[linux-audio-user] Speeding up Speech

[Dirk Jagdmann]

Phillip Blevins wrote:
> Ah, thanks for the suggestion, I had wondered about using a high
> sample rate at high speed and then slowing it down. Any idea how
> slowing the file down could be accomplished?

The trick here is to just record and don't make any conversion. How will 
then just need to alter the WAV header to the correct sample rate (this 
can be done on good audio software). I'll explain this closer.

We assume you have a tape running at 4x speed. Then you'll sample with 
88.2KHz. If you would play back this recorded sample with 22KHz then it 
will be played 4x too slow. But has we have recorded it 4x too fast 
you'll have the correct speed.

Every good audio/sample editor can adjust the sample rates *without* any 
conversion (so they just alter the values in the file headers). If you 
favourite sample editor insists on converting sample rates you should 
either get yourself a better program or just load a hex editor and 
change those 4 bytes for yourself (assuming .wav).

> As for my track splitting question - I've found tracksplit
> http://sourceforge.net/projects/tracksplit/  which allows me to
> specify a minimum track lenth the and the amount of silence to split
> on. This turns out to be perfect for voice wave data.

[I posted this answer again to LAU, as I think other might be interested 
in my anser as well.]

-- 
---> doj / cubic
----> http://cubic.org/~doj
-----> http://llg.cubic.org