Archive of the former Yahoo!Groups mailing list: MOTM

previous by date index next by date
previous in topic topic list next in topic

Subject: OT kawai additive info + (weird) module idea

From: norman fay <NFAY@...>
Date: 2000-12-05

In message <4FC23B4FFF62D311854F00508B08E81DD7FD31@...
m>, Brousseau, Paul E (Paul) <noise@...> writes
>So they do it the (ouch) hard way. No wonder they were never entirely
>successful.
>
>I suppose the problem with bunching oscillators is that you're assuming that
>you want the oscillators bunched at those fixed intervals. I could see
>having them run at ill-tempered (heh) intervals for some nicely clangy
>sounds (bells, anyone?).
>
>Perhaps you could have oscillators bunched together, and them manipulate how
>the bunches are offset from each other. The oscillators ∗within∗ each bunch
>would be tempered to particular intervals. Or is this what you had in mind
>to begin with? (Sorry for the poor description, I'm writing
>free-thought...)
>
It depends on wether you have the K5 (which I have) or thee K5000 (which
I want. REAL bad) On the K5, you get two "oscillators", each with 63
harmonics. Each "oscillator" has four envelope generators, and you
assign each harmonic to one of these. There are also two independent
pitch envelopes (loopable) two independent envelopes for the weird
digital filter, and two for the "VCA" stage, so that's (count on
fingers) FOURTEEN envelopes
(okay, I had to count up with a couple of toes too ;) )
is that enough?
Obviously not, because each 64-harmonic voice channel in thee K5000 has
INDEPENDENT ENVELOPES FOR EACH HARMONIC!! that's as well as pitch, lo
pass filter, formant filter and amp envelopes, of course.
These instruments rule, and to be brutally honest, I think this sort of
synthesis just isn't practical in analogue hardware form. There was a
cool PC editor for the K5 called "overtone" which allowed you to extract
harmonic spectra from samples and translate them into K5 patches, but to
the best of my knowledge, no such utility exists for the '5000, which is
a shame.
I think a better MOTM additive-type waveform module would be an
implementation of the GRAPHIC OSCILLATOR - 12 or 16 sliders, VC fast lag
time to smooth off those pointy edges and scaleable modulation inputs
for each slider. This would RULE, but even then, you'd have at least 16
sliders, 18 knobs 19 sockets and a big panel = big $$$. What the hell,
I'd buy one...eventually...

I think that ruediger lorenz guy had something like this in his
(awesome-looking) home-brewed modular, BTW...

best etc
--
norman fay