S775 Hidden Voices

[BogdanH]

Hello
I discovered (by coincidence), that S775 plays few voices which aren't published in S775/S975 Data List AND which aren't installed from some pack either. It's described in this thread how all started:
https://www.psrtutorial.com/forum/index.php/topic,48581.0.html

I hope forum mods don't mind I started new "documented" thread related to "hidden" features.

As it seems, S775 (and S975?) have at least 17 voices which user can't access (select) directly on keyboard. These voices are used in Yamaha's Vietnamese pack. Anyway, if you install Vietnamese pack into S775, keyboard will use correct built-in voices (thus, makes no sense to put voices from pack into custom made pack -only styles is enough).

Ok, but what's the use, if one can't select these voices to use them? Well, if you have some style editor on PC, then you can assign desired voice(s) there, import modified style on keyboard (for further fine voice settings) and save style back on PC (i.e. for custom packs). Yes, some will say "who would listen to Vietnamese instruments?". I say, adding some "exotic" voice in few music notes can add that certain "spice" to music piece -my humble opinion.

Vietnamese voices have following data:
MSB=51
LSB=3
PC=0 to 16*

*-that's the PC value one should write directly into Style file (PC values in Yamaha's DataList are increased by 1).

For those interested (to confirm it's the same for S975?):
I've prepared a zip file (see attachment) which contains one (same) style each, however they contain all hidden voices in OTS1-4. First style has voices 0-7 for R1 & R2, and the second one has voices 8-15 for R1 & R2. Voice 16 is defined for L in both style files. Btw. when installed, "new" voice names will appear as they should.

What does all above mean? Well, fact is, these voices aren't published and that means, they can disappear in case new firmware is installed. On the other hand I wonder... why didn't Yamaha made them available on keyboard?

Greetings

[attachment deleted by admin]

[hans1966]

Hello BogdanH, it happened the same to me two nights ago with my S775, editing a style from the Onacimus style collection "GENOS, for the S970". some instruments sound, but I do not see them, instead the box appears with a piano keyboard icon. the sound in question was like a synthesizer and pad at the same time. but for more that I looked for in the corresponding banks (Chorus and pad - Synth) I did not find it. On the other hand, I do not know if Yamaha will release a new firmware update for S975 / S775, that it will be able to show those hidden voices, or even implement new voices. we would have to wait and see. Greetings. Hans

[Mega Voice]

The same applies to the Indonesian 2 package.

https://www.psrtutorial.com/forum/index.php/topic,48561.0.html

[BogdanH]

@hans1966
Thank you for showing interest. I've installed few "Genos for S970" styles and checked some of them (too many to check them all). What I could see is, those styles use many voices which are not present in S775 (dashes instead voice names). Many times however, S775 still plays some of those unknown voices. I assume (pure guessing), S775 substitutes them with some "best match" voices. If we try to change such voice, S775 will direct us to the voice, but we can see that voice isn't really selected. In case of "hidden" built-in voices, S775 won't direct us to the voice which is being used -even if it's name is shown.
As said, I didn't check all GenosForS970 styles/voices, so there might be some hidden voices used...
I mentioned firmware case only to alert on "what could" happen.. meaning: if you have voice files, keep them safe somewhere -even if you don't need to use them inside custom packs.
And finally: because hidden voices can't be selected on keyboard, having them inside pack is the only way -at least as long we don't decide for "final" (revoiced) style version.

@Mega Voice
Thank you very much for Indonesian styles! Not only I like few of them (incl. some voices), it also helped me to map Indonesian hidden voices. Here it goes:

Indonesian hidden 22 voices:
MSB=51
LSB=2
PC=0-21*

*-that's the PC value one should write directly into Style file (PC values in Yamaha's DataList are increased by 1).

Greetings

[SeaGtGruff]

Based on the values of these "hidden" voices, one wonders what's at the following values:

MSB = 51
LSB = 0
PC = 0 to ?

MSB = 51
LSB = 1
PC = 0 to ?

MSB = 51
LSB = 4 to ?
PC = 0 to ?

A few years ago I thought I had found a couple of "hidden" voices on my PSR-E4xx models, because there were a couple of MSB/LSB/PC combinations used in a few "dual" or layered preset voices even though there were no "single" or non-layered preset voices with those MSB/LSB/PC combinations.

But I eventually concluded (based on how the "hidden" voices sounded when selected via MIDI) that those values were probably just selecting standard XG voices as described in the XG Specifications. In certain cases— that is, for certain MSB/LSB values— requesting a voice that doesn't exist will result in the keyboard falling back to a standard XG voice; while in other cases requesting a nonexistent voice will produce silence.

In the case of the voices I thought I'd discovered on my PSR-Es, it's possible they had been used on other models but had been replaced with similar voices having different MSBs or LSBs. Or I suppose it's possible they may even have been typos, since their MSBs or LSBs were usually 1 off from one of the existing voices.

I'm not necessarily suggesting that this is the case with the voices you've discovered, especially since it sounds like you're already aware of the XG "fall back" behavior for requested-but-missing voices, so I assume you've already verified that this isn't what's happening. But it's worth keeping in mind for anyone who might discover "hidden" voices on their keyboards.

By the way, have you tried creating user voice files that select these voices? That might be a way to actively use them on the keyboard without having to put them in style OTS voices.

[EileenL]

Many of the voices on Genos will not be on other keyboards so as you say they will try to substitute these with XG voices. That is why styles converted from Genos will never sound as they are supposed to in any other keyboard.

[BogdanH]

@EileenL
I assume (considering price difference), Genos might use more "advanced" signal processors and thus, even if we would install "native" Genos voices (by using original wav files), those voices probably wouldn't sound the same as on Genos -seems logical to me. But then, we poor guys are happy with "close enough" 

@SeaGtGruff
Happy to see you in this discussion. I'm no keyboard expert by no means (I'm just very curious person), so any info or hint, is very welcome for me.
Yes, LSB=2.. LSB=3.. I noticed the pattern too. So I made a file containing MSB=51, LSB=0 to LSB=7 and PC=1, hoping there would be more hidden voices. Especially because Yamaha introduced more than just new Indonesian and Vietnamese icons (new South America, Africa, etc. icons). But I found nothing. That doesn't mean there aren't some more.. but checking all banks would be very time consuming.
For checking voices, I made a simple program, which changes MSB,LSB,CP values inside style (it does that in "dirty" way without parsing style file -without evaluating midi structure that is).
I also tried to change voices by sending midi commands from PC (that way it goes much faster). The problem is, in that case, Voice icon and voice name on S775 display doesn't change -so I don't know am I listening some "hidden" voice or not. Again, I'm not expert in midi.. and considering many Yamaha Sy*** commands are secret (from what I've read), I probably won't put much work in that direction. I mean.. I need to learn how to play music better 

About

...have you tried creating user voice files that select these voices? That might be a way to actively use them on the keyboard without having to put them in style OTS voices.

I don't know what exactly you mean. Ok, I've made a new voice by using Yamaha's Expansion Manager -just to learn how it goes.  Would you please explain how to make user voice that would use another (hidden, in that case) voice?
Thank you!

[pjd]

@SeaGtGruff
AboutI don't know what exactly you mean. Ok, I've made a new voice by using Yamaha's Expansion Manager -just to learn how it goes.  Would you please explain how to make user voice that would use another (hidden, in that case) voice?

Just a little background information. When a user selects a panel voice, the instrument reads and interprets a voice file. A voice file has one of several possible file name extensions depending upon the type of voice: VCE for regular voices, LIV for Live voices, SAR for Super Articulation voices, etc. Sometimes people call these "VCE files" because of the original file name extension.

A VCE file is actually a Standard MIDI File (SMF) in disguise! If you rename the file extension to MID, the file can be opened in a DAW for editing.

The VCE file selects the actual (internal) voice using MSB, LSB and Program Change number. Then, the VCE file sets the DSP effect, EQ, filter resonance, and other Voice Set parameters.

If you change the bank select MSB, bank select LSB and Program Change number within the VCE file, it will select a different internal voice when the edited VCE file is read and interpeted.

Hope this info helps! I'm pretty excited about your exploration. 

-- pj

[SeaGtGruff]

@BogdanH, what @pjd said is exactly what I had in mind.

First, you would create and save a "user voice" on your keyboard using its "voice set" functions, or whatever they're called. None of my entry-level Yamahas have that functionality, so I'm unfamiliar with it myself, except indirectly by reading what other people have written about it. My understanding is that when you select one of the preset voices, modify its parameters-- even as simple as just changing its volume or panning, but potentially including things like modifying the attack time, release time, filter cutoff, filter resonance, reverb depth, chorus depth, etc.-- and then save the custom settings to a "user voice," the keyboard creates a data file that has a particular extension such as "VCE," "LIV," and so on, as @pjd said. It doesn't contain any new sound samples the way a new voice from an expansion pack does; it's really just a MIDI file, except it doesn't contain any Note On/Off events, just the Bank Select, Program Change, and Control Change events which select a specific preset voice and then modify its voice parameters from their preset default values.

Second, you would save or copy that "user voice" to a USB thumb drive, copy or rename it to a file that has the standard "MID" extension, load it into a MIDI editor such as a DAW or MixMaster, change the values of the Bank Select and Program Change events, save the changes to a new "MID" file-- probably giving it a new name to reflect the "hidden" voice you're selecting with it-- and then copy or rename the "MID" file back to a "VCE" file on the USB thumb drive.

Finally, you would try selecting the modified "user voice" from the USB thumb drive to see if it successfully selects the "hidden" voice as intended.

[BogdanH]

Thank you both for clarifying. I assumed you might mean this, but it was too obvious, so I though I better ask.

Yes, I did what you say and anyone can do that with files I attached in my first post. But to speed up the process of checking, here I attach zip file which only contains VCE files. If you choose those (midi) voices from USB or copy them in user area, you can select/use any of them in your own styles. Needless to say, because these voices are built-in, you only need to save style.
Btw. You can rename voices, of course, but it's icon can't be changed (because it's defined in keyboard).
As said, I assume it works only on S775 and probably on S975 (latter needed to be confirmed).

I'm glad that I could make my small contribution to this generous forum.

[attachment deleted by admin]