<<i disagree. the real advantage of 24 bit is in capture. the extra
dynamic range - there is some, only not 144 db of it - really makes
it possible to do very good field recordings. think of setting up so
the loudest sound will not overload and be sure even of the quietest
ambience is captureed above noise level. changing to 24 bit later on
provides nothing.>>
Umashankar:
I'm not sure what you're disagreeing with here, since you're making
the same point I was advocating, i.e. initial recording at 24 bit, &
the futility of adding empty bits to a 16 bit word prior to DAW post
processing.
<< and if some processing requires a larger number of bits (as in
some kinds of filtering) software these days scales the samples
automatically to create the headroom. >>
Actually, all processing within a DAW involves a math operation upon
the original values in the digital word, & thus a deeper bit depth for
any DSP is needed to avoid degradation from rounding off resulting
fractional numbers. And yes, it is built in to all DAWs so we don't
ever have to think about it.
<<most software mixes can be set to be done in 32 bit, for instance.>>
Selecting 32 bits for a mix still involves a conversion process, I
believe, since you're going from 32 bit floating point (or 48 bit
fixed point in ProTools) to 32 bit fixed point words in the mix file.
In other words, you're not truly preserving the same 32 bit word
values from the internal data bus. At this point I question if there
is any point in presenting these extra bits to the DAC, as opposed to
a 24 bit word, which also contains unresolvable bits.
Scott Fraser
|