On Tue, Feb 01, 2005 at 05:44:24 +0200, Juhana Sadeharju wrote:
Hello.
Convolution reverb might not be a best approach.
The impulses I have seen (including at Noisevault) are
puzzling. One impulse has square wave impulses. Another
has exp decaying step impulse. Third has something close to
sinc impulse. What is the "correct" impulse type? How the
reverberations differ in the above examples if all have
sampled the same device?
It doesnt matter too much what the source impulse was, as long as its
wideband, you just deconvolve the output impulse wiht the source to
extract the dirac impulse response.
Worse: The big Eventide impulse response archive at
Noisevault
has a sample with good impulses in left channel but fuzzy
impulses in the right channel.
Thats not great, but you get good was well as bad impulses.
- Steve