Just a note that if your show sampling frequency is 44100, the highest reliable, accurate audio frequency that the system will manage is 17KHz. Audio tones up to 20KHz are the result of system noise, inaccuracies and guesswork. So massive frequency conversions to a higher pitch are likely to generate garbage, not anything useful. I suspect Audacity is trying to figure out where all the foldover frequencies are and other artifacts from trying to do something impossible.
If you force it hard enough, the system will start generating whole new, clearly audible, non show-related frequencies as a result of fractional and odd multiple sampling.
I can confirm this bug, and it is something that I am currently working on.
Perhaps there are legitimate reasons, but sadly it may (depending on the % change that is required) be beyond the capability of “SoundTouch” (which is the shared library that Audacity uses for time/pitch stretching). Future versions of Change Pitch will probably not “solve” this problem, but will at least save the user from the “hang” that currently occurs by disallowing settings that are beyond its abilities.
To change pitch by a very large factor, do it in steps. For example, if you want a 50 semitone increase, you can do it as a 25 semitone increase then re-apply the effect for an additional 25 semitones.
