How far can you time stretch whilst maintaing original pitch

[raydaledower]

I have a recording of a piano being dropped and disintegrating that I would like to slow to 1/40th of the original speed - this is to sync with film footage shot at 1000fps being played back at 25fps.

Have slowed sound by resampling, but it becomes largely inaudible due to the now ultra-low frequencies, any advice on how to maintain pitch successfully or the limits of this code much appreciated.
With thanks,
Raydale

[neuronaut]

There is no limit as to how far you can go, however, the nature of the time stretching algorithm you're using will become more and more apparent with larger stretch ratios. If you express the signal by a sum of sinusoids you will get exactly that, even if your original signal contained noise. The noise will then also be represented by a sum of sinusoids that vary relatively quickly in amplitude and frequency, which is different to what the human auditory system would expect in this case (= just noise).

Dirac has a limit on its time stretch factor (4x is max), so you will have to roll your own time stretcher, for instance using a phase vocoder or a sinusoidal resynthesizer, or use one of the programs that specialize in this (such as Paul's extreme sound stretch: http://hypermammut.sourceforge.net/paulstretch/)

Hope this helps!
+Stephan