick something other than PART 1 in a multi-PART Performance. The difference is you cannot select "All" PARTs - this PART Category Search only lets you replace one PART with one other PART - not merging more than one PART (or entire Performance) with your current Performance.

So the system lets you add even fractured PARTs of other Performances if you want. The suggestion would be, until you get more familiar with how to deconstruct Performances and learn what is "inside" the Presets - to stick with single-PART Performances when replacing a single PART with another single PART.

So Performance Merge - yes, you would essentially be placing other Performances inside your existing one IF you select the source of "All". If your source is selected as anything else - then you are surgically selecting one PART of that Performance to place in your current Performance.

For Performance Category Search (used when selecting an existing used PART and selecting the "Category Search" touchscreen option) - this will only place one of the PARTs out of another Performance inside your existing Performance. If that source Performance happens to be single-PART already - then, yes, this is placing the entire Performance inside your existing Performance.

There are other details to consider when doing this:

1) The original PART you are pulling from and placing into your existing Performance may have had its own system or master effects. Your existing Performance may not have the same master or system effects. You, the programmer, are going to have to resolve this. If both Performances use different effects at each level - you may not be able to find the same effects to make both your existing Performance's PARTs and the new edited/merged PARTs sound as originally programmed. You'll have to decide how to sort this out. There are lots of options here (leave it alone, change to the 2nd Performance's effects, program new effects for all that compromise, turn off effects for the new PARTs by turning off the system effect sends, etc)

2) The original PART you are pulling from and placing into your existing Performance may have its own motion sequence, controller modulation, etc. that do not carry over 100%. I haven't sorted out how the latest firmware deals with this - but how the new PART is uprooted from a different Performance and placed into your existing one has to be considered. There may be incompatible controls that do something important for your existing PARTs - but would "break" the new PARTs somehow (do something undesirable). So these loose ends are something to consider.

For the most part (PART?), you will not run into any show-stoppers when it comes to loose ends. But every once and a while from shotgunning a PART from something else and placing it inside your existing Performance - you may run into something that doesn't sound right. That's when further digging would be necessary and likely will fall into one or both of the above considerations.

Thanks, I think I've got it! So now it makes more sense to me that the sounds that you choose from when assembling Performances don't have their own name per se... those things that come up are, in fact, Performances themselves, from which you can extract what you need. And the list of Performances is the same no matter how you get there (i.e. whether searching for Performances to play, or searching for Performances as a source for sounds to put into other Performances). Right?