Merging similar packs

[andy_v]

Hi all - I've just recently bought a Genos and am delighted to have "joined the club"!
As a voracious consumer of free stuff, I've naturally downloaded all the Middle Eastern expansion packs, and tried to cram them all onto my Genos, along with the other freebies (the Superior pack and the Best of Europe pack); but unfortunately, there's not quite enough voice memory to accommodate all of them on-board at once.
That's fair enough, of course - there was always going to have to be a limit, and at least it's much higher than it used to be - but the tantalising thing about it is that all of these Middle Eastern packs have a significant number of (presumably identical) oud and qanoun voices.
So what I was wondering is - what's the best way to merge two or more such packs, with a lot of common content, together, but still fully intact, to make more economical use of them within a combined "combo-pack"?
It looks like it could be done by making a note of the numbering of all the shared voices in all the relevant packs, "unticking" those voices in all but one of the source packs, then having built and loaded the "combo-pack" on to the Genos, working through all the broken styles and manually re-voicing them.
Does anyone know of a better way?

[DrakeM]

I have the S950 and Yamaha gave away at one point all the expansion packs for the keyboard.

The S950 can only hold one pack at a time. The solution is simple. You take the style and sub out the voices that are not on your actual keyboard that are used in the style. Then find and change out the voices in each of your OTS for the styles. Now I don't even load any of the packs into my keyboard.

I have never met a style I could not use in this way. 


Regards
Drake

[andy_v]

Thanks Drake.

As it happens, I can't see a single style in any of those Middle Eastern packs that depends on the common voices. I think some of the one-touch settings might be missing a voicing or two, but I'm happy enough to sacrifice those for the sake of well over a gigabyte of expansion memory!

[BogdanH]

hello andy,
Most important, when creating custom packs is, you also include all voices that are used by particular style. Let's say Eastern pack has 20 styles and 25 voices.. But if you only like 5 styles (for your custom pack), then obviously not all 25 voices are needed. It would be max 8-10 voices (guessing from my experience) that are needed for those 5 styles. Keep in mind, that you maybe don't need to include all OTS voices for each style, because you can very easy assign some "similar" OTS voice on keyboard -unless, for example, there's some "special" voice in OTS that is just needed to be there, so the whole result sounds "original".
Speaking for me, before moving from original pack to custom pack, I check every style for voices. Sometimes I replace some voices (re-voice) before I include style in custom pack. For example, I replace some accordion voice with similar one that's built in keyboard. I do all that on PC of course, because if dealing with many styles, it takes too long (and is too complicated) on keyboard.
My keyboard only has 400MB user space, so I need to be very rational with custom voices. However I never sacrifice voices that are really needed for style to sound as intended.

Bogdan

[andy_v]

Thanks Bogdan.

To be honest, in this case I'm probably going to get more use from the oud and qanoun voices than the styles, but I think they're all still intact, so it's a win-win!

If I do have a style with a reference to a missing voice, will there be an obvious indication somewhere on screen - e. g. a particular icon above that channel in the mixer?

[BogdanH]

...
If I do have a style with a reference to a missing voice, will there be an obvious indication somewhere on screen - e. g. a particular icon above that channel in the mixer?

On keyboard mixer you can see if some voice is missing for Rhythm1...Phrase1 but that's only partial info and can also be misleading:
1. In some cases, if voice is missing, keyboard will automatically replace voice with some built in keyboard. Result obviously depends on how good replaced voice "match" -can be anything from good to unusable. Point being, because particular section (i.e. Chord1) does play, you'd think everything is ok -except result isn't as intended by style creator. And only in case keyboard can't find substitute, you'll see missing voice icon (and of course, no sound from that section).
2. Rhythm1...Phrase2 (channel 9-16) aren't the only voices that style (usually) uses. Usually more voices are defined in channels 1-8, which are (together with those in channels 9-16) then used in style variations (Main A, Main B, etc.). If you watch mixer carefully, you might notice how voice in some section quickly changes now and then. In short: if voices are missing (or substituted by not so good "match"), style just won't sound as it should. My experience is, majority of shared styles are close to unusable because of that. Many still use them, though -without even knowing there's something wrong.

Sadly, Yamaha's YEM isn't of any help in this regard and so, separate (for re-voicing) tool is needed. There exist quite some even in this forum, for example:
https://psrtutorial.com/MB/bedesem.html
and https://psrtutorial.com/util/sorenson.html

And finally, if you prefer something more simple, here's a too I made for re-voicing:
http://digi-life.synology.me/software/YStyleRevoice.html
-it's meant for PSR keyboards in first place, but should be usable for Genos too as long "normal" styles & voices are used.

Bogdan

[andy_v]

Thank you Bogdan - that's really useful.
I'll give those tools a go and see which gives the clearest indication.
Am I right in thinking that there is a fixed MSB number for the expansion memory, and then each pack has its own LSB number within that?
If so, I suppose I'm just looking for styles which reference voices within that MSB/LSB; and then I can tell which voice was intended from the PC number.

[BogdanH]

Yeah, that MSB/LSB thing can be confusing when making custom pack (from several packs). From what I know (far from being an midi expert!), one need to pay attention on voices which have MSB 62 or 63, and if needed, you change that in YEM (for your final pack). When I say "if needed"... Again, YEM doesn't tell us there might be something wrong with our final pack -that is, you may have all needed voices in pack, however style doesn't "recognize" them.
That's why (among other reasons) I made Y-StyleRevoice. When my final pack is made, I open it with my tool and check styles inside pack again -and I can clearly see (without much thinking) if all voices are seen by style.

Bogdan

[Joe H]

andy_v,

I think there are issues with the LSB.  There is a couple of ways of dealing with this issue.  You can Export the Voices and Styles from each pack you want as "Content".  Then create a new pack and Import those Voices and styles from the various packs into a single pack.  Now... make sure you set the LSB on the new pack to whatever number you want.

The other option is to "untick" the duplicate Voices and save the packs, then reload them in YEM.  Now you will have several smaller packs. And you will probably have to remix some styles that use the oud and qanoun voices to get all your styles to play correctly. As you have stated; some of the Voices won't sound because of the LSB in those styles does not to match the pack Voices LSB. You can change the style LSB using MixMaster Batch Processing function to match the LSB to the pack Voices.

Hope this helps.

Joe H

[andy_v]

Thanks Joe H!

I thought MixMaster sounded like it would be useful but I didn't know about the batch processing capability - I shall certainly take a look at that too.