home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!att!att!allegra!alice!jj
- From: jj@alice.att.com (jj, audio curmudgeon)
- Newsgroups: rec.audio
- Subject: Re: 20 or 24 bit digital recording?
- Message-ID: <23346@alice.att.com>
- Date: 29 Jul 92 14:22:47 GMT
- Article-I.D.: alice.23346
- References: <23292@alice.att.com> <1370@eagle.ukc.ac.uk> <23324@alice.att.com> <1992Jul28.230217.4309@fid.morgan.com>
- Reply-To: jj@alice.UUCP (jj, audio curmudgeon)
- Organization: NJ State Home for Bewildered Terminals
- Lines: 43
-
- In article <1992Jul28.230217.4309@fid.morgan.com> sethb@fid.morgan.com (Seth Breidbart) writes:
-
- >That brings to mind an interesting question. DAT uses 16 bits, so it
- >has 96 dB snr for a full level signal. For a signal at -50 db, it has
- >46 dB snr. For a signal at -90 dB, it has 6 dB snr.
- More or less, there is some complication in the assumptions one makes on
- quantizer occupancy and peak level vs. rms level.
-
- >Therefore, it is correct to say that DAT has, simultaneously, X dB
- >dynamic range and Y dB snr, whenever X + Y = 96.
- No. Dynamic range doesn't add to SNR. Part of the problem
- is that many people have used different (so different in some
- cases as to be unrecognizable) meanings for dynamic range, and
- you are probably trying to sort the world out. Sorry, but
- "dynamic range" isn't particularly well defined, even where
- it IS well defined, so to speak. The Peak/Noise Floor dynamic
- range (that's what I call dynamic range) of both DAT and CD is
- 96 dB, considering the adjustments above.
-
- The SNR is indeed level sensitive, just like you'd expect with
- a medium with ONLY a noise-floor, and no signal-correlated distortions.
- This confuses some people who think that means that vinyl is better,
- but some people desparately want to be confused, too.
-
- >Now consider a floating-point system with 16 bits, 4 bits of exponent
- >and 13 bits (using the standard IEEE cheat) of mantissa. That system
- >has 90 dB of dynamic range (using the exponent only), and 78 dB snr.
- >Similarly to 16-bit linear, we can increase the dynamic range at the
- >expense of snr (but not vice versa).
- Well, the numbers are approximately ok.
-
- >Is that analysis correct?
- Except that the 78 dB SNR across the band will not always be enough to
- prevent unmasked "pumping" of the quantizing noise.
-
- You need more mantissa bits, at least 14, arguably 16. 16 also
- covers the necessary reproduction dynamic range (although not the
- recording dynamic range).
- --
- Bush *Copyright alice!jj 1992, all rights reserved, except transmission
- vs *by USENET and like free facilities granted. Said permission is
- Clinton -- *granted only for complete copies that include this notice.
- Just Say !NO! *Use on pay-for-read services specifically disallowed.
-