[linux-audio-user] Acid for Linux ? - Pitch shift

Mark Knecht mknecht at controlnet.com
Tue Dec 10 16:26:00 EST 2002


Jesse,
   It is, to the best of my knowledge, a misnomer that is flowing around in
this thread that Acid time-stretches anything in real time. I do not know
that to be the case at all, but also admit that I haven't tried to do it.

   My use of Acid has focused on the use of specific samples at the speed
they were recorded at. If the drum library loops were 145BPM, my song will
be at 145BPM. I think the likelihood of time-stretching a sample and having
it be sonically interesting is pretty minimal, but I'll happily respond to
the application that does it first! ;-)

   Acid does have 'Beat Mapper'. Most of us gave up depending on it a while
ago. It's a misnomer that you can use it on too much recorded material as
real musicians vary their tempo by miniscule amounts anyway, so it isn't
constant enough to depend on for that. It can be helpful in grabbing a
specific section of a song and helping you cut a loop out of it, but I
typically don't have the patience to do that on a regular basis when I can
by a loop library that has 500 loops for $50.

   I don't know which one of you, yourself or Steve, is smarter, but I'm in
awe of both of you, as well as about a dozen others here.

Cheers,
Mark

-----Original Message-----
From: Jesse Chappell [mailto:jesse at essej.net]
Sent: Tuesday, December 10, 2002 12:57 PM
To: linux-audio-user at music.columbia.edu; Mark Knecht
Subject: RE: [linux-audio-user] Acid for Linux ? - Pitch shift


Mark Knecht wrote on Tue, 10-Dec-2002:

 > Acid has the ability to shift samples by known amounts, so that if I have
a
 > bass line in C, I can shift it up a major 4th to F and it will fit into
the
 > sound quite easily. This function in Acid works pretty well over a
limited
 > range, say +/- a major 6th, but starts to fall apart pretty badly at an
 > octave, in my experience. (It depends a lot on the library.)
 >
 > Until this morning I hadn't considered using freqtweak to accomplish the
 > same thing. Today I would have to take an input wave file for the base
line,
 > then shift it appropriately and write it back to disk as a separate wave
 > file, using more disk space and creating more problems at the app level
 > dealing with yet another wave file to be read and played back.

This already exists in the form of Steve Harris's "pitch scaler"
LADSPA plugins.  In fact, the algorithm used in freqtweak is based
on the one in Steve's (ripped right out of there in fact).

Time-stretching is a showstopper for LADSPA however since
#samples must equal #samples out.....  I think this has been
discussed before....

jlc






More information about the Linux-audio-user mailing list