[sdiy] various scales for a Midi-to-CV ??

[Florian E. Teply]

harrybissell at copper.net wrote:
> Are you talking about "tunings / temperment" or "scales"
> 
According to what you wrote below, i'll rather be talking about 
temperment than scales, even though both terms are (at least in my 
opinion) somewhat related to each other. What i mean with that wille 
become clear a couple of lines below.

> Temperment is the actual frquencies used...
> 
> "Pythagorean tuning" uses whole number ratios related to the
> harmonic series i.e. a fifth is a 3/2 frequency ratio
> 
> "Equal Temperment" has modified the frequency of the fifth (etc)
> to permit playing in ~any~ key
> 
> Some of the Arabic tunings you mentioned have non-western intervals...
> 
> Scales, on the other hand... are selections from the available
> pitches. Things like "major", "minor" etc have different half /
> whole steps.
> 
I absolutely agree with you here. Given an equal temperment, one could 
play different scales with all the same frequency values. Say you got 
some fixed frequencies according to C Major. With that same values, one 
could also play A Minor or all those different modes like d Dorian or F 
Lydian or e Phrygian and all that. But that's not quite the point i'd 
like to address.

> Assuming you want 'temperment' you could make up an offset table. Rev
> 3.2
> Prophet V allowed you to use the twelve knobs in the center of the
> unit
> to tweak the twelve notes +/- a small amount (one half strp iirc).
> 
> If you want to make 'scales' I'd use a MIDI-CV converter driving a 
> programmable A/D - D/A such as the Wiard-Blacet MiniWave.
> 
Sounds like a good idea to me. What i was thinking about was some sort 
of #2 just omitting the A/D part and doing soem calculations in a 
microprocessor instead so one can have a MIDI-CV converter with some 
selectable (and maybe user-defined) scales. What i'm not quite sure 
about is (given the CV-controlled oscillators around) whether or not it 
is worth all the computation of frequency values (and Voltages according 
to that) if the change in frequency is quite small (probably less than 1 
percent or less than a couple of millivolts for a 1V/octave law). Or in 
other terms, are the oscillators around good enough to get reproducible 
frequency shifts for those small changes in CV Voltage. Another (maybe 
more complicated) option would be to build some sort of MIDI-controlled 
DDS oscillator, which could have the advantage of some user-defined 
waveforms and frequencies using some sort of lookup table. Actually that 
was the thing i was thinking about initially. But with oscillators good 
enough around, there would be no need to reinvent the wheel, ya know ;-)

Greetings,
Florian