kavaler@zion.berkeley.edu.UUCP (06/30/87)
It will be interesting to see how the record companies are going to add this new CD feature. If they tell people ahead of time, the price of non-notched CDs will rise dramatically (can you imagine the resale value of a non-notched CD when the notched CD has an audible and possibly bothersome notch). Maybe this way the record companies can produce 2 versions of each CD, one with notch (costing $14.00 each), and one without notch (>> $14.00 each). I also wonder if there will be any marks on CDs that are notched. How does one know? Are they already notched?? Anyone have a spectrum analyzer handy? Robert Kavaler
taras@utgpu.UUCP (07/17/87)
In article <19530@ucbvax.BERKELEY.EDU> kavaler@berkeley.edu (Robert Kavaler) writes:
# It will be interesting to see how the record companies are going to
# add this new CD feature. If they tell people ahead of time, the price
# of non-notched CDs will rise dramatically (can you imagine the resale
# value of a non-notched CD when the notched CD has an audible and possibly
# bothersome notch). Maybe this way the record companies can produce
# 2 versions of each CD, one with notch (costing $14.00 each), and one without
# notch (>> $14.00 each). I also wonder if there will be any marks on
# CDs that are notched. How does one know? Are they already notched??
# Anyone have a spectrum analyzer handy?
#
# Robert Kavaler
Since the human hear does have a notch in it's hearing range of about 3khz,
I wonder just how noticeable this will be. I would rather doubt that the record
companies, atleast some of them, would want to sell a non notched CD at any
price considering how much they are fighting for the legislation that makes
use of it. I also wonder how much of a premium will be charged for DAT units
that do have the special anti pirate circuits defeated? They will be available,
and they should not be all that hard to build.
--
Taras Pryjma
uucp: taras@gpu.utcs
bitnet: tpryjma@utoronto
Bell: +1 (416) 536-2821
Ok then, It's settled. I will have my computer call your computer and work
it all out. ....... But wait ...... it won't do any good, never mind.jj@alice.UUCP (07/22/87)
In article <1987Jul16.222805.5650@gpu.utcs.toronto.edu>, taras@utgpu.UUCP writes: > Since the human hear does have a notch in it's hearing range of about 3khz, > I wonder just how noticeable this will be. I would rather doubt that the record > companies, atleast some of them, would want to sell a non notched CD at any > price considering how much they are fighting for the legislation that makes > use of it. I also wonder how much of a premium will be charged for DAT units > that do have the special anti pirate circuits defeated? They will be available, > and they should not be all that hard to build. I've got a few problems here. First, there is no "notch" at 3kHz in a normal individual's hearing. In fact, 3kHz is very close to the lowest (i.e. most sensitive) absolute threshold, which usually comes around 4kHz. [Fletcher, Harvey, "Auditory Patterns", Reviews of Modern Physics, 1940, 12, 47-65]. <many other, newer works are published, but the old standard is STILL damn good> For any given individual, an ear canal resonance can cause something that looks like a notch, however the ear canal resonance of individuals varies enormously, with effects measurable between 1kHz and 6kHz, depending on individual. Saying that this causes a "notch at 3kHz" cannot be supported in the general case, and can only be supported in very specific (and unusual) individual cases. Second, the 100Hz wide notch at 2828 Hz about 1/4 of a critical band wide, and does NOT come between "notes on the piano" or "harmonics" as is some currently claim. What this means is that it is NOT likely to be inaudible under all 'normal' conditions, and it will likely be audible to most people given a signal with white noise filtered to 1 critical bandwidth around the notch frequency. (It represents a change of ~1.25dB in the noise energy centered around that critical band, which is audible under controlled conditions. This ignores the frequency shaping due to the notch filter, which will enhance the detectibility.) [Schroeder, Atal, and Hall, "Optimizing Digital Speech Coders by Exploiting Masking Properties of the Human Ear", JASA 66(6), Dec 1979, pp 1647-1651] has a good discussion that suggests how the frequency shaping will work to enhance detectibility. Many other references follow from that paper. Third, the above statements are made assuming that the filtering method that removes the 100Hz wide notch are linear phase. If there is a significant phase shift in the filter, that phase shift will be within a critical band, and more modern results <that I don't have at hand> show that significant phase shift, within a critical band, will cause noticable changes in timbre, even assuming no amplitude distortion. There is current debate on what "significant phase shift" amounts to, but the amounts are smaller than can be generated by simple analog filters. This behavior can be confirmed by any text on Fourier analysis and/or signal processing. Fourth, the issue of impulse response (related to 3) has not been discussed. The 100Hz notch MUST have an impulse response longer than a few mS at the very least. Depending on the phase characteristics, this impulse response could easily be much longer (in the 100mS range). Fifth, I can easily imaging electronic music that has fundamental signals (let's say sine waves for the time being) within the notch area. Clearly, if there is any significant energy in the notch area its removal will not be masked by the auditory process. Witness Dire Straights "Love Over Gold" for examples that might be interfered with. <I haven't measured it objectively, so I'm not going to stick my neck out, but it sounds likely.> This condition can also be met by a piano with "stretched" tuning, many string instruments, some wind instruments, and so on. If various Baroque or Rennisance (sp, I know) tunings are used, (say A=435, A=415, A=460, for example) the idea that "it's between two notes on the scale" is invalidated. So. I'm not comfortable with the idea of the "A flat remover" as currently described in the literature I've been able to find. I don't appreciate the idea of legislated "low-fi", and I am very annoyed by the hysteria that is permitting the mandating of (probably) permanant inferiority for all recordings using the "copycode" idea. -- THERE'S BEAR-LY ENOUGH FOR ALL. "Juggle or die!" (ihnp4;allegra;research)!alice!jj Copyright JJ 1987. All rights to mail reserved, USENET redistribution otherwise granted to those who allow free redistritution.