It seems I'm always finding new parameters/settings that are often uncommon and not always easy to find when modifying certain existing Performance PARTs & I can't easily find what is set that's causing the sound or behavior to be different than it should be or would be if I loaded those Elements to a new Init PART.

Been there, done that, and the T-shirt is being printed. Bad Mister gets his T-shirts faster because he's a 'preferred' customer of 40+ years.

That usually happens to me when trying to use a preset but haven't taken the time to go through it thoroughly to try to look for things like you describe.

For me it was the 'Cutoff/Key' defaulting to +31% on the 'Filter -> Type' page where the default filter type is 'LPF12+HPF12'.

I eventually started cross comparing & checking settings of this PART against another PART that worked normally the way I'm used to, and I saw that the "Vel Cross Fade" on that Element was set to '4'. As soon as I set it to '0' it worked normally as I'm used to.

Page 95 of the Montage ref doc

Vel Cross Fade (Velocity Cross Fade)
Determines how gradually the volume of an Oscillator decreases in proportion to the distance of Velocity
changes outside the Velocity Limit setting.
Settings: 0 – 127
The higher the value, the more gradually the volume decreases.
0: No sound

A value of 4 seems low to make much of a noticeable change but some parameters use different (meaning 'unknown to us') curves - some are exponential.

But now knowing what the setting does, it could be useful in the future if I want to have one Element 'blend' into another Element in a good way (cross fade), instead of a complete switch over...

I just traded in my Modx for the M8X so have to do it all over again but what started doing was creating my own library of 'standard' elements, operators and parts.

Then I used my 'standard' versions when building new performances. I always knew what was in a standard element/operator/part because I had reviewed and/or set the values myself.

On the new M models you cna have up to 128 elements for an AWM2 part. So you could burn ONE PART and define 128 custom elements that do different things.

Then when you need a custom action you just add that one part to your performance and you can use SHIFT + EDIT to bring up the Copy / Exchange screen and copy the custom element out of your custom part and add it to whatever part and element you want to.

That guarantees it will be set up the way you want it.

There can still be 'gotchas' if you use a user waveform that isn't available anymore.

But in general creating you own set of 'default/standard' components prevents a great deal of those surprises.

[quotePost id=124067]
For me it was the 'Cutoff/Key' defaulting to +31% on the 'Filter -> Type' page where the default filter type is 'LPF12+HPF12'.
[/quotePost]
Yeah I can hear that one 😉

I've found various settings in a PART under Mod/Control settings whereby parameters were set to specific assign knobs and either via the PART & the position of the Assign Knob, or via the SuperKnob position, whereby I had to adjust those or in some cases remove the Assignment.

A value of 4 seems low to make much of a noticeable change but some parameters use different (meaning 'unknown to us') curves - some are exponential.

Yeah, and it was up quite high in velocity so it didn't drop too fast making it more difficult to understand why it sounded just a tiny bit less loud. I might actually set that back to a high number so that there is a bit of subtle blending (and also set the ...Hrd Element with a higher number as well)

Then I used my 'standard' versions when building new performances. I always knew what was in a standard element/operator/part because I had reviewed and/or set the values myself.

Smart idea and it will easily be ported into the Montage M.
I do something kinda similar but not as standardized like a template.

There can still be 'gotchas' if you use a user waveform that isn't available anymore.

Are you referring to waveforms from a Library or Custom waveforms in User memory?
My understanding was that all Preset waveforms are still there the same on any Montage, Montage M, MODX(+), and most/all from the Motif's (although I don't know if they would have the same MIDI listing number, or whatever identifies it)..!?

An uncommon setting that's been discussed previously are the note ranges and how the low note when higher than the high note will invert the note ranges such that you have a "hole" in the middle and two ranges from the bottom to the first MIDI note and then from the 2nd MIDI note to the top of the range. Basically defining a note range that would be the keys not available if the range was standard (left note a lower note than the right hand note in the range).

I've never really used it but it was a nice surprise when I learned about this.

[quotePost id=124069]An uncommon setting that's been discussed previously are the note ranges and how the low note when higher than the high note will invert the note ranges such that you have a "hole" in the middle and two ranges from the bottom to the first MIDI note and then from the 2nd MIDI note to the top of the range. Basically defining a note range that would be the keys not available if the range was standard (left note a lower note than the right hand note in the range).

I've never really used it but it was a nice surprise when I learned about this.[/quotePost]
Wow, you nailed it! Although in my case I figured it out because I needed it, not tried to figure it out because I didn't understand it (which is most of the time)!

All my piano Performances are custom. I don't touch a Preset unless for demo purposes, but when playing pianos, I always use my custom ones. They all have the EnvFollow Source of the Keyoff PART pointing to the PART beside it, which is just a piano PART, such that the Curve type is Dogleg (Uni) with a value of -26 so that the Keyoff doesn't jump out at you after 10+ seconds and has the right amount of volume for how long the notes have been held. You gave me the idea and how to configure it @Jason, which is so awesome. I would have never known about a dogleg curve otherwise... I don't care that it burns an extra PART, because the Keyoff is sooooo awesome and perfect!

And my string resonance is very custom based off the Epic Grand, as I use the string resonance sample from that Library, blended with the Resonance effect on the Montage, such that the effect amount is lowered considerably and you can't notice it, but the blend of the two together sounds quite natural.

However, getting back to my original point regarding the "hole" in inverted note ranges that you mentioned. One of the piano songs I play is "Faithfully" by Journey. https://www.youtube.com/watch?v=YalgKfrOdsI If you listen to it, the D#3 note is THE most important focal point note in the entire song (not to take away anything or importance of the other notes); However it just so happens that in the CFX sampled waveforms (whatever piano they sampled it from) that one particular note is a bit "dull" even at high velocities. Play all the notes around it and you'll easily hear how it doesn't have that bell-like snap to it like it should. So I had to replace that one note with a completely different Element in a 2nd PART. This is something that even with the Montage M's new 128 Element spec, I still couldn't do in a single PART and it requires a 2nd PART. Because the only way I found to add that note (PART 2) and remove it from the main PART 1 was to create a "hole" on the D#3 note in PART 1 by inverting the velocity ranges of that PART such that the Range shows as E3 to D3 (the Elements note limits are all C-2 to G5). And PART 2 (the D#3 note) just has a range of D#3 to D#3 with the PART Common Pitch 'Note Shift' set to -1 and all the Elements have the Tune 'Fine' set to +63. Essentially using the D note & stretching it up onto the D#3

The only other note out of the 88 that were sampled for the CFX that is not quite perfect, which I wish could also be changed is the F2 note. Compared to all the other notes, it's a bit tube-like sounding (square wave'ish). I've watched a few other videos with a CFX grand and that particular note is not tube-like on the other actual CFX grands... But all the other notes are right on par and sound perfect, so it must have just been the CFX that they used to sample the waveform velocities from!
If I could afford to trade up to a Montage M, I could fix that second F2 note as well, but it would require combining 3 PARTs into 1 and using the extra 2 PARTs to create another "hole" and note stretch to fix it... So for me it sucks not having the money to be able to get the Montage M8x and having to listen to others who have it (or can afford it) complaining way too much, when I would literally give my left nut to have one. I'm not even joking about that. If someone would hand me a Montage M8x I would donate my left testy...

Because the only way I found to add that note (PART 2) and remove it from the main PART 1 was to create a "hole" on the D#3 note in PART 1 by inverting the velocity ranges of that PART such that the Range shows as E3 to D3 (the Elements note limits are all C-2 to G5). And PART 2 (the D#3 note) just has a range of D#3 to D#3 with the PART Common Pitch 'Note Shift' set to -1 and all the Elements have the Tune 'Fine' set to +63. Essentially using the D note & stretching it up onto the D#3

You've highlighted one of the biggest issues (IMO) users face. Lack of detailed knowledge about things like sampling.

You said 'using the D note' and stretching it up. In a past thread you said you tried using the E note and stretching it down.

But the lurking question is whether those D and E notes are actually samples themselves. Or are they really stretched versions of other notes like C?

all the Elements have the Tune 'Fine' set to +63

In your older thread you mentioned using a fine of -64 for when you wanted to tune them down.

It might be easier to just turn unneeded elements OFF in your extra part rather than have them on at all.

I think you, and others, will find an interesting read in Bad Mister's comments about this Key Off issue in this old thread:
https://www.yamahasynth.com/forum/sample-robot-release-samples-volume#reply-95231

Another common, but 'hidden', issue BM describes there is in the Fingered Steel performance.
He describes how that performance has two DIFFERENT finger 'zings' based on velocity.

If you play a note REALLY hard you can hear the quick slide of a zing.

He talks about how many players of that performance never really play it in a way (i.e. with the needed, various, velocities) that highlights all of the programmed features like those zings.

Except none of those 'nuances', in fact NOTHING, about the performance is even documented. It is left up to every, single user to discover (or not) those things for themselves.

On the M models you can only see ONE element at a time out of a possible 128. You no longer have (at least I can't find it) even that row of 8 lights that tells you which elements are sounding when you hit a key or play a chord.

So I haven't figured out just HOW you tell if an element is actually sounding or not. I'll create a new thread to ask about that but so far BM hasn't replied to model M threads.

Just FYI, knowing how you said you love the C7. My M8X has both a one part, 16 element C7 performance and a one part, 11 element C7 Grand. I can't speak for how they compare to any previous versions.

Most pianos sample every key. As an acoustic piano manufacturer, Yamaha pays a lot of attention to acoustic pianos. Even if they do sample in a "character" key here and there.

You're able to tell what's a sample and what's stretched by using pitch offsets and seeing how far the ceiling is. If the sample is already stretched some then you'll have less chromatic notes before you hit the ceiling.

Although I generally agree with the sentiments you are expressing the details of that 'sampling' info are essentially unknown. You can probably rely on what you said for most use cases you might have but for outliers a lot of 'trial and error' will be needed.

Most pianos sample every key.

The reality is that one has no way of knowing exactly WHAT was sampled or how.

A 'waveform' can contain multiple keybanks with each keybank having different constraints such as velocity and note ranges. The details of the preset waveforms (and keybanksk) are NOT publicly documented.

So we have no way of knowing just what keys were sampled and how many different velocities they were sample at.

For OPs example what works for one velocity/range may not work as well for a different velocity/range.

As an acoustic piano manufacturer, Yamaha pays a lot of attention to acoustic pianos. Even if they do sample in a "character" key here and there.

I agree - but they don't actually publish the details of what they put into various instruments. Any given 'waveform' may have different internals for different instruments. Some waveforms may include/exclude certain keybanks or fewer actual samples that are then 'stretched' when needed.

You're able to tell what's a sample and what's stretched by using pitch offsets and seeing how far the ceiling is. If the sample is already stretched some then you'll have less chromatic notes before you hit the ceiling.

You can certainly make that effor if you want. But Yamaha already has that info anc chould make it public if they chose to.

I don't disagree with the basic intent of what you said - just reiterating that the burden is on the user to try to figure out what was done for those use cases where more detailed info is needed.

[quotePost id=124077]
You said 'using the D note' and stretching it up. In a past thread you said you tried using the E note and stretching it down.
But the lurking question is whether those D and E notes are actually samples themselves. Or are they really stretched versions of other notes like C?
In your older thread you mentioned using a fine of -64 for when you wanted to tune them down.[/quotePost]
You are extremely observant as that was how I originally tweaked the D#3 note, stretching the E note down; However I eventually tried stretching the D note up and it sounds even better, so yes, I changed which way I stretched it per PART and then per Element! 😉
Also, I did listen to the C#3 & D#3 notes to see if the D is already a stretched note and I don't believe it is...definitely not stretched down from D#3, as that note is not that great and the whole reason I needed to replace it. The new D#3 note stretched from the D3 of the waveform doesn't sound stretched and so much better than what is there in the Presets. Also the E3 note doesn't appear to be a stretched note either (definitely not stretched up from the D#3), and most likely not from the F3 note, as the stretching is usually up, as stretching down can lose more quality...
When I used SampleRobot to sample my Synthogy Ivory C7 piano library I got really good at telling which were the best notes and how they sounded when stretched. In general I did about every 2nd note so only single note stretching and you can tell which ones are pairs as they both have the same core, bell-like and overtone sounds to them...
I also on a few choice notes (and all the layers of those notes), used Wavosaur or Audacity along with SampleRobot, whereby after sampling all the specific best notes via SampleRobot, I took did a reverse stretch in a sense, because stretching always from the lower note sample to the blank note with not sample. So for instance there was a C note that was excellent, but the B note below it wasn't great and the the A# note was better than the B note, but not nearly as good as the C note. So I took the individual C note that SampleRobot captured, then using Wavosaur or Audacity I stretched it down to a B & saved it (which still sounded better than the A# note stretched up). I then removed the C note from Sample Robot and imported the new B note that I had saved. So now the B note sounds better and it stretches back up to C, which sounds identical to the original C does since it was created (stretched down) from C to begin with...
So essentially in the end, I fixed up any choice notes from the original sampled Synthogy Ivory C7 VST, that weren't as good as they could be, and made that piano sound even better across the entire range of notes than it does when played on the VST..! 😉

It might be easier to just turn unneeded elements OFF in your extra part rather than have them on at all.

If you are referring to the extra PART that I have for playing just the D#3 note, I need all 8 Elements, because I have it set for 8 different velocity layers of that one note...

On the M models you can only see ONE element at a time out of a possible 128. You no longer have (at least I can't find it) even that row of 8 lights that tells you which elements are sounding when you hit a key or play a chord.
So I haven't figured out just HOW you tell if an element is actually sounding or not. I'll create a new thread to ask about that but so far BM hasn't replied to model M threads.

Oh wow, that's not good at all. 😮
I depend on those lights to ensure I know what Element I'm hearing, especially when programming velocity layers of a piano. To get it sounding seamless when switching between two different Elements, I need to be sure that I can see the switch point between the layers and not hear any difference when the higher velocity is at it's lowest vs the lower velocity at it's highest point...they should sound exactly the same thereby making it a seamless switch. I gotta have those Element lights!! Unless there's a way to see it via the new top left screen, perhaps!?
Anyway, I so wish I had THAT problem to deal with & solve!!! 🙁

Just FYI, knowing how you said you love the C7. My M8X has both a one part, 16 element C7 performance and a one part, 11 element C7 Grand. I can't speak for how they compare to any previous versions.

It sounds like the first C7 piano is an 8 velocity layer C7 and the other is a 5 velocity layer C7..!?
I'm only guessing, but it would make sense that the first C7 has 8 layers of Element waveforms for the lower-mid 3/4+ keys of the piano, 7 waveform layers for the highest notes of the piano and 1 for the keyoff waveform (for a total of 16 Element waveforms)..!? And I would guess that the 5 layer C7 has 5 for the lower-mid keys, 5 for the highest keys, and 1 for the keyoff for a total of 11..!? Same as the CFX Stage piano on the Montage, except that the highest keys only use 2 layers instead of 5. But that makes sense for the new AWM2 128 Element enhancement!! 😉

[quotePost id=124083]The reality is that one has no way of knowing exactly WHAT was sampled or how.
So we have no way of knowing just what keys were sampled and how many different velocities they were sample at.
[/quotePost]
The one thing we can know for certain is if there are any stretched keys side by side, as you can use SampleRobot to sample the notes of a piano, remove the note above it making it blank, and check the stretched note against the original.
So if for example you wanted to check a C# note to see if it is a stretched note or not, you can just sample the C note below it via SampleRobot for a particular CFX waveform layer, ensure that the C# note is blank in SampleRobot, allowing the C# note to be stretched up from C, and then compare the stretched note to the original C# note. If the stretched note sounds exactly like the same note from the original waveform, then the original is a stretched note...

Here's another Uncommon or hidden settings/parameters that I encountered recently. I loaded one of Manny Fernandez library ('MannyFMXplr') and tried to remove all the ARP and Motion Sequencing of PART1 of the Performance named "MFMX Krazy 8 HarmSwp" but had a hard time figuring out where the extra motion was coming from after I disabled both ARP and Motion Sequencing. I removed all effects, all Mod / Control settings, but it was still there.

Came to find out that it was happening within one of the carrier operators (#8) whereby the Time Attack was set to 49, Decay to 53, and the Level Initial set to +20 & Attack set to +41 under Form / Freq. I hadn't encountered any FM-X PART that utilized these settings under Form / Freq before... Such a great way to change a sound in a Motion Sequencing or ARP-like kind of way, but right within an Operator of the PART itself..!! 😉

For some use cases you can 'boost' a problem key by adding another element that plays just that key.

So the original key would stil sound like it does. But the new element would add it's own tone to the mix by, maybe, detuning a bit one way or the other.

The irony, of course, is that the 'tricks' you add to get what YOU need become someone else's 'Uncommon or hidden settings/parameters'. What goes around comes around - LOL!

Yes, it is a catch 22 sometimes, as there is so much unknown and untapped potential in the Montage(M)/MODX(+) synths, that most never use half of it. It's so feature filled, with so many parameters, and so powerful with an extreme potential of depth, such that it sometimes makes it difficult to reverse engineer and figure out how/where some configurations were set in a Performance.

In the MONTAGE/MODX/MODX+ (first generation Motion Control Synthesis Engine) a programmer could break the usual 8-Element per AWM2 Part barrier by linking multiple Parts to create one instrument with lots of nuance.

The “CFX Concert” is one such example with its 18 Elements. This acoustic piano used 4 Parts (linked by the KBD CTRL feature) to accomplish this goal.

In the next generation, MONTAGE M, the architecture has been changed so that a programmer can have virtually unlimited Elements to construct a single instrument all within a single AWM2 Part (using a single MIDI channel). I say “virtually unlimited” but it is actually 128 maximum. (When you get into programming you’ll likely run out of ideas before you run out of Elements… anyway, it’ll keep the programmers busy).

If/when you call up one of these “Extended Element” sounds the HOME screen will indicate the number of Elements inside this Part. The “CFX Concert” will indicate “Elements: 18” on the top line of the HOME screen (along with an activity meter).

If/when you recall a traditional 8 Element AWM2 Part, this ExElement count is replaced with 8 radio buttons indicating which of the 8 Elements is active at any moment. If you recall “CFX Stage” which is an 8 Element AWM2 Part, you will see the activity of each Element (You can see Element #8 is the KeyOff Element.

This is also true for any AWM2 Part in your Performance — simply *select* it.

Select the Part —> EDIT —> “General/Pitch” —> “Element Settings”
Currently, the “Ex Elem Sw”, which allows you to increase the number of Elements (Element Count) as you require, the 8 radio buttons that show activity disappear and the number of Elements will be shown along with an overall activity meter.

Alternatively: From the HOME screen select a Part —> then using Knob 6 under Main Display set “View Mode” = “Part Info”. Here you can see the first 8 Element Switches, Element Levels, Slider position and a radio button indicating which Element is currently sounding.

For the many users who found it a problem to manage the multiple-Part, single-instrument Performances, now can address the very nuanced “CFX Concert” on a single MIDI Channel (instead of four), from a single slot (instead of four).

You can now freely decide what Parts are set to receive on what channels…
Your KBD CTRL Parts can be set to Receive (all on the basic selected “MIDI I/O” Channel) — and all non-KBD CTRL Parts can be set to “Tx/Rx Ch” on any channel of your choosing.

Final thoughts: But it’s the potential here that is really exciting. You can setup an Element Group that is set to XA CONTROL “Cycle” or “Random”… used on things like repeated notes this can breath new life into sounds that typically suffer from always being exactly the same when played over and over in a phrase. New possibilities for “Legato”… different combinations of Elements via AssignSwitches… etc., etc., this could be fun!

The key system (M8x) has the ability to sense the position of key at all times — this makes for a monster of a nuanced controller…. I’m looking forward to some really exciting developments in the way of sounds and the control over their behavior.

CFX Stage in the Montage classic uses 8 CFX waveform elements and all 8 sample each key including key off.

Takes less than a minute per element to determine including adjusting limits so all velocities and all notes are available to test.

Set pitch bend (up) to +24, hold the pitch bend all the way up which stretches the played note by two octaves, and play a chromatic scale listening for repeated notes.

The CF3 waveforms are mostly not sampling every key although there are a few that do like CF3 Flat Soft St-

S6 waveforms are mostly not sampling every key although there are outliers like "S6 Stretch ff Mn-" that do .

The S700 waveforms sample every key.

And so on.