[sdiy] CPLD shield for arduino/AMANI 64/AMANI GT

[Scott Nordlund]

> From: tom at electricdruid.net
> Date: Tue, 27 Mar 2012 11:09:21 +0100
> To: brockr0 at shaw.ca
> CC: synth-diy at dropmix.xs4all.nl
> Subject: Re: [sdiy] CPLD shield for arduino/AMANI 64/AMANI GT
> 
> Whilst I applaud the effort and ingenuity that is going into these attempts to build modern top-octave generators, I don't understand the need.
> The switch from separate master oscillators to one master osc and a top-octave generator chip caused a notable flattening in the sound of the instruments that did it, since everything got phase-locked together.
> Wouldn't we be better off putting twelve surface mount oscillators on a chip? Or are you guys so far ahead of the original clock speed that it stops being such an issue? What are you intending to do with these chips? Just curious, not criticising.
> 
> T.>
What's the advantage of 12 master oscillators? Are you thinking of independently applying vibrato to each one, or imagining them drifting over time? The crappy phase locked sound is due to the octave dividers, not the TOS. I can't really imagine anyone subconsciously perceiving some sort of least common multiple "super period" in the sound and suddenly exclaiming, "Hey! That's not an irrational frequency ratio!" The only intervals that can really be "fudged" that way are the fourths and fifths, but I don't think any of them are actually a perfect 4/3 or 3/2, at least in the MK50240. It would be electronically convenient, but you'd end up with a weird sort of quasi-Pythagorean scale.   
On a not really related note, I recently found that the Yamaha PAS organs (E70, D85, etc.) deliberately mistune all notes by a different amount to avoid any perception of phase locking and to make octaves, coupled keyboards, etc. much thicker sounding. But some notes are actually detuned quite a lot, so it seems to have introduced some "wolf" intervals.