it's not that simple. MS20 has 4 VCO pitch related mod inputs: VCO 1+2 CV IN, VCO 2 CV IN, TOTAL EXT MODULATION and VCO 1+2 EXT FREQ. VCO are used in linear (Volt/Hz) mode regarding keyboard CV (for octave up keyboard generates 2 times higher voltage, for 2 octaves up it's 4 times higher). After keyboard you have portamento and after that VCO 1+2 CV IN.hakey wrote:
per-oscillator pitch modulation
a) If you patch anything in VCO 1+2 CV IN, keyboard+portamento will be disconnected from both oscillators and wont affect osc pitch.
b) If you connect something in VCO 2 CV IN, only VCO 2 will be disconnected from keyb+porta.
In both situations octave switches and pitch knobs for oscillators will still work with external mod source patched.
Now, VCOs in MS synths have in essence both linear (V/Hz) and exponential (V/Oct) inputs. Linear inputs are driven by keyboard (this simply to avoid temperature drift issues, because dominant temp dependency is introduced by exponential V-I converter in VCO core), and exponential inputs are driven by TOTAL EXT MOD and VCO 1+2 EXT FREQ (if you use expo input just for modulation, that is over narrow range, lets say less than one octave, temp drift is much less of an issue than with multi-octave V/Oct VC like in Moogs). Both of these expo inputs are same regarding VCOs (TOTAl EXT MOD feeds VCF as well).
Exact replica of this MS20 behaviour would not be best approach in Diva. IMHO, this would cover all MS quirks and add much more, if added for all osc variations:
1) Option for linear VCO frequency modulation
2) Option to disconnect oscillators from main (MIDI) pitch input (that is, VCO frequency is not fixed regarding note being played). This is super nice thing when "fixed pitch" osc is synced to "keyboard follow" osc
{maybe you can do both of these already in Diva, I've only yesterday "found" duophonic mode, to my immense pleasure}
Further more, linear portamento could be added (when you do portamento on V/Hz CV and on V/Oct CV there is difference - glide up will be "faster" and glide down "slower" than in V/Oct - but it starts to become really noticeable on "wide" glides, like two octaves and more).
Other differences beside envelopes:
MS20 has that variable shape LFO unlike Diva and
Ring modulator in MS20 is CMOS XOR type (whatever you set as waveform shape for VCO 1, ringmod will spit out multiple of square waves), I think Diva is VCO2square*VCO1 (not sure, have to check when I get home)
