View Single Post
  #77   Report Post  
Jay - atldigi
 
Posts: n/a
Default 16 bit vs 24 bit, 44.1khz vs 48 khz <-- please explain

In article ,
(Garthrr) wrote:

Ok, considering the post below, then the question is "Where is the
disagreement between the two camps-- the one camp who says 16 bit is
as good as 24 bit for anything but very, very low level audio and the
other camp that says 24 bit is better even at higher recording levels?
Where is the point at which the two camps begin to disagree?


It's not the "if", it's the "why". The subject is so complicated that
it's hard to give simple answers. For instance, 24 bits for delivery or
initial A to D is unneccessary because analog electronics (mics, pres,
consoles, the inputs of the ADC itself, outputs in the DAC, and
reproduction equipment) can't produce that kind of dynamic range, and no
practical recording or listening environment will ever be that quiet. So
there's the "marketing bits" problem right of the bat. This is not to be
confused with digital processing, where I want even MORE than 24 bits.

However, that point is unnecessary for a basic answer to your question;
but remember, there are lots of little things like that where you have
to say "it depends", or could offer a caveat or a distinction.

I think you're making the mistake of assuming that the "low level stuff"
means that if the average recording level is above 16 bits' -96dB limit
then the extra bits are somehow not helpful. That's not neccessarily so.
Unless you have steady state noise, the signal is constantly rising and
falling. Also, there are quiet components to a signal (or even a mix),
and there are louder components. If not, you'd never hear the voice
above the string pad. In some cases, depending on what you are
recording, and where, and with what, and how you play it back, little
details that are quiet unto themselves can still have a positive impact
on the performance you capture.

Just because they are quiet components doesn't mean they are useless
when there's anything else going on. Sure, sometimes they are masked and
are indeed useless, but sometimes they are not. Remember what I said
above? It depends... There are indeed some things that do not benefit
from more than 16 bits, but some things really can use more, and for
that reason, I definitely support the idea of A/D conversion, mixing,
and processing at higher than 16 bits. It's not, however, because the
loud signals are captured more accurately. It's because the more subtle
details (which are lower level signals) are also recorded.

--
Jay Frigoletto
Mastersuite
Los Angeles
promastering.com