ark@rabbit.UUCP (Andrew Koenig) (02/09/84)
I just read a long article by Tom Buckley. Its main points can be summed up roughly as follows: "I know how specific instruments and performers should sound. None of the digital recordings of these performers and instruments has sounded as much like this ideal as the best analog recordings I have heard." Jim Johnston has posted several articles recently that I can summarize roughly as follows: "The digital recording system used in CDs produces an output that is close enough to the input that it is almost impossible for anyone to tell the difference." I want to make the point that these two statements can both be true at once! Consider the typical recording process, analog or digital. Sound is picked up by microphones, amplified, mixed (if there are more than two microphones), and usually processed in other ways to make a master tape. This other processing can include noise reduction, equalization, editing, and so on. The result of all this is a master that presumably sounds "right" to the producer, or whoever else might have a say in this part of the process. Now the contents of the master must be transferred to the final medium, be it records, tapes, CDs, whatever. In the case of records, the signal goes through a good deal of additional processing: RIAA equalization, more equalization to compensate for the mastering engineer's guess as to how much of the high end will disappear in the vinyl, dynamic range compression and bass rolloff to keep the stylus from flying out of the groove, and so on. The object of this game is to produce a master disc that sounds as much like the master tape as possible, though there will inevitably be differences. There will be further differences when the records are pressed. Let us assume, for the moment, that our producer has had much experience in such things. It seems sensible that the producer will then try for a master tape that he thinks will give the best possible records, not for one that sounds best by itself. Enter the digital process. If Jim Johnston is correct (I am inclined to think he is, but I do not have enough first-hand evidence to convince me completely), the result is a CD that is audibly identical to the master tape. But that master tape was produced with the assumption that it would manipulated somewhat further. Might this not account for some apparent problems? "I can also simulate the analog signal's apparent extra sweetness by tilting the digital frequency response down about 1 dB through the midrange (It's surprising how small a change in midrange frequency response can noticeably affect harshnewss and stereo imaging!)" -- Richard Burwen Tom and Jim: if I have done injustice to your positions, I apologize.