Old Motifator threads are available in the Archive.
Jote
Total Posts: 1549
Joined 07-29-2002 status: Guru |
https://www.instagram.com/richarddevine/ Three short FM-X snippets |
davlippo7
Total Posts: 252
Joined 10-06-2010 status: Enthusiast |
Check it...... |
Jote
Total Posts: 1549
Joined 07-29-2002 status: Guru |
Speaking of Ujiie, his demo just got the English subtitles |
DmitryKo
Total Posts: 1483
Joined 07-25-2002 status: Guru |
Korg Kronos has a 12 GByte preset ROM (of which 6 Gbyte is EXs12, the SGX-1 “Austrian Grand"). These samples are stored on the internal SSD, as already mentioned, and streamed from disk on demand, using the “preload" bank which contains program data and initial sample data.
Nope, they just need to design their tone generator chips with 64-bit memory adressing and integrate a compatible flash memory controller.
And if the current trend continues, in 3-4 years 32 Gbytes of integrated flash RAM will the the norm in phones and other embedded devices instead of today’s 8 GBytes. That’s 80 Gbytes of samples if using Yamaha’s LPC compression (2.5:1 ratio) - about 975 times as much as on the original Yamaha Motif from 2001.
Yamaha has no incentive to do this, they are better off with dedicated VLSI circuits.
They have product portfolio ranging from very expensive digital mixers like the Rivage PM10 ($200 000) to cheap PSR-E253 portable keyboard ($170). All of these use the same or very similar proprietary VLSI chips, so Yamaha can sustain fabrication costs for large batches of these.
Nope, 4 Gbytes of flash RAM. 5.67 GBytes is a figure for 16-bit linear format (i.e. uncompressed original PCM) - the physical size is 2.25 Gbytes assuming 2.5:1 LPC compression. 1.75 Gbytes is, conversely, physical size - which would give you an equivalent of 4.38 Gytes, assuming 2.5:1 LPC compression. To pity only Yamaha sample libraries seem to use compression, as they use it as a form of DRM/copy protection and hold their compression tools to themselves. |
DmitryKo
Total Posts: 1483
Joined 07-25-2002 status: Guru |
Mostly because there are many different variations for XA control.
Again, that’s triple the work - not just make large new samples and relevant programs, but also upgrade old samples and ensure their compatibility with old programs, all for a mild perceived quality difference - which could even be a nuisance for those who depend on these waveforms to sound exactly the same. Yamaha clearly follows the same rule for Waveforms as with Voice/Performance names - if it’s the exact same Waveform name, then it’s the exact same Waveform. We shall have further information when the Data List for Montage is released.
I’d rather put the resources and effort into sampling the newest Yamaha CFX concert grand (2010, 9’, $150,000) rather than “upgrade” already existing sample set for the CFIIIS (1991, 9’, $125,000) and S6 (1994, 6’11”, $70,000) / S700 (1994, 7’5”, $250,000). From all accounts, it’s exactly what Yamaha did in the past with the CFIIIS samples on the Motif ES (~40 Mbyte sample set from the P120/P90) and S700 samples for the Motif XF (~90 Mbyte sample set from the S90XS).
Not exactly. Orchestral libraries have separate samples (programs) for every intricate nuance of playing technique. But they’re still one-dimensional snapshots - you can’t really change the articulations while you are playing, unless you select a new program. It’s not something you’d want on a live performance synth. On the other hand, the Motif/Montage use clever programming tricks to make these articulations playable live in realtime.
It has 10x the sample data but the same number of Motif XF waveforms (3997) plus ~3000 new waveforms for a total of 6900. According to my 5th grade math class, it’s not possible to have 10x the sample data just for old waveforms and have 10x the sample data for all waveforms (both old and new) at the same time…
The numbers like “PRE01 G08” had very specific practical use because you could quick-dial them with dedicated buttons. There is no use for Performance numbers if you can’t dial them anywhere. |
davlippo7
Total Posts: 252
Joined 10-06-2010 status: Enthusiast |
“ Nope, 4 Gbytes of flash RAM. 5.67 GBytes is a figure for 16-bit linear format (i.e. uncompressed original PCM) - the physical size is 2.25 Gbytes assuming 2.5:1 LPC compression. 1.75 Gbytes is, conversely, physical size - which would give you an equivalent of 4.38 Gytes, assuming 2.5:1 LPC compression. To pity only Yamaha sample libraries seem to use compression, as they use it as a form of DRM/copy protection and hold their compression tools to themselves. “ So the 5.67GB is actually an inflated number to appear more impressive but in relative terms considering the compression that’s a fair number to advertise. The good thing about the physical of 1.75GB is that if we purchase Yamaha expansion packs it will be compressed so in relative terms we are actually getting a relative 4.38GB of space. The bad news is that If we load our own custom stuff then we’re not compressed and could be using up valuable space quickly. |
dsetto
Total Posts: 434
Joined 01-24-2014 status: Enthusiast |
You sure about expansion packs intel? |
Jote
Total Posts: 1549
Joined 07-29-2002 status: Guru |
Beg to differ. There’s still use if one finds it easier to memorize numbers than names. Even the “G08” format has its numeric equivalent of “104”, that can’t be really dialled anywhere on the machine itself. But it’s there, and it’s displayed on the screen (the “AXX” that can actually be dialled is actually in parentheses!) |
DmitryKo
Total Posts: 1483
Joined 07-25-2002 status: Guru |
It serves as MIDI Program Change number - I’d guess these will be provided in the Data List and insrument definition files for the sequencing software. |
lordbachus
Total Posts: 85
Joined 08-28-2009 status: Experienced |
Then Yamaha needs to deliver a compression tool, so we can compress our own stuff on a pc to the efficient awm2 format |
Jote
Total Posts: 1549
Joined 07-29-2002 status: Guru |
I don’t think Yamaha is keen on releasing their proprietary compression algorithm into the wild. |
davlippo7
Total Posts: 252
Joined 10-06-2010 status: Enthusiast |
Memory used to be coveted and scarce and extremely expensive. Thus the “proprietary compression algorithm” was born. Nowadays 4GB of flash is chump change. We all kinda thought that the Montage had somewhere around 8GB but apparently it’s only 4GB. I know it’s high speed but how much more to double that size since it’s not expandable? Somewhere on the budget chopping block it got cut out. This doesn’t make much sense with the money making expansion packs on the way which I expect very similar to the Tyros system in musicsoft.com. “Vintage Keys” due at release along with the Bos piano. How many times over would that doubled flash expenditure have paid dividends to expansion geeks like myself. I’m a sucker for those things. I think I bought 75% of the Tyros 5 expansion packs at $160 to $200 a pop. Now I’m concerned about running out of space and I don’t even own one yet. |
lordbachus
Total Posts: 85
Joined 08-28-2009 status: Experienced |
https://www.youtube.com/watch?v=jObR8avIAkY listen to the Latin Flute at 6:10 Thats technollogy i only know from SA2 voices in Tyros 5, EAM technollogy.. Now they got me wondering if the Tyros 5 ensemble technollogy is also hidden somewhere deep into the instrument… |
Jote
Total Posts: 1549
Joined 07-29-2002 status: Guru |
I don’t think (I’m pretty certain as a matter of fact) this has much to do with AEM. The “trill” is an arpeggio (it was mentioned in another video). XS has a very similar flute sound that works in a similar manner, but the arpeggio is velocity triggered. This one is a bit weirder because it does require you to hit two keys, but only one note is actually used which leads me to believe that the arpeggio is simply bound to an element set to “legato” (which itself is not new). So, nothing special really. |
DmitryKo
Total Posts: 1483
Joined 07-25-2002 status: Guru |
We know why it was used with 1980s/1990s EEPROM and 2000s flash memory, but today in 2016 a 16 GByte NAND flash memory chip is US$5. I’ll refer you to chickensys.com/products2/translator/formats.html
|