Start with Oxygen

Can you suggest a different Performance to start with? "Oxygen" isn't a preset.

I'm not saying this to undermine the rest - just that it would be good to be able to collectively get on the same page.

Picking a random pad preset "Dark Light" I would change this to a pluck using AEG:

1) Part 1 element 2 - turn OFF (just for demonstration - element1 and 2 are related with one more left side and one more right). This is the equivalent of pre-cooking in the cooking show and pulling out the baked goods.
2) Element 1 - "Amplitude" -> "Amp EG" menu:
Time: Attack=60 Decay1=40 Decay2=87 Release=59
Level: Initial=0 Attack=127 Decay1=35 Decay2=0

Bass sounds like a plucked bass of sorts. Just a little body before tapering off. Mid and high end is staccato.

Say I want to add some kind of "pick" or percussive sound to the start. Randomly, I decide to "turn on" element 3. It happens to be a piano sample which has a percussive sound in the beginning already. Lucky. I want to just chop off and tone it down to add a smidge of percussion to the start.

3) Element 3 - turn on and change same "Amplitude" -> "Amp EG" menu:
Time: Attack=0 Decay1=21 Decay2=3 Release=50 (doesn't matter, release that is)
Level: Initial=0 Attack=53 Decay1=32 Decay2=0

On my Montage the [PERFORMANCE CONTROL] button is lit and the last row of buttons that shows "ARP SELECT" over it works as mute control here. I press the "3" button to mute and unmute Element 3 to hear how that sounds with and without "percussion".

I set the attack level to 53 here just as an inexpensive (lazy) way of setting the relative level of this element. I could have also used higher values and then go back to the element's Level/Pan menu to scale it back. Either way. This way required less work.

The effects are a little wet for my taste - but that's another thing.

I'm not sure what you're trying to do with FEG. But if I wanted to get more of a "duck" sound rather than "pluck" - I'd FEG it up.

4) Back to Element 1 "Filter" -> "Filter EG" menu
Time: Hold=0 Attack=45 Decay1=93 Decay2=38 Release=75
Level: Hold= -128 Attack= +127 Decay1=0 Decay2=0 Release=0

If I wanted something more "laser" like, I'd take the FEG Depth and change it from where it started at +44 down to -64

The nature of a pluck is it doesn't last long so filter changes only have a limited time before the core sound goes away. The filter changes can't do much over any long period of time.

Each filter FEG starts with each new note and follows the note. Filter sweeps that span multiple notes belong to an LFO of some sort (including Motion Sequence) or a control (i.e. not note) Arpeggio.

I don't know where you're trying to go and started here only because it's a preset. It may be better to start with the one you want and try to sculpt something closer to what's in your head.

If you mute element 3 from here and change element 1's FEG Depth to +63 you'll get more of a pan drum sound. Not too far from the +40 FEG "duck" sound though. Pan seared duck.

SORRY. I think I must have grabbed Oxygen off SoundMondo.

Can't check right now... as I can't login to SoundMondo. Something new and different...

Also, Dark Light is an AWM.

There's a MASSIVE difference between how filters work between FM-X and AWM, not just in terms of where they are (each element etc vs Common), but also in terms of the shape, timings, sculpting and "curve" of the "ADSR" envelopes, and this is most noticeable when there's some resonance both in the filter and the FM sounds - both of which are the strong suit of the filters and the FM-X realisation of FM... making this problem with the envelopes much more pronounced, and annoying.

Which is explicitly why I'm talking about FM, and it's in the title.

Ok - we're getting closer. I have to scarf off of Soundmondo.

https://soundmondo.yamahasynth.com/voices/23648

There I take it.

Can't get SoundMondo to even load on one computer, at all, in all browsers it has (it does not have Chrome and it won't ever have it, but it has every other browser known to man).

On a Mac, with Chrome, and permissions allowed, 3rd party cookies, etc... won't login and can't get its treasured secure connection.

But yes, that's the one!

I've seen Soundmondo work with Opera before. Chrome is the only one supported though. What I got to work here was an iOS device and the Soundmondo App. With Mac I have trouble getting the Montage recognized except with BLE and BLE doesn't work right with Chrome for me (on Mac). Not going to address that here though. I'm guessing you may have saved Oxygen in its original form on your MODX somewhere or can figure out the computer issue later or can start with a different preset.

... pressing ahead - my approach to Oxygen:

1) Turn off all effects - there's a lot of echo type effect in this so get rid of that. Press the "FX" button at the top of the screen and turn off insertion, system, and master FX
2) Recognize the carriers are 2, 4, 6, and 8 from looking at the algorithm picture
3) Mute 4, 6, 8 as a group and 2, 6, 8 as a group == etc. I'm listening to see which carrier (2, 4, 6, or 8) is "dominant". I just want to work on one carrier at a time.
4) From this, I determined carrier 2 was the loudest/dominating carrier. Muted OP4,6,and 8 (all carriers except 2).
5) Edited Op2 Level menu:
Time: Hold=0 Attack=35 Decay1=37 Decay2=85 Release=47
Level: Attack=99 Decay1=0 Decay2=0 Rel(Hold)=0

or (pluckier)
Above but Attack(time)=25

or closer to my AWM2 example
Time: Hold=0 Attack=13 Decay1=41 Decay2=52 Release=47
Level: Attack=99 Decay1=64 Decay2=0 Rel(Hold)=0

Do something similar for each of the carriers OP4 and OP6 and OP8 (be sure to mute all even OPs except the one you're currently editing to isolate - I do this rather than "solo" because it makes more logical sense to me - you don't have to strictly follow).

Then you can turn the effects back on and determine what you want to keep or tone down.

Myself I would turn off both of the insertion effects and reduce the sends to system effects.

I didn't see a huge difference in approach between the two. FM-X has more options since you can have infinite hold times which I find useful for other purposes (mostly level generators for envelope followers).

The first part is exactly as I've been doing, to isolate which operators and carriers do what. In other endeavours!

This whole questions is about the elephant in the room... the envelope on the filter.

I, too, know how to fake a subtractive synth with FM by carving the individual operators.

But that's not what's being asked, here.

Think about this from the perspective of wanting to use the Filter to do fancy things to a well defined, quite nice FM-X sound... that sound is the Oxygen. It's a quite nice sound.

But once you start trying to play with it via the filter, particularly a normal low pass filter with a bit of resonance on it, you'll notice that the Release cycle is the opposite of nice in how it squelches the filter's resonance.

Just think about that one thing, to start with... see if you can describe what's going on, and wrong, with that release phase of the "ADSR" EG on the Filter.

That's the part I'm really struggling to understand, before anything else.

Getting as near to a useful envelope on the filter as I can, and still needing to "fix" it via the amplitude envelopes then reveals a further set of problems, including that this is no way to try to resolve the issues of the release on the filter.

I'm working my way there. I don't know if you can tell or not - but I'm trying to get on the same page. Maybe you can just skip all this and post your version of what you think is a bad Performance on Soundmondo and I'll start there instead of working through this methodically. Despite what you think you've been clear on - I like to verify because there are lots of steps to get there. The information I had to work with on what was going wrong was subjective and not specific enough to have walked in your shoes yet.

First pass I thought you might have been using FEGs - but now I think maybe you were just using an AEG, modifying the filter, and something sounded off to you. If you think there is crystal clarity - there isn't. I'm starting (now) with a common Performance with a specific link. I'm communicating specific parameters. You have a chance to communicate your own if I'm off. Or supplement with specifics so I don't have to infer or upload your "bad" example to Soundmondo if that ever can work. Or if Montage Connect works for you then you can upload a ZIP of your X8B file showing the bad sounding Performance.

When I'm working in a group setting and someone is doing some "work" - someone else doing some steps will often verbalize what they are doing. Step by step. So I can follow. Not because they're teaching me anything. Not because it's a lecture of how to do the process. It's merely documentation. Maybe you can understand better why I am verbose in my steps? Also, it's your chance to say - oh, no -- at whatever step don't do that because (perhaps realizing there was ambiguity) - XYZ needs to be done. Or - you got that FEG part wrong because I'm really just talking about how the standard filter works after exciting with the modified "pluck" AEG. It's not a fight - it's collaboration.

This is a response addressing your previous message (now shown as #7 in this thread). Thanks for taking a step back, realizing some need for detail, and supplying additional in the next follow-up.

Let me know if you want to somehow convey with a zipped X8B file attachment, Soundmondo upload (I know that's problematic) or detailed delta of parameters between stock Oxygen and the problem. If you want to do this first without my working to get there -- or if you are OK with the longer road of me inferring what I can as I make best attempts to reproduce the version of Oxygen you have which has the problems you describe.

Have a listen to the release aspect of the filter EG, on Oxygen. Isolate the best little stack of operators, make the filter a strong (definitely doing something) low pass, with a good amount of resonance.

Now try to release it in a nice way. Examine the release "curve".

Let's start there.

Because at this point you're seemingly looking to obfuscate and/or deflect from examining the envelope's effect on the Filter, making me think that the distasteful effect (it makes to my ears) is not just subjective.

obfuscate

You're wasting your time laboring on this. I'm not paid by anyone to lead people astray nor do I have any personal motivation to do so. I've got a record of being highly critical of Yamaha and even when I see both sides -- I have empathy as a user more so than a company. I don't have strict loyalty to Yamaha - I use other gear. I just think you're barking up the wrong tree and have perhaps some bias that what you're doing is painfully obvious to everyone. I'm just a keyboard player. I've got a gig tomorrow. I've got to practice my parts and fix up some Performance problems. I'm not sure why I would waste both my and your time by blowing smoke. You're free to have a contrary opinion - it's just hard for me to dream up of any plausible motive for what you're suggesting.

Starting where I left off:

Common: "Pitch/Filter" -> "Filter Type" left this as LPF24D. Cutoff at 140. Cutoff/Key +84%. Resonance: 52. I believe the rest is as-is but for completeness Rate/Vel: +7, Cutoff/Vel: +21 Gain: 255

Filter EG
Time: Hold=0 Attack=48 Decay1=35 Decay2=0 Release=0
Level: Hold= -75 Attack= +127 Decay1=0 Decay2=0 Release=0

Of course I'm shooting in the dark here.

Is it trolling?

You have a release time of 0. There's no way you're going to hear the release curve if you set it to that.

And just in case you're fixating on the goal, which is really a mere example being used to examine the release curve, consider that I know the trick of using the first decay at zero Level to approximate a normal ADSR, but you've left off a little trail on the Release, that's needed to make sure it behaves nicely on quick key lifts, which is probably obscured by having such a long attack time.

Remember the example, a pluck sound. You're going to get a bow sound with that long of an attack.

But back to the the question, it's about the release portion, and the "curve" it has, and the way it interacts with a resonant filter... IF you put the Decay's up high enough that you're able to hear the Release portion carve the filter's influence off to zero.

It's in the original question. It's what I'm trying to understand, by way of this example, and the commentary.

It can't be that the release portion of this "ADSR" is so bad sounding and poorly shaped on purpose, can it?

it's not linear, that's for sure. And its influence on the filter is... odd. Have a listen. Try to create an "ADSR" curve that lets you hear that release portion carve off the filter to zero.

Then experiment with it, and see if YOU think it's pleasing and right and proper in the way it deals with resonance in both the filter and the operators/carriers.

Pluck doesn't have a release (or much of one or ... it's a subjective call). I've already requested you post up your Performance.

I expected to get here - a place where something was wrong (with my guesses) and you'd have to correct it. Give me the parameters. I want to start in the same place. I told you I'm taking guesses. I can read what you suggest - but it's not crystal clear. Crystal clarity is communicating specific parameters either by a Soundmondo download, X8B bulk dump, a sufficient list of specific parameters (as I've modeled) to get on the same page. I'm not claiming I'm matching your Performance. That's for you to communicate. That's YOUR responsibility. I'd be perfectly happy with starting with your broken Performance and go from there.

Again, despite your notions - I really don't have a concept of what exactly you're after as an end product. I know there's something you don't like. You've described something that I am telling you isn't sufficient for me just to plug in and be on the same page. Meet me where I am and tell me what dial to turn to break this however you currently have it. This is not the most efficient way - but I'm willing to put up with this in order to help. I'd love for you to find someone with more clairvoyance than me - but you're presently stuck with my extreme lack of it. So do something constructive.

You're fixating on finding a specific result.

Read the initial post again. Slowly. See if you can see that I'm trying to understand, through examination, what's going on with the falloff created by the final stage of the "ADSR" as it pertains to its influence over a resonant filter and/or resonant operators and carriers in an FM sound.

If you're not able to create versions of a Release that cause you to question what it's doing, and make you wonder what's going on when Release is carving a falloff out of resonant Filters and FM sounds, you're never going to grok what I'm asking, because... well I don't really know why someone wouldn't be curious about the Release of these unique "ADSR" envelopes.

If you're struggling to understand the difference between objectives that are specific and more generally being in search of insight and understanding throug analysis and experimentation, you're simply not going to be able to grok why someone might be this type of curious.

So rather than trying to second guess me for the umpteenth time, explore that Release and let me know what YOU think of it, please.

EDIT: just realised you're definitely taking the piss, or trolling, or both.

There's no way you could have tried 44 for the attack time and not realised that's the last point before the value starts going exponential in a big way. Which is the exact same thing it does on release. Well played.

Anyone else got an opinion on the filter adsr release on FM-X?