Prophet 2000

WeAreAs1 at aol.com WeAreAs1 at aol.com
Sat Feb 15 09:31:33 CET 1997


jocke at netcontrol.fi (Joachim Verghese) wrote:

<< I think the main reason why the P2000 sounds good is because of the
uncompromised sample read-out and DAC mechanism.  First of all, each of the
eight voices has its own 12-bit DAC - there's no demultiplexing of analogue
signals. >>

Yep, that's probably the main reason.

<< Second, the P2000 uses variable rate playback, not the add/drop (or linear
interpolation) method which was commonly used in fixed-playback-rate designs.
In other words, less digital distortion. >>

The big trade-off, in this case, is that the P2000 doesn't really have a very
wide pitch transposition range.  When using the highest sample rate (I think
it was about 40kHz??), you could transpose samples only 7 semitones upward.
 The downward pitch transposition range was not limited, but even using the
P2000's tracking filters, there was still a ton of aliasing junk in the
sound.  This was to be expected, given the system they used.  It worked great
for non-transposed samples though, such as drums/percussion and sampled vocal
phrases and sound effects.

<< Obviously, fixed-playback-rate sample read-out can sound good too, but you
need to implement some pretty advanced DSP algortihms for that.  And 24 or
32-bit DSP ("internal") precision is essential. >>

The Roland S-50/330/550 12-bit machines used the fixed-playback-rate system
(with a lot of add/drop sample interpolating to transpose pitch), and hence,
was able to transpose over an incredibly wide pitch range, and with much less
aliasing noise at the low end.

The Prophet sounded better, though...



More information about the Synth-diy mailing list