Subject: Re: [linux-audio-user] LADSPA Module
From: Jason (hormonex_AT_yankthechain.com)
Date: Mon Dec 10 2001 - 20:20:18 EET
true, but if you're only changing the tempo a little bit, it's faster to
do it that way, and up to a certain point it sounds better. My evidence
for this is that I happen to know that MOTU does it the way I described in
digital performer, and their time compression sounds much better to my
ears in most cases than most of the third party algorithms which take a
different approach, in which there are often glaring artifacts. The only
time I can imagine there really being a problem is if you were doing the
same amount of time change to multiple files with a lot of common
information, in which case, a great deal of phasing would occur.
On Sun, 9 Dec 2001, Steve Harris wrote:
> On Sat, Dec 08, 2001 at 04:40:24 -0500, Jason wrote:
> > oh, time compression. you just need an algorithm that searches for zero
> > crossings in a wave file, creates region markers at the crosssing, and
> > then crossfades the regions together.
>
> Thant only works if you don't care about the realtions of the harmonics,
> if you do then you need to use a FFT (or Wavelet?) style appraoch.
>
> - Steve
>
-- YankTheChain.com - You can pretend we're not here. That's what I do.,
This archive was generated by hypermail 2b28 : Mon Dec 10 2001 - 20:13:58 EET