AW: AW: Mellotron/sample/wave type question.

[Haible Juergen]

	>In a mix it would be very hard to decide if it's an original
mellotron
	>or just a multisample, I think. If you do a mellotron only
recording,
	>well that's like with any other machine, if you listen carefull
there
	>will be differences, for shure. But maybe the latter situation is a
bit
	>academic or unusuall.

I never listened to a solo mellotron, it was always in a mix (;->).

I cannot imagine how you would play things like King Crimson's
"In the Wake Of Poseidon" or Genesis' "Watcher Of The Skies"
with a multisample and no one notice it. You're not done with
a separate sample for each note; you will need separate samples
for the same note, too, and a clever algorithm to model the effects
of different pressure on the keys (which has a direct mechanical 
effect on the tape transport). If you have a sampled mellotron, and
repeating the same note will produce the same sound twice, it's
so easy to detect it as a fake. Now let's assume you have 10 separate
8-second samples for each note - how do you interpolate between
these when your finger pressure changes while you play one note ?
You cannot sync the various samples, 'cause different pressure
means slightly different pitch, too, so you will need a clever algorithm
to adjust the phase of your 2nd sample during the crossfade.
I have no idea if it would sound like a mellotron if you were doing
all this - but I'm quite sure that it sounds different from a tron if you
*don't* do all this - and so different that you hear it in a mix easily.

JH.