I recognize your problem, but have no idea how this is caused.
However, I do have some workarounds for these unexpected experiences;

1) If you start from an original Yamaha performance be advised that sometimes the multi-knob influences part 2 or 3, based on the input on 1. So your part 1 is not that clean as you expected it to be. So start part 1 with a single part performance. or leave all related parts of a multi-part intact and do not replace them by others.

2) Delete all default scenes assigned to the original performance. And start creating scenes from scratch.

3) Sometimes, after all scenes are turned off, I still have to assign scene 2 BEFORE assigning scene 1. Otherwise, somehow the memory of 1 keeps involving the settings.

These options work for me.

For real backgrounds and explanations you probably have to wait on Phil.

Rod wrote:

Hello - I must start with an apology - I know I've asked this same question before, but about 5 weeks ago I had a computer wipe-out that lost everything, including all my notes, so I'm having to reinstate a lot of stuff. My question concerns Scenes ... I have created a number of 4 Part Piano Pfs, and extended them by creating a 'Piano Bar' or 'Cocktail Lounge' piano for each simply by dropping one of the Parts by one octave. However, I thought I could create 4 variations on each Pf by dropping one octave on each of the 4 Parts in turn, storing each one as a Scene with the 'SHIFT + SCENE' buttons, then storing the whole thing as a 4 Scene new Pf. It doesn't work like that! Dropping Part 1 through 'Part Settings - Pitch - Note Shift' and then 'SHIFT - SCENE 1' is lost as soon as I return Part 1 to normal (ie: Note Shift = 0). When I switch to Part 2 in order to drop the second Part by an octave - Part 1 has returned to normal. I expected it to stay as stored, but it does not.

I've watched Blake's video, and it seems that the Scene variations are achieved only by using the 'Mute' function. Is that all it can really do? That's no good for my purpose, when I need all 4 Parts playing, but one of them varied in turn. Is that not possible? If it is, how can I do it, please?

The scenes can store a variety of things. There are the different categories (which you have to enable individually) like Mixing 1, Mixing 2, Motion Sequence, ARP etc. If you go to the scene screen and then select one of the mentioned tabs after the other you can see what you can store per scene. The transpose does not belong to the things which can be stored for each scene as far as I know! You could maybe do this via the superknob, you can certainly mute individual parts, but transpose directly is not there.

Hi, extract from the Yamaha Music Production Guide (by Hape or Peter i do not remember), here is settings that can be store with scene.

Octave and note shift are not stored with scene memory.

What you can do is copy the part you want to have a different tone, change the tone shift of this copy part and use "mute" with scene to change from one part to another.
Scene 1 part at tone zero not mute, part with tone +12 mute
Scene 2 part at tone zero mute, part with tone +12 not mute

Hello Jan, Stefan, and Joel - good stuff! Thank you! I''m piecing together all you've told me, and will report back with what I think I can do once I've established that it will work - Montage has this ability to catch you out by not doing things the way you expect ... lateral thinking that often gives a better result.

OK, satisfied myself that 'Scenes' will not do what I want them to do in the way that I want them to do it. However, by repeating my initial 4 Pfs as Parts 5 - 8 but an octave lower, I can do a better job. My original thought was 4 variations of a 'Cocktail' piano, but the 'repeater' system allows up to 64 variations (although only 8 can be selected as 'Scenes'). The results are necessarily related - I call it 'Shadeplay' - but offers me a whole new concept of sound creation, built with Shadeplay in mind. That ol' devil Montage has come up with the goodies yet again ...

Jan, I do use only Montage Single Part Presets as offering the least complications - I like to keep things simple. Stefan and Joel - I don't use MS or Arps, and it surprises me that 'Scenes' remember stuff like that, but not simple Pitch shifts! However, it is what it is. The info is safely tucked away for future reference should I evolve into using MS and Arps. For now, if they are present, they are turned off as getting in the way ...

I don't use MS or Arps, and it surprises me that 'Scenes' remember stuff like that, but not simple Pitch shifts!

So that you are not "surprised" about what a SCENE remembers and does not remember, perhaps a little background on the SCENE function is in order. You may even discover that there is a way to accomplish what you need... (not necessarily how you thought it would work, perhaps) but the history can help see how we got here.

1983 - year 0 for MIDI, Yamaha introduced the first Clavinova, the DX7 and the CFIII Acoustic Concert Grand (a fairly good year, as product introductions go). At this time, synths with samples still sold for tens of thousands of dollars (SynClavier, Fairlight), 'digital' meant either samples or FM. The "PRESET" was still fairly new - they started to appear on analog synths in the late 1970s.

1984 - Yamaha TX816 introduced...

By 1987 - with the success of FM, Yamaha's pro Synth setup was TX816 (eight DX7 in a rack), the KX88 keyboard controller, and the QX1 (5-1/4" floppy drive sequencer). The Keyboard connected to the Sequencer, and it was connected by eight (count them, eight) MIDI cables to the TX816. There was also a single (global) MIDI I/O that could be used to address the unit as an 8 Part multi-timbral Synth (hmm)! You could link them all together to make one huge eight part sound.

It was also this year that Yamaha introduced their first Digital Mixer... it had 8 inputs, and was the perfect accessory for the setup. You could send a Program Change from your KX88 that would/could address the TX816 to recall and harness your 8 DX7s... back then synths did not have EQ, did not have Effects, did not have Pan controls (they weren't stereo), they didn't have many of the things you take for granted now... you wanted EQ? - you did that in your mixer, you wanted to setup for stereo, you'd better setup your mixer, Effects? If your mixer did not have built in effects (and most at the time, did not), you needed to connect to them via AUX SEND /RETURNs - a Send from each channel and a stereo Return. Digital reverbs and delays were sold separately for hundreds of dollars.

Oh, if you wanted to change EQ and/or use a particular Effect setup... you had to do it yourself manually between songs... ENTER the Yamaha DMP7, a keyboard player's mixer, to be sure. You connected your TX816 to the DMP7 with eight (count them, eight) signal cables. You had a MIDI Junction box that allowed you split your MIDI to multiple locations.

It was then that Yamaha invented what amounts to a "Program Change" for the mixer. It could recall, in response to MIDI messages, and entire setup for the mixer all channels and settings. This was the dream setup, your mixer was 'total recall' and you could do it all with a single button press from your keyboard, the KX88! Along with this mixer "Program Change" Yamaha introduced the MUTE (clearly a mixer term), MIDI spec has no official mute function as MIDI is not sound - with MIDI messages you typically use volume = 0, the Mixer "Mute" actually disconnected audio on that channel. Effects? Buil-in. EQ? Built-in. Heaven!

Well, this was too cool for school, the Scene "snapshot" memory was used to create synchronicity between your Digital Mixer and your Synth rig. It made your mixer intelligent. You had things called Program Change Tables... when you sent a certain PC, you could have that recall a specific SCENE in the digital mixer's memory. So not just for per Song setups but any time you changed patches on your Synth your mixer became a part of what automatically changed. It was like having a sound person who was your biggest fan, because they could make every change you needed, every time... and they would execute it immediately.

When you think about these components separately, I believe you get a clearer picture of exactly how these things evolved (logically)... to where we are today...

__ The KEYBOARD > SEQUENCER > TONE GENERATOR > DIGITAL MIXER > Monitor Speakers... were all very separate devices... in your recording rig.

__ The KEYBOARD > TONE GENERATOR > DIGITAL MIXER > Sound Reinforcement... were the separate devices in your live rig

They still are, separate, in spite of the fact that today you find most of them combined in one product. The MONTAGE has keys, has a Synth tone engine or two, and has a serious Digital Mixer w/automation and effects, all built-in! You can see that the items memorized in the SCENE are principally those concerning the Digital Mixer (Volume, Pan, Mute status, Effect Send amounts...) plus some that clearly address the Synth engine.

But let's focus in on these Synth engine parameters... they are mostly general OFFSETs, not direct parameters deep within the Tone Generator. They are mostly collected from your "QUICK EDIT" knob area and it memorized those things temporarily attached to the Part (which Arp, which MS, and are they swinging, how they reference tempo, etc)).

Not that is not to say you can't automate functions deeper within the Synth engine... because you can... anything that can be assigned to the Super Knob can be recalled because the Super Knob value is apart of the SCENE memory. Anything that can be manipulated by a Motion Sequence can be recalled because the MS is stored in the SCENE. Even Arps can be used as an automation tool. Whether or not, it makes sense musically is tbd (to be determined)

And this brings us to some of what's new (evolutionary, when you know the history) about MONTAGE... It is able to add control over a wide variety of parameters and you can (if you're crafty) create the result you desire. A SCENE is not a "snapshot" of all parameters, for many reasons that's not a good idea. The features that are memorized directly are carefully selected. Those that you assign indirectly (via a surrogate AsgnKnob/SK, or MS, etc) are for you to select and make work.

NOTE SHIFT is Assignable as PART "PITCH"... Whether it serves your idea, you will have to tell us. On Synth sounds stretching pitch maybe fine, depends on how sensitive your ears are to "munchkinization". Mileage varies.

Transfer the above discussion to the Real World - what makes a good candidate for SCENE automation:
NOTE SHIFT... think of Note Shift as a guitar player using a Capo to shift the key of their instrument.
Now picture that guitar player in a recording studio... what button in your control room is going to move that Capo up a half a step? What automation control would you activate to make a change on that guitar player's instrument? You see?... changing the instrument in this case means editing the sound... and because NOTE SHIFT is a real time accessible parameter, it can be a candidate for SCENE memory... but you will need to experiment to see if this will work for you.

The correct answer to the question "What button in your control room... " is the Talkback button, because you will need to ask that musician to edit their sound at the source, because in your control room you don't have access to automate that parameter!

What you can do:
You can assign Part PITCH control to a Controller.
If you assign it to a controller, you can then investigate whether the controller movement serves your purpose. As you know, if you assign it so that it is linked to the position of the Super Knob, and the position of the Super Knob is stored in a Scene, therefore you should, in theory, be able to accomplish your goal via SCENE memory

Hope that triggers some thought.... using Note Shift (which tweaks the tone engine) to raise or lower something an Octave may (or may not) give you the effect you want, I'm not at all sure what exactly you're trying to capture. But MONTAGE does allow PART PITCH to be memorized when you assign it to a knob and link that knob to the Super Knob... you can jump to that tuning by storing the position of the SK.

@Rod. Yes, that's also how i do it. Copying a part and set the pitch on the copy.

Hello BM - you know your stuff backwards, thank you - I follow the evolution path, but am still puzzled. For example, the Scene does remember the pitch change so long as I don't change it after storing it. Thus, I can put Scene 1 at -12 and store it, and it stays at -12 when I change Scene 2 to -12 and store that, and so on. But if I store Scene 1 at -12, and then switch it back to 0 before moving to Scene 2, it reverts to 0. Why does it do that when it's stored as -12? It must be stored somewhere because it stays there so long as I don't change it after storing it. How can it remember a 4 bar Arp that bounces up and down over 2 octaves, but not remember a single one octave switch? It may be clear to you, but to a layman it's far from it! However, as I said, it is what it is and we have to adapt to it as it can't adapt to us. Still a blinder of a machine, warts and all!

Jan, I figured that's what you were doing, and I've extended it into a genre in itself (I call it 'Shadeplay') aimed at those seeking a very particular sound - 64 variations in a single Pf, each of which can be varied still further, based on a type of sound. Anything from a smokey dive piano in Marrakech to Star Fleet vanishing down a Black Hole. Or it could be a mixture of types of sound, as expounded by Blake in his videos. Limitless. Unfortunately time-consuming, and my problem is lack of time ...

I don't think it has warts (as it concerns your post), at all. I think I should probably just directly speak to your issue as simply as possible.

Note Shift is NOT, repeat NOT memorized directly by the Scene. See Joel's post (with graphic) above. Do you see Note Shift? No.
Because NOTE SHIFT is NOT memorized directly by the Scene function.

If you are curious as to why, reread my previous post.

What is stored in the MONTAGE version of the "Scene", that concerns your request, is the position of the Super Knob. Each Scene can have a different position of the Super Knob.

Note Shift is the pitch (tuning) of the tone generator Part. If you assign this Pitch control to a PART AsgnKnob, then link that AsgnKnob to the Super Knob, then you can indirectly determine the current Pitch of the Part by memorizing the position of the Super Knob. The Super Knob is smooth continuously variable control, so you will need to "tune" the Pitch... the Super Knob will work like the tuning peg on a stringed instrument.

I mentioned you will need to determine if this meets your needs, sonically. If you need to store an instrument at a different Note Shift as much as an octave, I would go with the recommendation above to use a different Part. I'm not sure what particular effect you are going for (like why not just hit the OCT button when you need to drop it an octave, for example?) ... hitting the OCT button "Transposes" the Keyboard, instead of retuning the instrument. Yes, there is a difference.

You can use the indirect method as BM recommends (superknob with destination as pitch) or you could use multiple PARTs as suggested by other users.

The multiple PART idea is as follows - say you have a string PART you want to use scene 1 to play normal pitch, scene 2 play an octave lower (-12 note shift) and scene 3 you want strings to play an octave higher (+12 note shift). And all the while, you want a piano layered with the string sound which does not shift.

You could accomplish by assigning PART1 to a string part with no note shift. PART2 as a string part with -12 note shift. PART3 as a string part with +12 note shift. And PART4 as your piano part. Each PART spanning the full keyboard.

You would have scene 1 _mute_ PART 2 and 3. This would combine the non-shifted string with the piano.

You would have scene 2 _mute_ PART 1 and 3. This would combine the -12 note shifted string part with the piano.

You would have scene 3 _mute_ PART 1 and 2. This would combine the +12 note shifted string part with the piano.

One method frees up superknob for other uses. The other method frees up PARTs (no need to "burn" PARTs). Depends on what you need.

Hello BM - Ah! I see your reasoning for the workaround by using the Superknob position as a means of 'remembering' the note pitch change. That could work, thank you, but I think your other advice of going with a second Part is more practical for my purpose. Which is to create a new Pf, not for a piece of music, but just for itself, so using 'Transpose' wouldn't work for me. I hope though, that having to go from A to D by going through hoops B and C will illustrate my puzzlement - it's a bit like saying I can only fly from Miami to Boston by going through San Francisco and Seattle. But it is what it is, so we must work with what we have.

Jason - the possibilities are endless! My current exercise is to produce a series of single (8 Part) playable Pfs that offer a wide range of quite closely related sounds so that a musician can easily audition and select the one closest to what's in his head. Waste of time, I know, as no-one is interested in what I do, but it does keep me off the streets ...