From: Rob Danielson <>
>
> At 9:06 PM -0400 7/24/04, Walter Knapp wrote:
>
>> if so, the 24 bit equipment is wasting 4 bits per sample, they
>>won't contribute to the
>>accuracy. And that's under ideal conditions.
>
>
> One non-ideal condition is when recording distant sound sources in
> remote locations and overall sound file saturation is low. The extra
> bits contribute to some verfy useful resolution. Rob D.
Pohlmann's point is that, no they don't. The resolution you can get is
20 bit or thereabouts, limited by the analog part of the A/D. Regardless
of how many bits you have. The extra bits in 24 bit are resolving the
electronic noise of the components of the A/D. The reason I say that
it's ideal conditions is that we have only considered the error
introduced in the A/D, not all the rest of the analog chain before. We
are assuming it provided a perfect copy of the sound being recorded.
As I noted, once digital if you wish to preserve what accuracy you got
then before doing any processing you should resample to a much higher
bit count. It's that transition from analog to digital we are talking
about, not any processing beyond that. And since storing the data in a
recorder is not processing the data to change it, no need to up the bit
count for that.
Walt
________________________________________________________________________
________________________________________________________________________
|