rmd@hpcnoa.UUCP (01/31/84)
Here's some ammunition for CD foes: I bought a CD recently with an audible defect which cannot be removed by cleaning. The defect is clearly in the CD itself and not in the master because it does not always appear at the same place in the music. The defect is a band of very small bubble or dust particles in the CD near the outer portions of the music. The band is about 1/8th of an inch wide and is straight, intersecting the disc tangentially. It is 2 or 3 inches in length. The audible result sometimes sounds like a brief tape dropout and sometimes sounds like a low-level 'pop' on a LP. The CD player is a CDP 101, which of course has only mediocre error correction characteristics. A better CD player might very well be able to play this CD with no audible problems. My conclusion is that error correction ability is more important than I had previously thought. We can only expect the quality of CD manufacturing to go down as the volume increases. Even on this disc, the defect is not as bad as the pops and scratches on LPs, but it is enough to be irritating. Anyone have any constructive comments or similar experiences? Rick Dow inhp4!hpfcla!rmd
stenerso@orstcs.UUCP (02/09/84)
The CD debate is now getting petty. Getting a few bad CD's is no reason to damn the whole CD process. Just as with analog discs (LP's) you tend to get a few rejects. When this happens you simply return them to where you purchased them for an exchange. I would expect quality to go up as quantity increases since any bugs should be eliminated or reduced as the process of producing discs is 'perfected.' The error correction used is a Cross Interleave Reed Solomon (CIRC) Code. The maximum correctable burst length is 3500 bits (~2.4mm) and the maximum acceptable burst (error correction + interpolation) is 12000 bits (~8.5mm). However, when Sony developed the CIRC code they made it possible to use all of its power or part of it. The idea being that more expensive players would use all of the code and less expensive players would only use part of it. At this point I do not know how much of the error correction capability is used in present players. I must admit that I do not own a CD player at present. They are still a bit spendy for a struggling EE student! But I have listened to one though a pair of Stax headhones and do plan on buying a player by the end of the year. As it see it there are not any major problems with the following: 1. sampling rate (band width) 2. quantization (16 bits linear/channel) (I do wonder why a nonlinear quantizer was not used though, ie. CCITT.) 3. error detection and correction 4. features (ie. direct access to any part of the disc) The major limitations are more concerned with the present implementation then with the disc format. 1. bad software (discs) 2. the analog and digital filters could be a problem if not designed properly. I will not give details on the above since they have already been beaten to death on the net! Flames to /dev/null! Roger Stenerson