Here we go again... 256 note polyphony??

[keynote]

I had to revise my whole post because of what I consider uncertainty regarding the polyphony issue. In trying to wrap my head around this issue I am learning more, I think, about how the polyphony is allocated on the Genos and how it functions in the real world. Now first of all the Genos has (utilizes) true stereo polyphony and that is according to Yamaha's own official statement.

That means when you play a "stereo" sample (voice) you're using only one note of polyphony not two as normally would be under the old system of allocation on the Tyros and Motif series, etc. I think Yamaha tech support staff were also unsure exactly how the polyphony worked on the Genos and the Montage which has led to some confusion among end users. The tech support representative I talked to at Yamaha USA wasn't too resolute in his answer but I came away with the impression the 256 was available across the entire range of the keyboard. But in reality it's sub-divided into two sections i.e. preset and expansion, albeit, under the new "stereo" system of allocation. So you have 128 notes available for the preset voices but under the new system of allocation those same 128 notes of polyphony go further because the stereo samples only use one note of polyphony instead of two under the old system. Once you reach the limit of polyphony available from the preset voices you can tap into the 128 notes of polyphony available under the expansion voices for a total of 256 Max.

PS: I noticed over at the Yamaha Synth forum that Montage owners also seemed confused about the polyphony and Bad Mister aka Phil Cleninden seemed evasive in his answers which seemed to lead to even more confusion. I wish Yamaha would come out and give us a direct answer so we can put this issue to rest once and for all. Montage owners seem to feel the same way I might add. Where is Yamaha when you really need them? Or perhaps my explanation says it all? If they want to hire me I'm pretty close geographically.

Mike

[Spirit of the old South]

Actually Montage polyphony is quite simple

128 stereo for AWM2 voices
128 for the FMX engine

Its obviously that both these instruments have a comparable construction
With 2 chips as the sound engine

Genos uses 1 for normal voices and the 2nd for expansion voices
Montage uses 1 for AWM2 voices and the other for FMX sounds

While identical chips, in both instruments, they don't seem to integrate well with eachother.

Another example of this is the 256 voice polyphony of the CVP 709
Where one chip is for all instruments with 128 voice polyphony
and the other chip is for Piano's only with another 128 voices of polyphony

It is just how Yamaha instruments work these days, its hardware based.

[keynote]

Spirit of the old South thank you for your input regarding this matter. The Montage has two sound engines i.e. AWM2 and FM-X but that is not the case with the Genos. It has only one sound engine. Whether they share the same chip(s) I'm not sure but they seem to work in a similar fashion. In other words, once you reach the limit with the Preset (on Genos) and the AWM2 engine (on Montage) you can then tap into polyphony from the expansion voices (on Genos) and FM-X on Montage. I wish Yamaha would've utilized a different approach where the polyphony was available across the entire range of the keyboard instead of separating it into two sections. Take for instance the Yamaha P-255. It has 256 note polyphony and costs $1,299.99 yet the polyphony is distributed across the entire range of the keyboard. In other words you don't have to worry about running out of polyphony in one section e.g. preset or expansion, AWM2 or FM-X, which necessitates having to tap into the other section(s) - if required - in order to achieve desired results. The Genos and Montage approach is unconventional to say the least and consequently it has led to some confusion on both the Yamaha Synth website and here at PSR Tutorial and on other forums too I might add. I'm as curious as anyone yet real definitive answers are hard to come by if you know what I mean.   I would love it if Heratch would chime in but Yamaha Japan seems to keep their employees on a tight leash unfortunately and therefore I don't expect any clarification from Yamaha USA regarding this matter. Oh well, it is what it is but perhaps Genos 2/Montage 2 will finally correct this anomaly and unconventional approach to polyphony allocation that is on the current generation.

Mike

[pjd]

Hi Mike --

I think I understand what you want -- the ability to play any voice (preset or expansion) through both tone generator (TG) chips thereby getting full 256 voice polyphony for any Genos voice. Did I get that right?

Here's a highly simplified diagram of the two tone generator integrated circuits (SWP70) in the Genos (Source: Genos service manual):

          SWP70 <----> Wave memory (2GBytes)       [Expansion]
            |
            + -----> DAC -----> Output
            |
          SWP70 <----> Wave memory (4GBytes)       [Preset]

Please think of the SWP70s as fully independent computers. Each SWP70 gets commands from the same main CPU which catches notes from the keyboard, MIDI, etc. However, each SWP70 runs as an independent hardware engine. The output from each SWP70 is mixed (digitally) and sent to a digital-to-analog converter.

The limitation in Genos -- 128 voice polyphony preset and 128 voice polyphony expansion -- is due to the memory organization. Each independent tone generator computer (SWP70) has its own dedicated wave memory. One wave memory holds the factory preset samples and the other wave memory holds the user/expansion samples. One TG chip cannot read samples from the other TG's wave memory.

This is why there are 128 voices of polyphony for preset voices and 128 voices of polyphony for expansion voices.

Hope this helps. I avoided any discussion of Montage since that wouldn't really clarify the Genos situation. I didn't give chapter and verse about why sharing isn't possible with the Genos memory technology. Also, suffice it so say, the memory organization is different in Tyros and Motif products with dual tone generation chips. Montage and Genos are a genuine break from Yamaha's past engineering practice.

All the best -- pj

Music technology blog: http://sandsoftwaresound.net/

[keynote]

Hi Mike --

I think I understand what you want -- the ability to play any voice (preset or expansion) through both tone generator (TG) chips thereby getting full 256 voice polyphony for any Genos voice. Did I get that right?

All the best -- pj

Music technology blog: http://sandsoftwaresound.net/

Hi Paul,

That's exactly right. To be able to play any voice whether preset or expansion and have all 256 notes of polyphony available at one's disposal. Perhaps Yamaha will rectify the current situation with Genos 2. Stereo polyphony is one thing and a nice addition I might add too. It goes a long way in solving any note drop off one could potentially experience otherwise under the old system. But having all 256 notes of polyphony available across the entire range of the keyboard would be optimal needless to say.   

All the best,

Mike

[Spirit of the old South]

Hi Paul,

That's exactly right. To be able to play any voice whether preset or expansion and have all 256 notes of polyphony available at one's disposal. Perhaps Yamaha will rectify the current situation with Genos 2. Stereo polyphony is one thing and a nice addition I might add too. It goes a long way in solving any note drop off one could potentially experience otherwise under the old system. But having all 256 notes of polyphony available across the entire range of the keyboard would be optimal needless to say.   

All the best,

Mike

Mike, as long as Yamaha is sticking to their current hardware (swp70) i highly doubt that is going to happen. Unless the chip is much much more powerfull then we currently see, which is something i doubt, because otherwise they would not have put 2 chips inside but a single one.

Also the remark about the P255 having 256 voices polyphony, that is also only because the sound engine is much less advanced and does not need to have 16 or 32 tracks capacity at the same time but only 2. So it can be done with a single soundchip.

Since we all seem to agree the problem is hardware related, this makes a software solution highly unrealistic.