naturerecordists
[Top] [All Lists]

Re: 16 bit vs 24 bit Information Theory Hypothesis

Subject: Re: 16 bit vs 24 bit Information Theory Hypothesis
From: "Dan Dugan" dandugan_1999
Date: Sat May 26, 2007 12:01 pm ((PDT))
George Paul, you asked,

>If we have more information, then why aren't we creating a
>representation of reality that is superior?

Good question, deserves a thoughtful answer. It's a matter of where
you draw the line between "good enough" and "overkill," and that's a
personal, artistic, and economic decision.

It's going to be very difficult to tell the difference between an
excellent 16/44 recording and a corresponding 24/96 recording, for
example. But there's a general principle in media production that the
master should be of the highest quality possible, so that it has a
quality edge that will help it survive the necessary processing to
produce a useful end-product. And "legs" for future distribution
media that might have higher quality.

24-bit recordings take up 1.5 times the file space of 16-bit
recordings. 96K recordings take up 2.18 times the file space of 44.1K
recordings. So the 24/96 recording will use  1.5 x 2.18 =3D 3.27 times
the file space of a 16/44 recording. The rule of thumb for 16/44
stereo is about 10MB/minute. 24/96 recording will use 33MB/minute.

Storage is becoming cheaper by the day, so that consideration isn't
as important as it was when digital was young.

Since converters top out at around 21 bits, 24-bit seems to have
sufficient overkill. Gives us some wiggle room for conservative
levels. I don't see that changing in the near future.

Stockham, inventor of digital audio recording, found that 50K was the
optimum sample rate--anything more would be overkill. So 96K is
already probably wasting a lot of space on no useful information, and
the only reason to go higher would be to extend bandwidth into the
supersonic, useful only in science work.

If I had an unlimited budget I'd record 24/96. But I don't, and like
Walt and others, I manage (most of the time) to make recordings that
are both scientifically useful and aesthetically pleasing with MD
(ATRAC encoded) recording that is 16/44 with a low-pass at about
Message: 16KHz.

Subject: 
-Dan Dugan




<Prev in Thread] Current Thread [Next in Thread>
Admin

The University of NSW School of Computer and Engineering takes no responsibility for the contents of this archive. It is purely a compilation of material sent by many people to the naturerecordists mailing list. It has not been checked for accuracy nor its content verified in any way. If you wish to get material removed from the archive or have other queries about the archive e-mail Andrew Taylor at this address: andrewt@cse.unsw.EDU.AU