I tried the process in Wavelab.
Here is my file:
https://dl.dropboxusercontent.com/u/22100835/Material.mp3
Isn't that what you'd end up with if you digitized a 30 minute cassette
side,
if the recording speed had gone slower during the recording?
I didn't use an analog recorder for this file, I made a demo file in the
DAW.
At the end of the recording the pitch is 16 semitones higher than in the
beginning.
The test file duration is 29'26".
Here is the result of the WaveLab "Pitch Bend" process:
https://dl.dropboxusercontent.com/u/22100835/Lowered%20speed.mp3
The result is 45'26", which is quite close to my original starting
audio, 45'06".
Could have tweaked it.
The Wavelab process doesn't seem to be linear. The pitch lowered in the
middle
of the sequence, but was ok in the end. I adjusted the contour to
correct this.
Untick the "Preserve duration" box in the Pitch Bend window.
The maximum range of the WaveLab Pitch Bend is 36 semitones.
If you just need to make the interwiew easier to listen to, this seems
to work.
In practise, you had better first find by ear and using the speed adjustment
the needed speed correction at the end of the recording and in a couple
of spots
along the way. The recording speed most likely hasn't dropped linearly
and it
drops steeper towards the end. The cassette recorder speed control has
tried to
keep it up, but when the battery voltage drops, the speed will also slow
down
badly.
Did I miss something?
An automated process would surely speed up the job.
Eero
-------------------------------------
http://www.tonfiks.fi/
www.facebook.com/tonfiks/
|