Rev2 Alternate Tunings - Part 2: OSC FREQ

[OakBloodThree]

The Manual Is Wrong!

What it says:

Osc Freq: Sets the base frequency of Oscillator 1 or 2 over a 9-octave range from 16 Hz to 8KHz (when used with the Transpose buttons). Adjustment is in semitones.

How it actually works:

OSC FREQ Sets the base frequency of Oscillator 1 or 2 based on the currently selected Tuning system selected Global Parameters.

When standard Equal Temperament is selected, OSC FREQ Sets the base frequency to the indicated note over a 9-octave range from 16 Hz to 8KHz (when used with the Transpose buttons). Adjustment is in semitones.

When alternate tunings are selected, OSC FREQ sets the base frequency to the frequency designated in the tuning table by the MIDI note number represented by the indicated note.

What it means:

As long as the OSC FREQ settings for both oscillators (all 4 in a Stack program) are set to the same value, there's no difference when using alternate tuning.

But for a tonal patch, the oscillators are often set in octaves, or other simple harmonic or near-harmonic relationships.

Examples:

For most equal tunings, you can make adjustments to get back to the same timbre: Say you had the two oscillators an octave apart.  Set tuning to 24-Tone Equal Temperament, and they will come out an equal-tempered semitone apart. So you set them to be two octaves apart and you’re good to go.

For non-octave equal tunings it's similar, but the math might get a little harder and you might need to use the fine adjustment.

What if you had the Oscillators tuned to C₂ and G₃+2¢ to add some extra 3rd harmonic (like the quint stop on an organ), and then you used an non-equal octave tuning?  Some notes will sound exactly as you want, but some clash with themselves!  There’s no way to correct because the space between notes varies through the scale and the difference between the oscillators won’t stay fixed.

Pick a tuning that’s non-equal and non-octave-repeating, and you might not be able to find any acceptable settings but with all oscillators set to the same frequency.

This highlights one difference between the DCOs (as programmed) and a VCO.  With a VCO, a volt is a volt, and an octave is an octave, so in a 1V/octave VCO, when oscillator 2 gets a control signal that’s 1 volt more than oscillator 1, they are an octave apart.  Not 12 MIDI notes apart, but their frequencies are in a 2:1 ratio.

It seems there could be a selectable mode on the DCOs where oscillator 2 is set in relative terms to oscillator 1.  I suppose I should write my first feature request.

[maxter]

Alternatively, if the modulation matrix could modulate pitch independently from the tuning table selected, with DC as source. This would allow for more tuning possibilities, enabling detune by more than one scale simultaneously, ie. Ability to transpose both by the chosen tuning table, aswell as equal temperament. "While you're at it", so to speak, in case SCI should implement it. Unless it's impossible, or would take too much programming to be considered. Otherwise a finer resolution (and/or downscaling/dividing, as proposed by creativespiral) in the mod-matrix for tuning destinations would probably make any other eventual shortcomings in this area overcomable, one way or another.

Anyway, it would make sense to at least somehow enable clean octave dividing in alternate scales. I imagine the potential for sound design being quite limited as is, when using alternate tunings, and this rendering many otherwise good patches less usable and in some cases maybe even unusable.

[OakBloodThree]

Say you had the two oscillators an octave apart.  Set tuning to 24-Tone Equal Temperament, and they will come out an equal-tempered semitone apart.

That should have said "equal-tempered tritone". In any case, dissonant.

[OakBloodThree]

A follow up to this for record, in case someone comes across the thread in the future:

Pym has posted an alternate version of the OS firmware that addresses the issues I described, and also gives a way to directly edit the scale from the Rev2 front panel.