Re: [ARSCLIST] Dynamic pitch correction programs (was: wire recorders)

[Aaron Z Snyder <aaron@xxxxxxxxxxxx>]

Don Cox wrote:

> I think the problem in a disc that was recorded at varying rotation
> speeds is one of time correction, not pitch correction.

Indeed it is, so digital algorithms yield excellent results sonically. The
problem is finding an algorithm that will do the job automatically. So far
I've found that none of the automatic speed/pitch correctors will work
reliably, or even well, for that matter, with orchestral music. I've been
restoring recordings in which the speed variations are anything *but*
periodic and predictable, and about the only method I've come up with is a
manual one using the Cool Edit Pro/Adobe Audition pitch bender. It's a bit
of an art which one develops over time, although even when one knows how to
do it, occasionally the results are hit or miss. Fortunately I've found that
I can correct my own mistakes and end up with a good-sounding recording.
Multiple iterations of correction seem to result in no sonic anomalies
whatsoever! The real problem here is the amount of time necessary to fix the
recordings in question. It would be a big help if someone could develop an
algorithm which would do this sort of correction using serendipitous fixed
frequencies, such as hum, as the correction reference.

Aaron Z