6 Apr
2004
6 Apr
'04
12:27 p.m.
Hi Daniel,
there are quite a few solutions for musically monophonic sounds out
there that work very well indeed. It's the complete mix of instruments
that gives them a hard time, which is exactly where ClearScale comes
in: it is intended for both musically monophonic and polyphonic
material, the goal being artifact free (or at least sonically pleasing)
time and pitch changing over at least one octave, or changing the
length by a factor 2 in both directions.
Additionally, formant correction is planned which avoids the common
"munchkinization syndrome" (Mickey-Mouse-Effect) when changing the
pitch.
Cheers,
Stephan
Am 06.04.2004 um 10:34 schrieb Daniel James:
There are a
few Open
Source solutions around. I no experience using them
but I suspect they don't compare with the design goals
of ClearScale.
Comparing the Audacity effect with the ClearScale example at:
http://www.clearscale.org/html/results.html
yields an obvious improvement. However, I've used the Audacity
timestretching tool on less complex material with good results, such
as single vocals or drum breaks.
I'm sure the Audacity team would be interested to check out
ClearScale - I'll cc this email to Dominic.
It should be announced on the LAD list where
Stefan
can get more feedback and publicity.
I've already suggested to Stephan that he does just that, because I
agree this is something that people on the lists will be interested
in.
By the way, I just noticed that the number of subscribers to the users
list has passed the number on the developers list for the first time
- is this a sign of something? There's a graph at:
http://www.linuxdj.com/audio/lad/subscribe.php3
Cheers
Daniel