well its much the same argument with the matrix 1000 as it is with the commodore computers ..theres alot of people from the computing world which would object with horror at old commodores being completely trashed for a sound chip.
I’m one of them! I couldn’t bring myself to trash my old C64 either, and hate the idea of perfectly good hardware being ransacked for two chips, like a rhino for its horn, elephant for its tusks, shark for its fins, gorilla for its hands, etc. they’re all perfectly good animals! Of course the hardware isn’t alive (though some would argue otherwise 
) … nonetheless when I could buy a working Commodore on the 'bay for $15 and sell one chip inside for $20 to a wider market than the original hardware and at 1/8 to 1/10 my actual shipping cost for the C64 breadbox+PSU, the economic equation was pretty simple to junk the rest. It’s changing somewhat and C64 prices have risen, but there are still hundreds of thousands out there. No idea how many M1000 / M6R’s may be out there.
whereas adding a module to a m1000 in a way which it could be removed if it was really desired is pretty non destructive. and going the route that requires the original hardware, could help prevent the remaining stock of spare chips for the synths drying up so quick.
Agreed. Another issue here is the fact it IS old hardware and that stock would be needed for everyone, *including* the DIY-module builders. Keeping up the DIY-only, open-source philosophy would at least limit the commercialization, though speculative buyers are always a problem for any such market.
unfortunately speeding up the program alone wouldnt work, amoungst other things, the timing specifically relates to the midi & chip bus rates & without finding out where their ratio flags are; it would completely knock everything out of sync & out of touch.
far easier to take note of the 3396 or multiplexer chips (if fast enough to be useful) communication protocols & work a mios implementation from there.. which allows for the addition of things like boatloads of envelopes & modulation sources
Right, I mis-spoke a bit, in that I wasn’t implying use of the original firmware (which is for our purposes a “black box” with little to no documentation or meaningful source code), just an emulation of that processor if necessary for precise timing issues or whatever else, in order to avoid sync problems. I was assuming with a whole-number multiplier, stepping down the timings accordingly (i.e. processor runs at A, emulator runs at 3A, certain clock signals need to be emitted 3x slower than emulated CPU etc) would be fairly easy… then again who wants to work on that old processor? You’re probably right that the ‘brain’ can simply be replaced - of course whatever hardware was dependent upon idiosyncracies of its behavior or timing would have to be worked around. And I am guessing that they had to play with a lot of clever hardware tricks for performance to make this work back in the day it was engineered.
tho yeah ..what they got those chips to do in the first place was pretty impressive for the money.
Still is, when you think about it, for vintage pseudo-analog! These M1000’s go for $150-220 regularly on Ebay. It’s about the price range of the AN200 or a DarkStar XP2 territory. Those are good little synths but a lot more specialized/limited than a programmable M1000 in many respects.
so what the heck] not to support a bankstick or the like for patch storage. It could be hard to remove the biggest complaint about the Matrix-1000 (or at least the black-faced model like mine), the power-supply hum, but in terms of just the audio output of course, plugging it into my Firepod (balanced inputs and pretty good pre-amp) seems to take care of that quite nicely, and even the X-Station’s balanced stereo inputs cleans the signal of any noticeable hum. The unit itself, of course, still hums in the rack 
..however for critical performance, using another keyboard / a dedicated controller maybe advantageous.