tristram%conan.asd.sgi.com@SGI.COM (03/06/91)
See Mix, April 1991... A maniac at Mix magazine took selections from the Prosonus reference disk and other sources, and used a pair of common pro DAT recorders to make *NON-DIGITAL* copies back and forth on virgin tape for *TWENTY* generations. A panel of three "golden-eared" producers were unable to reliably distinguish approximately one third of the material from an original (one-generation) source DAT. The author indicated that this result renders some of the current hoopla regarding SCMS (and digital copies) moot. Interesting? -- David Tristram 1661, tristram@sgi.com True words seem contradictory.
konar@lennon.SRC.Honeywell.COM (Mithat F Konar) (03/07/91)
In article <9981@uwm.edu> tristram%conan.asd.sgi.com@SGI.COM writes: >See Mix, April 1991... > >A maniac at Mix magazine took selections from the Prosonus reference >disk and other sources, and used a pair of common pro DAT recorders to >make *NON-DIGITAL* copies back and forth on virgin tape for *TWENTY* >generations. A panel of three "golden-eared" producers were unable to >reliably distinguish approximately one third of the material from an >original (one-generation) source DAT. The author indicated that this >result renders some of the current hoopla regarding SCMS (and digital >copies) moot. Interesting? If you re-read the conclusion, you'll discover that the panel was able to reliably distinguish a difference between the copy and the original on most material. What they didn't do reliably was indicate correctly which was the original and which was the copy. A subtle but important distinction. Mithat Konar
chrisc%gold.gvg.tek.com@RELAY.CS.NET (Chris Christensen) (03/07/91)
In article <9981@uwm.edu> tristram%conan.asd.sgi.com@SGI.COM writes: >See Mix, April 1991... > >A maniac at Mix magazine took selections............... I believe that Jim Paul's article in MIX was very interesting. It is always humorous to note reactions to professional audio persons subjective evaluations of sound. Your comment in "quotes" seemed to indicate animosity towards this type of person. I consider myself a professional audio person and fully admit that I don't understand all I know about audio, but I am successful in my endeavors. I won't make any assumptions as to your experience in audio but I will say this and DON'T QUOTE ME please. Most professional audio people don't consider themselves audiophiles. I believe that Jim Paul and the MIX editorial staff had honorable intentions in presenting this article. Please feel free to flame, but understand that my intentions are in defense of my industry. Chris Christensen
jfw@neuro.duke.edu (John F. Whitehead) (03/08/91)
In article <9981@uwm.edu> tristram%conan.asd.sgi.com@SGI.COM writes: >>See Mix, April 1991... >> >>A panel of three "golden-eared" producers were unable to >>reliably distinguish approximately one third of the material from an >>original (one-generation) source DAT. In article <10017@uwm.edu> Mithat F Konar writes: >If you re-read the conclusion, you'll discover that the panel was able >to reliably distinguish a difference between the copy and the original >on most material. What they didn't do reliably was indicate correctly which >was the original and which was the copy. However, on certain types of material they were indeed able to differentiate between the original and the copy. For instance, the original solo trumpet and the solo voice were generally picked correctly. The denser or more obscure types of sound, including music ensembles and sine wave sweeps, were the ones more often chosen incorrectly. The differences in sound were attributed (in their opinions) to phase problems from the repeated copying, which noticeably (although not significantly) degraded the sound of the solo instruments. The denser music masked any phase problems so they could not reliably pick the correct one. An interesting note is that none of the engineers seemed particularly fond of DATs -- i.e., they do not rely on them/use them extensively in their own applications. But one did say that he would now not have any problem with making a dub from the analog outs since the sound degradation is so insignificant. John Whitehead Internet: jfw@neuro.duke.edu Department of Neurobiology jfw@well.sf.ca.us Duke University Medical Center Bitnet: white002@dukemc Durham, North Carolina
david@agora.rain.com (David Robinson) (03/08/91)
In article <9981@uwm.edu> tristram%conan.asd.sgi.com@SGI.COM writes: >See Mix, April 1991... > [Delete a line or three...after 20 generations of DAT duping...] > A panel of three "golden-eared" producers were unable to >reliably distinguish approximately one third of the material from an >original (one-generation) source DAT. The author indicated that this >result renders some of the current hoopla regarding SCMS (and digital >copies) moot. Interesting? > No, merely predictable. %-) Actually, I find it *much* more interesting to note the two-thirds of the material that they apparently *could* tell the difference on. There's nothing like showcasing a logical flaw. Adam and Eve headed for the bushes... Nothing has been rendered "moot." As the Stereophile test CD (with its "original" and its "100 generations later" samples) shows, things can most definitely happen to digital sound in the retelling. A theoretical basis for what we are hearing has not been definitely established...yet...but I don't believe that it's arguable that *something* goes astray. This calls for more work, I'd say.... david@agora.rain.com > > > > >-- >David Tristram >1661, tristram@sgi.com True words seem contradictory. -- My opinions belong to me...and vice versa. They're not copyrighted; third party thinkers should feel free to clone them at will. david@agora.rain.com davidr@glacier.UUCP
chowkwan@priam.usc.edu (Raymond Chowkwanyun) (03/09/91)
It's not quite the same but the Chesky test CD has a track which is a 100th generation copy of a digital tape. You can compare the copy and the original. To my ears the copy sounded muted and less detailed. -- ray
ed@mtxinu.COM (Ed Gould) (03/11/91)
> ... the Chesky test CD has a track which is a 100th generation copy > of a digital tape. You can compare the copy and the original. To > my ears the copy sounded muted and less detailed. This seems very odd to me. Has anyone tried to do a *digital* comparison of the multi-generation digital copies (which ought to be identical, no?) to see if there really are any differences? If there are no digital differences, then I would have to conclude that any sonic difference is psychological. Could there be any other explanation? If there are differences between the digital data, is this due to lack of error correction or something truely insidious? -- Ed Gould mt Xinu, 2560 Ninth St., Berkeley, CA 94710 USA ed@mtxinu.COM +1 415 644 0146 "I'll fight them as a woman, not a lady. I'll fight them as an engineer."
david@agora.rain.com (David Robinson) (03/12/91)
In article <10123@uwm.edu> ed@mtxinu.COM (Ed Gould) writes: >> ... the Chesky test CD has a track which is a 100th generation copy >> of a digital tape. You can compare the copy and the original. To >> my ears the copy sounded muted and less detailed. > >This seems very odd to me. Has anyone tried to do a *digital* >comparison of the multi-generation digital copies (which ought to >be identical, no?) to see if there really are any differences? If >there are no digital differences, then I would have to conclude >that any sonic difference is psychological. Could there be any >other explanation? If there are differences between the digital >data, is this due to lack of error correction or something truely >insidious? I'm not sure that the problem is in the digital domain (01001001 and all that), since the numerics check out. My personal suspicion is that there may be some "drift" in the time domain...along with an exacerbation of "digititis" due to a "squaring of the squared sonic wave"^100. Too much stair-stepping, and not enough curve, repeated over and over...the numbers can still "match" (*love* that rounding!) but be imprecise sonically. This would be akin to multiple generations in the analog domain building up hiss and 60 Hz layers; the music is still "there," and is still "right," but is fighting sonic drift. Disclaimer: My thoughts here are more visceral than analytical. Comments? david@agora.rain.com
ed@mtxinu.COM (Ed Gould) (03/15/91)
>My personal suspicion is that there may be some "drift" in the time >domain...along with an exacerbation of "digititis" due to a "squaring of >the squared sonic wave"^100. Too much stair-stepping, and not enough >curve, repeated over and over...the numbers can still "match" (*love* >that rounding!) but be imprecise sonically. >This would be akin to multiple generations in the analog domain building >up hiss and 60 Hz layers; the music is still "there," and is still "right," >but is fighting sonic drift. If that (whatever "that" is, I'm not sure I follow your argument) is the case, then it should show up digitally, too. If the digital signals are no longer in the right time relationship with one another, that's a problem, and should be digitally detectable. I still claim that if the digital signals are really the same (including their time relationship), then any sonic difference must not be in the recording meduim. If they're not the same, then there's a bug in the copy process. -- Ed Gould mt Xinu, 2560 Ninth St., Berkeley, CA 94710 USA ed@mtxinu.COM +1 415 644 0146 "I'll fight them as a woman, not a lady. I'll fight them as an engineer."