The current pitch shifting algorithms seem to be flawed. Small minuite(but obvious) artifacts are introduced into the stream. It seems like the algorithm works on small blocks of the audio then patches them together wrong.
Wavelab uses the dirac algorithm and it seems to work flawlessly
Europe, Middle East and Africa