[sdiy] Some Synergy info, was Re: Synclavier

[Veronica Merryfield]

All

The Synergy patent is numbered 4201105 although there are better explanations elsewhere. 

Speaking with Hal a couple of years ago now, the technology used was born out of the early digital phone exchanges, hence the origin of the patent being Bell Telephone Labs. I think Hal working working on filtering and compression. The actual use of this technology goes back way earlier than the 1980 patent but was only patented then because some bright spark in Bell realised that the synth industry was worth something.

Hal spoke at length on the spectra implications of adding phase shifted sines for amplitude control over that of the use of the multiplier, not just for the partials, but also when used as FM operators. In those days, the PCB real estate to create a multiplier was significantly larger than that for a phase shifted adder as well. He also talked about the phase alignment of the sine wave at note start. For fast rising envelope, this has a very marked effect on the rate of rise of the leading edge of the first wave and has a huge implication when used as a FM operator. I am not sure if this is unique to the Synergy and doubt it is but will say that a lot of control code is in place to ensure this is the case which may imply it isn't always the case.

I seem to remember there was mention of shift registers in this thread, but I might be remembering a different thread. The Synergy used the FPGA equivalent of a register file. The ALU, for want of a better term, cycled through memory locations to generate the output back to the register file. The output from one 'operator' could be fed to a few inputs of another very flexibly this way.

Various control and operational parameters were encoded in a log form to extend the dynamic range of values. These were mostly times and amplitude values but not exclusively so.

As Aaron mentioned earlier, the envelopes could be faded based on note (key position) and velocity. For it's time, this allowed some pretty good emulations of piano for instance. 

These days, an FPGA implementation would probably take more effort and logic to implement the phase control than using a multiple since many FPGA have the multiple built in, but it would be a fun exercise.

As also pointed out, the UI to these class of operator based synth is the key and close to impossible for something intuitive. 

As some of you may know, I started a soft version of the Synergy with a view to providing a emulation using original data and providing a modified mode. I have written the core engine and data set decode. It didn't take that much code to do but I stalled because of the UI. I can not figure out how one would even go about providing something that was intuitive. Before you ask, although I have a core and decoder, I have not put the two together in a AU or VST.

I also got side tracked into whether I could emulate a resonant sweepable filter without having a filter. I have figured how to do this for a harmonic partials timbre. However, sweepable noise is a completely different story. Noise is generated using over modulated FM and I can not yet see that it is possible to predictably control the noise spectra with such an arrangement.  

Veronica
 

On 2010-03-08, at 6:49 PM, Graham Atkins wrote:

> On 8 Mar 2010, at 22:13, karl dalen wrote:
>> 
>> Its interesting you mention first, where NED doing that FM thing
>> before Conbrio, DGS, Synergy etc etc or where they all similar in time?
>> 
>> Some NED service people claims that NED licensed Chowning/Yamaha
>> patent FM. But the Synergy oscillator was patented 1977.
> 
> According to Wikipedia the first Synclavier was working in 1975 although
> they had yet to market it.
> 
> Graham
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy