The cleanest way i’ve ever done this (changing the sample rate and replay speed/pitch) was by rewriting the WAV file header with another sample rate, which then makes the software either dynamically resample (usually with interpolation) or resample on import to match the project sample rate, in both cases ending at the speed related to the hacked sample rate. That was the only way I could align recordings from two different recorders, for surround use - had to set one up at 48004Hz. I’ve done this with an arcane piece of software called SoundHack, but anything that messes with the file header (metadata only) should do this for you.
7 Likes
Ah this is interesting, do you know how to calculate sample rate if I want to transpose by a semitone or two down?
Thanks for your feedback yes varispeed seems to work for this material and with Ircam Lab transposed audio files sounds noticeably better
I think it would be something like:
newSr = oldSr * 2^(p / 12)
with p being the number of semitones.
for example to transpose down a whole step:
48000 * 2^(-2 / 12) = 42763.13847073629
which you would probably want to round to 42763.
This is the formula for transposing a frequency by a certain number of semitones in equal temperament, but it should work for sampling rate as well.
4 Likes
Thanks so much much! This is actually very elegant and fast way of doing the transposing and it works beautifully with SoundHack. I have around hundred files to transform so it saves some time too…
1 Like
really nice track. Also, cool to hear about the process
1 Like
Sounds great, thanks - would love to try that module but I guess I have to limit to software.
1 Like