[sdiy] New DIY synth has born

[Thorsten Klose]

Hello,

Rainer already explained a lot of details about reading the SID registers, 
so I only have to add why my application doesn't get use of it:
I'm able to get much more accurate pot values with the internal ADC
of the PIC, and the OSC3/ENV3 values are not for interest, since 
MBSID provides 6 software implemented LFOs and 2 SW envelopes
with 16 bit resolution. The big advantage of this higher resolution is, that
the filter can be sweeped without crackling noise, and non-linear envelopes
are also possible for extreme punchy sounds, see:
   http://www.ucapps.de/midibox_sid/mbsid_curve2.jpg
   http://www.ucapps.de/midibox_sid/mbsid_curve3.jpg
(especially when an external 24db filter is added via CV out)

On Wednesday 16 February 2005 16:49, Rainer Buchty wrote:
> If you want to force oscillators to reset, there's a TEST bit for that 
> which sometimes is also needed to re-start the oscillators after 
> freezing it using "inappropriate" waveform settings. Flicking that bit 
> could also be used to manually hard-sync an oscillator to external 
> events.

thats correct, triggering the TEST flag is nice for percussions and bass,
especially when all oscillators are initialized with the same settings.
This can lead to some (nice) distortion at the SID filter stage - and
we all love such non-emulatable analog effects, no? :)

> Changing waveforms IIRC only became effective with also changing the
> gate bit, but maybe Thorsten can elaborate a bit more on this as he's
> implemented waveform sequencing.

Fortunately changes become active immediately. 

On Wednesday 16 February 2005 15:27, Batz Goodfortune wrote:
> Admittedly, at the time I was thinking of a box with 8 SID chips arranged 
> in 4 pairs for stereo. Each pair also having it's own OPL3. But this would 
> have given an interface for 16 pots and some really fine stereo blending. 
> But regardless of which, I certainly don't want to have to use a 6502. I 
> don't have anything against 6502s per se but I just don't want to have to 
> ever use one.

I'm not sure if it's worth the effort to use an 8bit CPU for controling so much
SIDs and OPL3s at once. Such a synth doesn't only live from a static
initialization of the sound registers, the fun begines once multiple
parameters are modulated by different sources (LFOs, EGs, 
"wavetable sequencers") and this requires a lot of horsepower

So, either the parallel approach will help (where small micros are 
prefered), or a 16bit uC with an external bus interface (for fastest
write access to the SID/OPL3). 16bit because of the sufficient resolution
and smaller memory consumption compared to 32bit controllers

Best Regards, Thorsten.