At 11:37 AM 8/22/2003 -0400, Steven Smolian wrote:
>I'm afraid this issue is more serious that Mike suggests.
>Classical 78 sets and some doulble-sided pop selections have link-up
>problems at the side joins when pitches differ. Tapes made on Ampex 300 and
>350 transports which used large reels were different speeds as the weight
>shifted from one side to the other. I've been jarred by splices made from
>different takes recorded at different ends of the reel and spliced together.
>I hear very clearly differences in vocal timbre when the speed is wrong. The
>percentage difference seems to vary from voice to voice, spoken or sung.
>Students of historical records should be given access to materials as
>accurate as we can make them if they are to draw accurate conclusions from
>what they hear. It's a duty we owe the original performers.
I acknowledge that I simplified. Another factor of interest is that speed
correction should take place *before* digitizing. Failing that, sampling
should be at a substantially higher rate than is to be used in the final
product so that faults in resampling are minimized. The hardware resamplers
I know are not of high quality; software solutions can be relatively good
but are very slow.
[log in to unmask]