A correspondent said something to me about MD and CD recorders that unneces- sarily (and inadvisably) resample all digital input, including signals that are already sampled at 44.1 kHz. I'd have to guess that resampling is like a round of DA-AD conversion and can introduce as much infidelity as an analog transfer (except that it wouldn't be subject to distortion or attenuation of an analog signal along its cable). To resample a signal already at the right sampling rate is not just a waste of processor cycles; it is extremely likely to damage the fidelity and cannot possibly improve it. The manuals for my MD equipment evade the issue, stating only that the built- in sampling rate converter allows recording of 32- and 48-kHz sources as well as 44.1-kHz sources. That is the case for all my Sony decks and the Aiwa AM-F70; the MZ-R3 has no SRC. I would assume that in monitor mode the Sony MD decks resample everything, but I can't see why they'd resample 44.1-kHz input that they're recording. Only the manual of my CD recorder (a Pioneer PDR-509) hints at bypassing the converter when the source is already sampled at 44.1 kHz, and even that is phrased in an ambiguous, contradictory way. The mono recording problem that I brought up last June or July didn't happen on the MZ-R3; I wonder if gratuitous resampling was the real culprit. Then again, Is there a way to tell if a signal is being unnecessarily resampled? I guess it's impossible, in making a mathematical comparison, to tell whether a dif- ference was caused by ATRAC encoding or by gratuitous resampling. For recording analog sources or (not that I have any) digital sources at 32 or 48 kHz, sampling or resampling is needed; for MDLP, it probably makes no noticeable difference. Are there any recorders with current-day ATRAC versions that definitely do not resample 44.1-kHz input, even if they don't support MDLP? I really would hate to use the R3 and downgrade to ATRAC 3.0 to avoid this problem, but it seems to outweigh the benefits of later ATRAC implementations. [On the other hand, the R3's better behavior with last year's problem could come from its recording only the left channel of digital input during mono mode rather than trying to average the two channels, and Jonathan Irwin's efforts to track down the explanation then might not have been in vain after all. Still, the R3 doesn't have a sampling rate converter and therefore can't possibly be guilty of gratuitous resampling.] In a less imperfect world, we could ask the manufacturers whether specific models resample 44.1-kHz input, and they'd give us truthful answers. ----------------------------------------------------------------- To stop getting this list send a message containing just the word "unsubscribe" to [EMAIL PROTECTED]
