Emu XL-7 & MP-7 User's Group

112 MB (option 1) still sounds pretty good, many of the ROMs were 16 MB. 112MB is a HUGE amount of samples when they are so well edited as the E-Mu sets tend to be.

In my opinion, 8 hours for programming a full set via JTAG is perfectly acceptable. In practice, most people will only want to check out the samples in the different ROMs before settling on the 4 ROMs they want to keep in the machine, or 4 user-hacked ROMs combining the best of the rest. This checking out could conceivably be done via the Proteus VX software. Then you'd leave it over night to write the samples.

Because if you are constantly switching ROM sets, you'd have a hard time keeping track of which presets and sequences use what ROM, etc. Adding a bank of custom samples to address some shortcomings of the CS architecture and of the ROM sample sets, that would make the Command Stations almost like the perfect instrument. But the Command Stations and their UI was not made to be used like a sampler and switching out the samples for every track, so not having the option of quickly switching ROM sets doesn't sound like too much of a drawback, imho. Though it would be nice, and for people with rack mount Proteus 2000's it might be more useful.

Now, I'd really like to see option 4 and custom software running on the ARM, with realtime sample streaming and transformation, and listening to the data lines to do all sorts of clever stuff, extending the functionality of the Proteus architecture. I'd also like to see the operating system reverse engineered, the CS mainboard FPGA reprogrammed, etc etc :-) But software development always takes much much longer than you think, hardware + software is even worse, and reverse engineering takes forever!

The fact that options 1 or 2 look like they could actually become a reality within the foreseeable future is very exciting! Fantastic job, Jack Pratt!