Difficult to diagnose some random user Performance with minimal information. You haven't communicated which PART is "popping" - maybe you don't know. To figure this out - leave all the PARTs active so they will create the artifact. Just mute each PART one by one until you narrow down which PART is "popping". Maybe it's the drums - but I can't tell from what you've communicated.

The issue with user-created Performances is that there are lots of possible pitfalls when assembling your PARTs to create your own Performance.

I did some additional testing today to see how the performance reacted to muting the individual parts. It appears that when part 4 is muted, the drums are not clipped. If the drums are muted (part 5), no problem either. However, I want a drum track, so it looks like I have to take a loss on part 4 (which I liked the sound of). I replaced part 4 with a different part and so far it seems pretty solid.

What is mainly happening here is that the drum part is being clipped occasionally when all 5 original parts are turned on. This results what I describe as a "popping" noise.

Btw - I'm not even sure what Part 4 was. The name of it, "Keys Electronic," is too vague to find in a category search.

I guess that it takes of bit of experimentation to find out which parts can work together. It's seems like what is happening is that the keyboard is reaching some limitation for how many sounds it is able to produce together. I'm kind of surprised at this, since in theory I should be able to include 8 parts in a performance.

It is a bit frustrating to put together a mixture of sounds (Parts) that I like, to find out that the parts are not totally compatible.

I did find a section in the owners manual (page 60) regarding sound distortion and cutoff. It's pretty vague though.

I'd be interested in finding out if Yamaha has any guidelines for adding Parts to a performance, or if it's just a matter of trying it to see if it works. It would be nice to know how to avoid the "pitfalls."

Btw - I'm not even sure what Part 4 was. The name of it, "Keys Electronic," is too vague to find in a category search.

You can be sure, the box to the left of the PARTS that reads "TYPE/NAME" - if you touch that box the NAME of the Electronic Key sound will appear (if it has a name) "Keys Electronic" is the TYPE, touching the box will toggle the function to reveal the Name of the programmer gave it a name.

I'd be interested in finding out if Yamaha has any guidelines for adding Parts to a performance, or if it's just a matter of trying it to see if it works. It would be nice to know how to avoid the "pitfalls."

It is possible to overload any bus with signal. In general you have plenty of headroom in the MONTAGE; for you to have used it all with just four or five PARTS means you have some settings that are, shall we say, a bit abnormal. Before we go hunting those down, lets start with a basic setting:
For example, at what OUTPUT level are you operating?

Press [UTILITY]
Touch "Settings" > "Audio I/O"
Under "OUTPUT" what are your settings?

Popping and clicking does not have to be overloading the internal bus. It could be coming from whatever you have your MONTAGE plugged into (you do not mention what the sound system consists of...). We can quickly - by process of elimination - remove doubt. Plug a pair of good headphones into the PHONES jack of the MONTAGE and see if you still hear the clicks and pops when you are monitoring the MONTAGE in the phones. If yes, you have an issue, If not, look elsewhere for the problem. Let us know.

If you are overloading your external sound system (which is a possibility you must consider), you can adjust the Output level of the MONTAGE to match what your system can handle. The MONTAGE is conservatively set as far as levels go, so it is quite unusual, to say the least, that you would be overloading the internal system with just four or five PARTS...

The original message has a screenshot of the NAME which is "KEYBOARD" for that PART. Which looks like a not-the-first-PART of some Performance. Generally, anything that's not the 1st PART of a mult-PART instrument is going to have a generic name in all CAPS. The 1st PART (the "anchor" ) is going to have the name of the Performance.

In this case - unless you drastically changed things -

PART1 = Anchor PART of "Full AP/EP/Pd/Str" preset Performance. Which is a CF3 Acoustic Piano
PART2 = "Finger Bass Amp AF" which replaced the original preset Performance's PART 2 (which WAS, incidentally, the Pad choir-like PART made up of Waveform "Fat Saw" )
PART3 = "STRINGS" which, I presume, is the original "Full AP/EP/Pd/Str" String part which uses the "OrchStrgs" Waveform (unless you hand edited)
PART4 = "KEYBOARD" which, I presume, is the original "Full AP/EP/Pd/Str" Electric Piano part which uses the "Digi" "DX" and "Bell" Waveforms. (unless you hand edited)
PART5 = Looks like you added the BeechwoodSnare Drum Kit to this PART. There's no "BeechwoodSnareKit2" (not the number 2) - so presumably you've made some edits to this and made it your own.

That's about all of humpty dumpty I can put together. Since it's a user Performance - there's no telling what's been edited.

Volumes aren't crazy.

PART1 = 68
PART2 = 31
PART3 = 62
PART4 = 74
PART5 = 80

Not all 127 - about 50% headroom. The drums are "hot" (relatively speaking) - but not wild.

This is just the pre-master. BM's question about the outputs are in a different gain stage (final boost/trim/unity on each output)

Note: when I pretend like I made your Performance as close as I think would be how you did it (Start with "Full AP/EP/Pd/Str", delete PART 2, [+] add PART 2 as the ""Finger Bass Amp AF", [+] add PART 5 as "BeechwoodSnareKit", turn ARP OFF for the bass=PART2, turn ARP HOLD ON PART5=drums, set note range for Bass upper limit to F#2 to form the split) - one thing "jumps out" as a possible residual issue of plopping (meaning retrofit) in the bass.

What I notice is that when superknob is full clockwise (also/or scenes 5-7. 7 is the worst offender, followed by 6, and least is 5), the bass (PART 2) really goes "crazy" - at least loud and a little "pop"y. You didn't mention anything about PART2 - so maybe you've already adjusted this. And maybe your PART4 (the EP) is not the same as the default "Full AP/EP/Pd/Str" part -- which is the only place I can leave mine without more info.

Here - I never hear anything cut out, drop out, clip, etc. What I mention about "pop" is kind of an aggressive attack on the bass that doesn't match the rest of the sounds that are relatively calm.

I have my outputs (utility -> audio i/o) set to +0dB across the board.

Thanks guys for spending the time and the homework for helping me troubleshoot this issue. I am going to work to put together some additional information and also try recreating the performance to see if I can replicate the issue in another performance.

I'll report back hopefully within a day or so, but want to take some time to provide you with a package of useful and accurate information and look over the information that you have provided.

I recreated the Performance this evening and also recorded a sample so that you can also hear what I am describing. Also, I kept everything as basic as possible, without doing anything fancy. Below describes the setup:

1. The sound system consists of a pair of Yamaha HS50M amplified speakers.
2. Playing through of pair of DT 770 Pro headphones produces the same issue.
3. Please see screen print of output settings (images 11 and 12).

Setup for the recording:

1. Select the "Full AP/EP/Pd/Str" performance (in this case I kept the part 2 "Pad/Choir" (which I omitted in the original) just to keep the factory performance as is.)
2. I added part 5 "Finger Bass Amp AF" and set the key range from C-2 to F#2, and set the pitch up an octave. Also, I turned the volume down a quite a bit.
3. I added part 6 "BeechwoodSnareKit2" which I believe may be an out of the box arpeggio (see attached screen prints). However, I also tried applying the "BeechwoodSnareKit" in its place and found it to be the same issue. Note: Arp 1 is being used in the recording.

That's about it. The SuperKnob is set at about 1 o'clock.

So, in this case, I'm using 6 Parts. If I missed any details, please let me know.

Below is a sample recording of the issue.

Maybe the BeechwoodSnareKit2 is a "drop-in" for the "Power Standard Kit 1". The datalist doesn't spell out the drum key usage for BeechwoodSnareKit2. At any rate, the MC_12/8 Ballad ARP is said to be built for "Power Standard Kit 1".

The ARP was originally "designed" for 60bpm. Your tempo shows as 100bpm. Shouldn't really cause a huge problem, I imagine. The beat didn't "feel" rushed.

Not sure if anything is connected to the external MIDI DIN port.

I don't have anything connected to the MIDI out. I also tried disconnecting all external devices (except for the speakers) and tried replacing the drum kit with the "Power Standard Kit 1," and the issue still persists.

It is not a clipping issue. Clipping being an overload of level. Your issue is a sustain pedal issue and running out of polyphony. Here is how that happens...

You have a couple of right hand keyboards sounds that are using stereo samples (the CFX Piano, Strings), listening to the MP3 you included, when you are doing the repeated chords with the sustain pedal down, you can run the instrument temporarily out of polyphony. This is why it does not happen all the time, only when the sustain pedal has been held down and the repeated chords stack up enough notes to hit the limit... this is manifesting itself most prominently in how it squishes the duration of the snare drum sound (cutting it short) in PART 6... you can avoid this by

1) thinning the Performance... With the Super Knob at 88 ( around 2 o'clock) as in your screenshot, you are hardly hearing the PAD CHOIR, or the STRINGS (Super Knob is Mixing their level) - they are there but given how expensive they are polyphony-wise I would consider replacing these or removing them. At least turn the Sustain RECEIVE SWITCH to OFF for these PARTS (pad and strings). This way you get their sound, but it does not unnecessarily bleed into the next chord, using polyphony up too fast. Leave the Sustain on for just the CFX piano... The "Keyboard" sound in PART 4 is using four notes for each note you play when you exceed a velocity of 93... (making it a major polyphony HOG... you could play softer to engage less notes of polyphony but that is not a good solution) I would evaluate this PART's contribution to your sound... TO TEST THIS OUT: Using the right front panel buttons [PART MUTE] all but the right hand sounds... see if there are any that when you drop them out makes no difference to the character of the sound you are going for - you may find that for using so much polyphony, PART 4 might not be worth its weight... You may find removing it makes not sonic difference in what you are going for... This will return tons of polyphony to the "pool". You can Delete a Part by touching the Part name box to see the pop-in menu > touch “Delete”.

2) Swap the Drum Kit to PART 1. PART 1 has priority for polyphony, meaning it will not get cutoff like when the Drums are in PART 6. This will mean if any notes are swallowed due to too large a polyphony request, it will not be a drum's note duration.
You can Exchange any Parts by: Select Part 6 > press [SHIFT] + [EDIT] > move option from COPY to EXCHANGE, set it to swap Part 6 with Part 1

3) Better Sustain pedal technique... Synthesizers must have an oscillator for each timbre you wish to sound. PART 4 (when velocity is greater than 93), uses four notes of polyphony for each key you trigger in its range. So the repeated 12/8 chords with the sustain pedal down, is very much a torture test for note usage. If you held the pedal through one whole measure you'd run the polyphony out - four note chords are 16 notes of polyphony each (just for PART 4), the sustain pedal down tells them all to keep sustaining... you CAN'T do that. It is amazing that it does not happen more often given your playing Toto's "Hold the Line" (one of my favorites).

I've attached a version of that program in MONTAGE CONNECT .X7B format... (it's zipped, of course) doing some of the edits I mention above (I did not remove any sounds because that is YOUR decision, but I recommend it) Lose PART 4..., it is much more difficult to run this one out of polyphony... of course, you can run it out of polyphony, that is not the point, really, ever... the point is to get the sound you like given the boundaries of the instrument.

Hope that helps

Hello Bad Mister,

Thanks for taking the time to listen to the audio clip and providing the information above (excellent write up! 🙂 ). I'll definitely add this information to my list of notes (btw - I am using Microsoft OneNote to document my findings as a go along. It's pretty easy to use and organize notes...in case anyone is interested).

The information on polyphony will help me manage this better, in case I run into it in constructing future performances.

I have a few follow up questions:

1. Is it possible to still run out of polyphony even if the sustain pedal is disconnected? I'm still able to reproduce the issue without using the sustain pedal in the current performance. For example: if I play 6 notes total (octaves with the left and 4 note chords with the right) or even 5 notes (only using one note in the left) with no sustain, I still seem to be capping out the polyphony playing the consecutive chord pattern.

2. I tried to do the math for figuring out how much overall polyphony is used per note, but I'm not sure if I can calculate it accurately based on all of the variables (such as velocity and the number of elements being used per part). Is there a "formula" for getting a ballpark figure for how much polyphony a performance might consume per note (assuming no sustain is being used)? Btw - I'm not sure if a feature exist yet on other keyboards, but it would be useful to have an indicator that showed how much polyphony is being used while playing. Maybe something to consider for a future firmware release.

3. Polyphony priority - Does the priority of polyphony depend on the part number order? For example, does Part 2 get priority over part 3, part 3 get priority over part, 4, etc...

4. Do you know how much of the polyphony "pool" the drum track uses (roughly)? Just curious how it factors into the equation.

Thanks for providing the .X7B file, I'll take a look at it tonight after work.

It would be easier if there was a display showing used polyphony - then you could disable PART(s) by turning off keyboard control and either enable or disable the ARP(s) for those PART(s) you want to see polyphony usage in the context of a running arpeggio.

You would go through each PART and see if, by the rules, any combinations of the Elements are mutually exclusive. For example - if some elements do not overlap with others by velocity or key limits. If there is overlap - add. If not - then only count one of the two non-overlapping elements. That's for a single note. Determine how many notes of the PART you'd, at max, play and multiply.

Arpeggios, you'd have to "break open" the ARP and analyze how many notes are played at once at maximum. Use this as the max # of notes at a time.

That's not it. Take a Performance like "Orch Glockenspiel". Play an 8th note of middle C, lift your finger, and quickly play the G above middle C. Maybe play the other way (G first, then C). What you'll notice is that even without sustain - you can clearly hear both notes "G" and "C" ringing out due to the long release time.

This has to be taking up polyphony (playing the release of a previously pressed note you are no longer pressing). So this behavior would need to be factored in. Even the drums, with say repeated crashes, should have some overlap where notes do not necessarily have to be triggered at the same time to take up multiple polyphony slots.

Also note that just because the Element Activity Indicator is OFF does not mean that the element is not consuming polyphony. What the indicator shows is if the element is still active for the LAST trigger of that element. Do this experiment - you can use "Orch Glockenspiel" for this. Go to the AEG (amplitude envelope generator) screen and set the release time to 127 for both elements 1 and 2. Play middle C. Now adjust the release time to 0 for both elements 1 and 2 and play any note. You'll hear a "thud" (no release) for the note - but the original note you played is still droning (long release). You'll notice the Element Activity Indicator lights are both OFF. However, you can hear the 1st Middle C you played is still sounding. So the lights "reset" on every trigger -- which means that the lights are not a true indicator of if the element is consuming polyphony or not (sounding or not).

Due to a number of interactions which "stack up" -- enough to reach 128 -- it's difficult to do this by some simple "rule of thumb" approach. Every performance is a snowflake with some homework to do in order to get a reasonable approximation. It would be better to have the keyboard tell you how close to the waterline you are so the musician can spend less time in engineer mode and more time in music mode. A bar graph would do. Using this information, one could quickly get a general feel for polyphony usage for each PART under different scenarios (sustain pedal or not, maybe an A.SW that causes a spike, a resource hungry PART that could be swapped with something "lighter" ).

That said - it's possible that the resource/variable/register that holds this information is not accessible to the OS-level of Montage. Which would make this an unreasonable thing to expect implemented.

Thanks Jason for providing the explanation and examples as to how notes can potentially remained sustained after they are played. I've learned a lot of valuable information in this discussion. You guys are very helpful, and I appreciate your help!

Now it's time to do some playing. Have fun! 🙂