Notifications
Clear all
Topic starter
Genos has a teaser video for v2.0 updates to come this winter. There are some interesting updates specific to that arranger platform which do not necessarily directly apply to the synth world (although some could - and one could argue the merits of each in synth-land. To me the most interesting development which "could" make its way to Montage is Genos's future increase in expansion memory from 1.8GB (that's a familiar value for Montage's sample memory) to 3.0GB. Nearly 2x increase. I'm not sure if the compression techniques (or other methods to increase available memory) in Genos is already in Montage or not. If not yet there - it'd be interesting to see a little less than 2x sample memory available for Montage and MODX and possibly squeak out some increase in the preset space available to include more native content (like CP uprights and other CP-exclusives, or any other new preset waveform content. Anything specific I list here will be biased and polarizing).
https://www.youtube.com/watch?v=KfrdEbbz1aM&feature=youtu%20be
... there's a bunch of blurred out "and more ..." items that may be interesting. The arranger is a different group and there's a bit of a great wall between the products. But the core hardware is very similar -- so cross-migration is possible although who knows what roadmaps and strategy will allow.
It's just interesting to try to read the tea leaves and think about how my investment may evolve.
Posted : 11/10/2019 3:51 pm
This comparison doesn't really make sense considering they have different hardware:
Genos
The main CPU is a 1GHz TI AM4376 Sitara Cortex-A9 ARM processor (AM4376BZDN100).
There are two SWP70 tone generator (TG) integrated circuits.
The master TG has 2GBytes (physical) of wave memory (Winbond W29N08GVSIAA).
The slave TG has 4GBytes (physical) of wave memory.
The internal memory is a Toshiba 64GByte eMMC device (THGBMGG9T4LBAIR).
- http://sandsoftwaresound.net/genos-internal-mem-speculation/
Montage
The main CPU is a 800MHz Texas Instruments Sitara ARM Cortex-A8 single core MPU (AM3352BZCZ80).
There are two SWP70 tone generator (TG) integrated circuits.
The master TG has 4GBytes (physical) of NAND flash wave memory (CypressS34ML08G101TFI000).
The slave TG has 0 GBytes (physical) of wave memory.
The Genos has 6GB of NAND flash whereas the Montage has 4GB.
Posted : 11/10/2019 8:21 pm
Topic starter
Certainly other things like super articulation may play some part in the possible memory consumption per atomic "PART" equivalent. So there are several differences - but I wonder if this was reserving memory (memory collecting dust on Genos until now) vs. a different encoding scheme to open up space. I don't have a memory map for usage of the hardware - the comment about same guts was more on a sub-component level. Not block diagram. Even at this level (how the "legos" are connected) - there's some ambiguity once you reach trying to assign a function for all the connected memory. I stick to the notion that if encoding plays a role here - then it's possible that similar techniques can migrate and that it's simply interesting to look across the product line and wonder. For the differences, the two use the same TG and same RAM technology. Physical memory part/size count differences are good to point out (you can't "grow" hardware that doesn't exist) - but one would have to have some better idea what those parts are used for than I currently see documented before one could call "foul".
Posted : 11/10/2019 9:40 pm
Thought I'd better drop a line.
The relevant components as far as (de)compression is concerned are the SWP70 tone generator ICs. These components handle the Yamaha proprietary compressed samples and the uncompressed user samples. Yes, one of the SWP70s performs FM-X synthesis and does not have a flash waveform memory connected to it. The other SWP in the Montage performs AWM2 synthesis, "sample playback", and, thus, has a flash waveform memory connected to it.
I'm one of the people speculating about the Genos V2.0 update on the PSR Tutorial Forum. Compressing user samples is one possible approach and reason for the increase in expansion memory capacity. Other members have speculated about reassignment of flash which is currently dedicated to audio style waveforms.
Fact is, folks, we really don't know and we are speculating (guessing) on the basis of very sparse information. Yamaha does not discuss its internal designs other than the patent descriptions which it publishes in Japan, the USA, and elsewhere. Even then, there is a "leap of faith" from the patent descriptions to real world products.
All the best -- pj
P.S. And the guy behind sandsoftwaresound.net 🙂
Posted : 12/10/2019 1:51 am
Topic starter
My original post quoting "could" denotes a high level of speculation. Seeing the trends in other products influences speculation. If nothing else - it shows that a product you purchase could "grow" memory through a firmware update. At least if you purchased a Genos in the past. And maybe others if similar roads are followed and the conditions are right.
Lots of "ifs" - but that was always how I presented the thread.
Posted : 12/10/2019 8:28 am
Hi Jason --
I definitely caught the "could." I just wanted to emphasize the speculative nature regarding Genos V2.0 and its increase in expansion memory. Plus, give at least one equally plausible explanation for the increase.
I also wanted to emphasize that people should take what I post on sandsoftwaresound.net with a grain of salt. I try to base my writing on service manuals and patents. However, that only goes so far. I do not have access to Yamaha design info ... or I would certainly be under a Non-Disclosure Agreement. 😀
It's really a check on my own intellectual honest...
All the best -- pj
Posted : 12/10/2019 5:09 pm
