[sdiy] DX7 hacking?

Magnus Danielson magnus at rubidium.dyndns.org
Sat Jun 30 21:49:24 CEST 2012

On 06/30/2012 09:06 PM, Scott Nordlund wrote:
>> When the DX7 came out I was one of those in awe.  When I compared it to my
>> simple modular I sort of just gave up on modulars and building and just
>> spent time playing.  Then I started buying analog synths in need of repair,
>> fixing them, playing them for awhile, and then sticking them in the garage
>> with the rest.  As analogs were not as well favored as digitals back then,
>> analogs needing repair were affordable - so it was a satisfying pastime.
>> Going through all this helped me learn a bit about synths that I never would
>> have been exposed to earlier - yet alone have the money to buy back then.
>> So anyway, as I progressed through this I occasionally went back to that DX7
>> and enjoyed its unique character but hearing it in most every patch was
>> audibly boring.  I didn't get into effects devices like Eventide or the like
>> until much later.  I always expected someone to hack the DX7 into more of a
>> monster.
>> I never saw it happen.  Has anyone ever made any progress or attempts? -
>> something that gives it a sound character other than FM?  Is the design too
>> "purposed" for FM to alter?
>> I'm sure I could just forget about it and realize that newer synths probably
>> have the FM as just one sound option.  I just don't have anything that new.
>> Barry>
> The DX7's sound production is pretty much hard wired. The main ASIC contains all the algorithms, registers and lookup tables. You do, however, have access to the amplitude data for each operator since it's passed from the envelope generator ASIC. The DX1 actually implements polyphonic aftertouch by modifying this data. But aside from that, I don't think there's much more to hack by modifying the hardware. You should instead be looking for alternate synthesis techniques that are similar to FM. In the 70s and 80s there were plenty of proposed algorithms that were discussed in journals and used in computer music, i.e. J. A. Moorer's Discrete Summation Formulas, waveshaping, asymmetric FM, formant synthesis techniques (FOF, VOSIM), etc. Yamaha also had a large number of patents covering similar things. I think most of these were never implemented commercially, though some products (Casio's CZ series) borrowed some from them. These techniques aren't exactly cutting edge anymore,
>   ut they can be pretty easily implemented in software. Yamaha's "AFM" synths (SY77, SY99, TG77) also usefully extended FM to include 3 feedback loops, looping envelopes independently adjustable pitch modulation depth for each operator. And I haven't used one, but the FS1r has 8 operators and includes formant operators. 		 	   		

I can't recall that I've seen much detail on the actual ASIC and its 
registers. Care to give a reference?


More information about the Synth-diy mailing list