On Mon, 2007-03-05 at 11:53 +1100, Erik de Castro Lopo wrote: > > This is basically an existance proof that there are normalised floating > point numbers that will not survive the > > float32 -> pcm24 -> float 32 > > transformation. i stand humbly corrected. i guess i confused the fidelity of pcm24 -> float32 -> pcm24 with its "inverse". thanks for the clarification. --p
