I haven't tried using a motion sequence for this, but I wonder if the release level can be controlled as a destination parameter which is reset on every 1st keypress so your chords will hold then release when a new chord is pressed (in your pad). I'm learning as I go attempting to do what you want. I'm certain that the tools available for Montage are completely different than Tyros.

I could do a motion sequence, but I'll do superknob this time since I haven't done much superknob before.

I'm going to just work with one performance/part - and I'm sure you can modify to fit your needs.

1) Select Init Normal (AWM2) Init voice [CATEGORY SEARCH], then bank/[SOLO], then [ENTER]
Note: my starting point, but you'll already have your pad PART in mind, not Init Normal (AWM2)
2) Edit superknob: press [PERFORMANCE] (HOME), touchscreen (TS): "Motion Control" -> "Knob Auto"
Note: under "Super Knob" menu, the Init Normal (AWM2) performance already has all assignable "global" knobs set at Full Counterclockwise=0, Full Clockwise=127 so I do not edit that.
3) From "Knob Auto" screen
a) TS (touchscreen): "MS Master" touch until on (highlighted green)
b) TS: "Super Knob MS" touch until on (highlighted green)
Note: superknob should start automating now
c) TS: "Key On Reset" change to "1st-On"
d) TS: "Cycle" - can use DATA DIAL to scale back to 1 from default of 16. Just want one cycle here
e) TS: "Loop" - press until off (NOT highlighted green)
Note: at this stage, press a chord and see the superknob respond. Press another chord. You can see superknob is doing something every chord press which, in theory, looks like the right thing to do.

At this point, I'm not sure where I'm going to end up since I'm not sure 100% my idea in my head will work, but I'll continue ...

I'll come back to this screen later to edit the pulse - but now I'll move on over to the destinations
4) press [PERFORMANCE] (HOME), then [EDIT]
5) TS: Press "Auto Select" until ON (highlighted green)
6) Turn the assignable knob 1 which is located directly above the left-most fader
7) TS: Press the "[+]" which adds a destination, then touch the default destination ("InsA Param 1") and change it to "Part 1 > Part 1 Assign 1", then press [ENTER]
Note: doing this step #7 just "ties a string" around the global assignable knob 1 to part 1's assignable knob so superknob can drill down to part 1.

8) Select part 1: press [PERFORMANCE] (HOME), then Number A [1]
9) press [EDIT]
10) TS: Common, "Mod/Control" -> "Control Assign"
Note: this brings up a screen where we can map part 1's assignable knob to a final destination
11) TS: "Auto Select" should still be ON (hilighted green), again turn the assignable knob 1 (see step 6 for location)
12) TS: Press the "[+]" which adds a destination, then touch the default destination ("InsA Param 1") and change it to "Element > AEG Release", then press [ENTER]
13) Setup the destination's control curve:
a) TS: "Polarity" change to "Bi"
b) TS: "Ratio" set to "+63"
c) TS: "Curve Type" set to user (pick a user curve not used)
d) TS: "Edit User Curve"
e) TS: Use faders to set OUTPUTs (#2-#8) to 127 or double-press output value and enter 127 manually. Use fader or use manual method to ensure OUTPUT #1 is set to 0
f) TS: Set "CurveType" to "Step" (which should be highlighted green)
g) TS: Change INPUT #2 to "3" (I believe the default is 18). The reason I do this is this timing sets how long you have to hold the chord before it "hooks in" to a sustained chord and also sets the duration of placing the last chord's notes to zero amplitude. You can use smaller or larger values here and see the results. Very staccato playing on pads doesn't seem to make much sense - but if you are doing that - maybe shorter is better. "Speed" will also be another way to control this. You have options.
h) press [EXIT]

This should be enough to demonstrate the basic idea - works over here.

Play a cluster on one area of the keyboard (lots of minor 2nd intervals) - This should ring out like the uncontrolled sustain problem you were concerned about.
Now play any single note - the single note now rings out and the cluster goes away.

I think this is what you're after.

Modify to fit your particular pad - as I suppose there may be multiple PARTs to the pad you use - in which case you can assign more than one destination in step #7 and recycle the remaining steps to apply to the second PART.

By the way, depending on your speaker response to a sudden 127-to-0 amplitude step function, you may want to modify the user curve to something else to keep the speakers from "clicking".

Maybe (step 13-e) a linear curve (NOT step) from {input #1=0, output #1=55} to {input #2=3, output #2=0} to {input #3=5, output #3=127} . This, for my studio speakers, gets rid of the "click".

I was going for something quick-and-dirty as a jumping point for your own creativity - but thought others may appreciate some better sonic design to keep the speakers from producing artifacts.

Also note, as a performance note, if you want to STOP your pad from sustaining - you're going to be asking yourself "how do I stop this train once it starts?" The answer is press the [MOTION SEQ ON/OFF] button so it is OFF (not illuminated). For this idea, the [MOTION SEQ ON/OFF] button can be thought of as an "Automatic Sustain Pedal On/Off" button.

And technical detail behind what pressing the button (off) does "under the hood":
Turning motion sequence OFF resets the pulse curve (controlling superknob) output to the "origin" value. If the pulse curve is Uni-Polar, then the superknob will be set to 0. If the pulse curve is Bi-Polar, then the superknob will be set to 64. Since I used defaults, I expect the pulse curve (controlling superknob) to be uni-polar. If not uni-polar, then the knob will not reset correctly when you press the [MOTION SEQ ON/OFF] button for this application.

Wow, thank you very much for this one Jason! I did what you said a few times in a row, step by step, but every time I got the same result - long, sustained tones which do not get re-trigerred, and become a mess, no matter is it a full chord or a single tone. Obviously I am doing something wrong, but can't figure out what exactly. Even the sustain does not stop when I turn MS off, despite I can see the knob moving every time, so..it was a long night, and definitely a long way in front of me for this hold/latch stuff.

Every time you hit a key, you should see superknob automate from far-counterclockwise, or position 0 (around "7 o'clock") to position 32 (around "10 o'clock").

I'm not sure if you started with the Init AWM2 performance or your pad. If starting with the pad, it's worth checking the Element 1-8 "Amp" -> "Amp EG" setting "Release" and set it to 50 to match the AWM2 performance value. This should not matter as much if you use the initial step-curve, but will come into play more if you use the linear curve later suggested for step 7.

Make sure [MOTION SEQ HOLD] is not illuminated brightly (can be dim). Press a few times to be sure.

I'm going to retrace my own steps - following the directions and see if there's anything I left off.

Step 3b - "Note:" Slight clarification: Superknob will start automation when you press a keyboard key.

Clarification: Just before step 4. "I'll come back to this screen later to edit the pulse - but now I'll move on over to the destinations" - this was not necessary although there is a chance for optimizing the pulse curve. I just let the standard pulse curve apply which is default for Init Normal (AWM2).

After step 4 (pressing home, then edit) - you may "end up" in a different screen than was assumed. please add:

4a) TS: touch "Control" -> "Control Assign"

NOTE: be sure on step 4 to follow this literally. Press (HOME) button then [EDIT]. No other touches or selecting the part inbetween.

...

Step 8 - you can also press [PERFORMANCE] (HOME) then TS: touch the part #1 area which should highlight the 1st (and only for Init Normal) part. It's easier to press the Number A [1] button which is the 1st left-most button under the silkscreen "PART SELECT". The button also has the silkscreen "1/1" under it.

After step 10 - not sure if you start from the Init AWM2 performance or not. I would suggest you do - but if not, be sure to clear out any destinations for assignable knob #1. Do this by TS: Touch Destination # (#=number 1-16), then touch "Delete" for all destinations of the AsgnKnob 1.

I followed the steps as shown and ended up with the correct behavior. The clarifications are more informational and do not modify the original sequence of steps.

I would start with the Init Normal (AWM2) performance first (not your pad or existing performance) - so you can see it working if not done already.

The last additional step I would add is to ignore sustain - since we are using auto-sustain. I did leave off handling the sustain pedal for the new PART with auto-sustain.

Add:

13) Press [PERFORMANCE] (HOME), then Number A [1] (to select part 1) then [EDIT]
NOTE: things in brackets "[]", when instructed to press, are ALWAYS buttons - not touchscreen. This matches Yamaha documentation convention.
14) TS: press "Mod / Control" -> "Receive SW", from this screen look for "Sustain" and touch until it shows "OFF" (is NOT highlighted green)

That should handle the auto-sustain response to the sustain pedal (ignore it).

Previously - you had to be sure the sustain pedal was NOT pressed or the notes would indeed run together.

That is about all I can do to help get you on the right track at the moment.

"3rd time's a charm" (?)

Finally managed to sit in front of my Montage properly, and try this again. I did everything from scratch, and - nothing. The only usable sound is the one from first pressed key - sustain stays when I press the second, the third and so on. Simply said - unusable mess. Definitely, I am doing something wrong despite re-checking everything at least hundred times.

''Every time you hit a key, you should see superknob automate from far-counterclockwise, or position 0 (around "7 o'clock") to position 32 (around "10 o'clock").''

Yes, that super knob light moving is happening. The thing which is not happening is that after I press second key, the sustain from previous key note does not stop, it continues.

Update - I finally did it, and it works, thank you Jason! And, it works smooth when this is assigned to the pad on part 1. When I put other instrument (in my case, lead) in the part 2 place, and make the split point, the super knob and this whole automatization gets triggered by the part 2 also, which I want to exclude, because, triggering the super knob it messes the whole thing on the part 1 (pad). Hope you can understand what I am thalking about.

I understand what you're saying. I outlined the PAD to show how that could be done. Glad you finally made it through. Now we got the PAD covered, here's the rest:

Using superknob, I realize I painted myself in a box. There's no limit on the notes (range for causing the reset) since superknob looks at the full keyboard when retriggering. The only option here is velocity - so to set that up, you'll need to setup an ARP to refactor your playing for the PAD to be set at a specific velocity (velocity=1) then adjust the offset so bring it back up. The trigger works by looking at the "original" velocity - not the offset velocity. So you can play tricks here assuming you are OK having your pad fixed at one velocity. It allows for the lead to respond to velocity changes - although you could switch this around as well.

Assuming the PAD at a fixed velocity is OK - have the ARP assigned to the PAD refactor your playing to velocity=1. This is similar to instructions on how to get velocity-mapped arps to work (like to play pick scrape noises on a guitar, etc). Go into your PAD PART and set the velocity offset to a value that makes the PAD sound fine to you.

When I say refactor your playing to velocity=1, I mean have the arpeggiator play your notes for this entire part at a fixed velocity. This is covered somewhat in the following article:

https://yamahasynth.com/resources/the-4-track-arpeggio-montage

For the lead PART, go into it and set the velocity range to 2-127 (instead of 1-127). There's no ARP needed here for the lead PART since we're not going to force the velocity to be fixed (just going to make sure it's not velocity=1). Although if you actually DO play a note (in the split range for the lead) that is struck so softly that it is actually velocity=1, the note will not sound (be "gobbled up" since we ignore velocity=1 by means of the range). This shouldn't be a problem to filter out velocity=1 and allows for this PART to retain velocity control if you want that.

Last, go into the super knob settings (motion control -> knob auto) and change the velocity limit to lower limit = 1 and upper limit = 1. This makes the reset-on-key only occur for the PAD notes.

It's kind of unfortunate that super knob doesn't seem to have an easier way to limit the keyboard range it looks at for note resets. In order to avoid having you generate an ARP to fix velocity, I suppose I could switch over to motion sequence instead of super knob. I really gave myself the challenge of doing this with super knob automation because I had relatively less super knob (w/automation) based performances I've worked before.

Last, I'm not sure if fixed velocity ARPs will be offset by the velocity offset when they are played by Montage. I would think they would be - but if not - this situation could be fixed by using a 3rd part which just covers the velocity=1 key reset. In this case, both parts 1 and 2 would have a keyboard velocity range of 2-127 and the 3rd part would have a keyboard velocity range of 1-127 with the "fixed velocity ARP" (velocity refactoring) on PART 3. This would be the source trigger that always generates velocity=1 notes. And this split (note range) would match your PAD note range. The 3rd part could be any sound - you would route the output (routing) for part 3 to OFF so it doesn't sound out your monitors/speakers.

Edit: also note that if we were going to have to make arps - then an arp itself is another way to get there.

An arp with a longer gate time than the longest pad you would ever play is essentially like holding the sustain pedal down. Set the arp on HOLD (gated in advanced arpeggio settings) and this will reset the originally held chord after moving. When you press a note or some combination within the split with the arpeggio + PAD, the arpeggio will actually play a very long note as if you were holding the keys down. Setting the gate time to be many many measures will accomplish the same goal. Gate time is the length the arpeggiator plays a given note after that note is triggered (either because it's the first note of the arpeggio, triggered by a keypress - or the second note of the arpeggio triggered by the elapsing of the 1st note's gate time).

Since the arpeggiator will be holding down the keys - the important amplitude envelope parameter will be decay (don't decay - levels should remain the same as the general level of the element). And release can be the default release (release return to zero level) since the arpeggiator version does not need to use a trick to keep playing when you release a key. When the arp is playing the notes - it's holding them down. Previously, when you let go of the keys - the fact that the release was kept at a listening level was the trick to keep playing the note when you let go (and nothing, not even arp or anything else was holding down a key).

Using an arpeggio falls more "in line" with an "up the middle" solution. I'll still give a motion sequence version. So there should be at least three ways to arrive at the destination.

I too wanted to see how to simply hold any notes I'm playing... and without the need to hold down a sustain pedal btw, whether its a slow evolving pad or a series of arps. I've been stuck scratching my head on this. It's a very typical thing players might want to do quickly and easily during a performance. Turn to the Montage... play a chord and hold it, pick up a guitar and then play a melody, just for example. Many synths have a HOLD button or a KEY HOLD or a LATCH button that handles this very common function and doesn't have to be programmed on a per patch basis... I'm honestly surprised how something so simple can be so complicated on the Montage... unless I'm totally missing something...

What would be ideal is a way to globally set one of the hardware buttons, say the ASSIGN 1 button, to act as a HOLD button that works with any performance without doing a deep dive into programming the Super Knob or messing with the Arps. Also I personally don't need it to auto-stop previous notes I'm holding when I play new notes... I'd be fine with just manually toggling a HOLD button for each chord change I might want to hold.

Overall I really love my Montage... it's pretty amazing... just wishing for a simple solution for this simple problem.

I assume you can assign a controller ass switch or an extra latch footswtch to trigger release parameter of a specific part.

Using a latch means you have to turn it off and back on every chord change to get what the OP wanted.

Since the thread was bumped - I was reminded of some unfinished business and did accomplish auto-sustain with one split and traditional sustain pedal sustain with the other side of the split. Pressing the "traditional" side notes did not cut off the sustained side notes like the super knob version does.

What I did was basically the same idea, but used motion sequence on the auto-sustain part. For SYNC (motion sequence) I used ARP since I needed a way to restrict the trigger for motion sequence to a region of the keyboard. One of the few ways to do this is to use an ARP which has its own key note range. I wish motion sequence had the same control where I could limit a range at least for the trigger. That's perhaps a niche feature - so the priority is relatively low for that but with each part possibly having its own motion sequence - and because there are plenty of parallels with arps - I do see some rationale for adding in future generations.

Back to what I did: I had to use an ARP to generate the motion sequence input values. The way this works is that the downbeat of the ARP measure becomes the clock. So my goal was two fold - 1) use a fast tempo so measures would "fly by" and 2) Find a short arp since there are some behaviors of the arps completing their cycle - so i didn't want to have to hold the chord down for too long before the next cycle could be re-armed.

I found the MA_Piano_Latin was fine for this.

Also, because I'm only using ARP to trigger MS - I made the key mode DIRECT to avoid having the ARP play the keys. arp velocity mode is THRU so the ARP is decoupled from the notes played both in sense of time (using DIRECT) and in sense of velocity (THRU). I set the tempo to 300, arp unit multiply to 50%.

I'll have to write it up if you're interested in something more 1,2,3.

Chris,

My original goal was not to use ARPs at all. Although those should get to the same destination as well. An ARP with an infinite gate time should be all that is needed along with just one note in the ARP so it is ready to be rearmed immediately. Maybe something like this already exists in the preset ARPs - but the search is limited (I cannot search for beat/measure count, can't sort by longest gate time) so finding exactly what I want, if it exists, has some limitations. I do not yet have custom ARP ability on-board (that's a whole other self-imposed limitation - no need for discourse here) so my later solution was to use any common ARP to work around what I see as a limitation in the triggering paradigm (all or nothing in terms of keys which cause a retrigger).

As far as stopping this feature - that was already built-in to the original solution. You may have missed:

Also note, as a performance note, if you want to STOP your pad from sustaining - you're going to be asking yourself "how do I stop this train once it starts?" The answer is press the [MOTION SEQ ON/OFF] button so it is OFF (not illuminated). For this idea, the [MOTION SEQ ON/OFF] button can be thought of as an "Automatic Sustain Pedal On/Off" button.

And technical detail behind what pressing the button (off) does "under the hood":
Turning motion sequence OFF resets the pulse curve (controlling superknob) output to the "origin" value. If the pulse curve is Uni-Polar, then the superknob will be set to 0. If the pulse curve is Bi-Polar, then the superknob will be set to 64. Since I used defaults, I expect the pulse curve (controlling superknob) to be uni-polar. If not uni-polar, then the knob will not reset correctly when you press the [MOTION SEQ ON/OFF] button for this application.

Certainly no need to turn off Montage since changing to a new PERFORMANCE would clear out the entire controls which cause the auto-sustain.

Oh - it may help to add links or plugin names (author, download name) or some "hook" that would allow for review and following better.

Surely making a plugin would be easy to do this. You would filter out note off - store the note on - then manually inject a note off of the previous note on list when a new note comes in (at a different time). Then I suppose, to "stop the train" (turn off the auto-sustain), you could use a specific note value or even assign 1 or assign 2 since they send MIDI CC.

If the plugin allows for on/off control with a CC - use 0x56 or 0x57. When ON - the value is 0x7F and when OFF the value is 0x00.

This thread covers a sustain for a part that is not triggered by the sustain pedal and allows the sustain pedal to still control other parts. So it meets completely different requirements. This is where the complexity comes in.

For your request, assign the switch CC to sustain's CC. You can follow the steps outlined as a kludge workaround 4 sources not able to reassigned along with their destinations for more than two destinations. Check out the following link for more info
https://yamahasynth.com/forum/about-motion-sequence-automation,-only-the-super-knob-can-be-automated-directly#reply-14796