On Jul 8, 2006, at 6:30 PM, Rob Danielson wrote:
> At 11:45 AM +0000 7/8/06, maxfrick78 wrote:
>> ...the conversion LR -> MS -> LR is fairly
>> simple and doesn't change a thing in the actual
>> source material.
>
> Have you tested this by chance? I've been
> thinking it might be prudent to confirm that an
> identical pair could be re-created if needed. Rob
>
It is highly unlikely to be identical. The conversion process involves
summing and renormalization. It is very likely that the low-order bits
will be different. But if you record 24-bit samples, the difference
doesn't matter.
Just the same, I am among those who usually record the M and S
channels. I like my recording to be as close as possible to the raw
source. But then I sometimes record the converted signal, but only if I
think I might need to play it back before I have a chance to process
it.
|