[sdiy] FPGA Synth Music

[Scott Gravenhorst]

Scott Nordlund <gsn10 at hotmail.com> wrote:
>
>>>Anyway I guess you probably know that the DX7, etc. used a single phase adder/sine ROM/
>>>accumulator thing with a 96 stage shift register to share it between all the operators.
>>>It's not hard to get different algorithms this way; the data is routed differently
>>>depending which operator is, uh, operating.

Actually, now that I read above again, it is very similar to what I'm doing (trying to do).  Though
not with a shift register, I do use the same logic to process the operators even in the current
synth.  The new design, based on something Tom Wiltshire described to me will make even better
re-use of logic.

I will look at routing possibilities, that sounds interesting.

>>
>> I know very little (close to nothing) about the DX7. My plan 
>>wasn't to clone it nor to develop a synth that does all of the 
>>algorithms of a DX7, rather just to get more than 2-OP 
>>generators. The scheme I'm considering will not do all of the 
>>DX7 algorithms, but it will do many of them and some that the 
>>DX7 cannot do. 
> 
>Sure, but my point was that FM was the first "interesting" and versatile form of digital
>synthesis to really hit the mass market, and for a good reason: 
>with clever implementation, it could be done very efficiently. 96 
>operators on a single VLSI is pretty good for 1983, right? 
>Everything else out there was either way more basic (Casio's 
>"consonant-vowel" stuff) and/or way more expensive (everything 
>else at the time). If you took advantage of some of Yamaha's 
>optimizations, you could have more operators and polyphony than 
>you know what to do with, and still at a really good sample rate. 
>Plus I think the "time sharing via shift register" thing does 
>make it easier to deal with multiple algorithms. 

Ok, you've convinced me.  I need to get back to this project.

I ought to be able to get 128 operators in a 400K or 500K gate Spartan 3A or 3E at 65.1 KHz.

>But, you know, as always the synth engine is kinda the easy part. 

Eh, well, I find it intriguing usually, sometimes frustrating.  Implementing something in such a
way as it fits and supports your feature list and voice count and sample rate...

>Finding a way to control all of that might be a real challenge.  
>People who make analog modulars take the easy way out, interface wise. 		

That has been a problem.  I take "an easy" way out using a separate serial connection to send data
from a PC where I can write my own GUI.  Certainly not as ergonomic as an analog setup, but I make
it functional.  

I also want to build a knobby synth out of my biggest dev board.


-- ScottG
________________________________________________________________________
-- Scott Gravenhorst
-- FPGA MIDI Synthesizer Information: home1.gte.net/res0658s/FPGA_synth/
-- FatMan: home1.gte.net/res0658s/fatman/
-- NonFatMan: home1.gte.net/res0658s/electronics/
-- When the going gets tough, the tough use the command line.