Steely Dan CDs Different Masterings: The Summary Thread

Discussion in 'Music Corner' started by bmoregnr, Jul 2, 2016.

  1. fast'n'bulbous

    fast'n'bulbous tight also

    Location:
    New York, NY
    Does this mean that 16 bit samples only have 90 dB of dynamic range, not the 96 (16 x 6) that's usually cited? I'd never thought about it this way, but since positive and negative peaks are in general symmetric it makes sense.
     
    Musicisthebest likes this.
  2. StingRay5

    StingRay5 Important Impresario

    Location:
    California
    No, it's still 96 dB. The only difference is whether the range is from 0 to 65535 (unsigned) or -32768 to +32767 (signed). Either way, the range includes 65536 possible values, so the resolution is the same.
     
    Musicisthebest likes this.
  3. fast'n'bulbous

    fast'n'bulbous tight also

    Location:
    New York, NY
    But it seems to me that 0 and 65535 in unsigned, or -32768 and +32767 in signed, don't describe two different levels of loudness, they describe the lowest and highest excursions of a generally periodic signal. I'd think there are only 32768 distinct values (think 'absolute value' if that helps) and thus 15 bits.
     
  4. StingRay5

    StingRay5 Important Impresario

    Location:
    California
    But the difference between signed and unsigned just changes where the center (silence) is. With signed values, a series of zeroes is dead silence with no DC offset. With unsigned values, it's a series of 32768's. The designers of the Redbook audio format could have done it either way and it would have made no difference at all to sound quality. Either way, you have a range of 65536 discrete values in which to describe a waveform, and the center of the range is where silence is.

    I don't think absolute values are helpful here. There is a difference between +x and -x. They have equal magnitude but they aren't equal. Loudness is not a function of magnitude alone; it has to do with the magnitude of the difference between the peaks and troughs of the waveform.
     
  5. CPRAudio

    CPRAudio Well-Known Member

    Location:
    MA, USA
    Lurked here for years and decided to post. I have been looking for a Hoffman Katy Lied CD and had about given up. Went into a small town library yesterday while on vacation and for $1.00 scored a #3!
     
    Keith V, Just_a_Dude, defmoot and 7 others like this.
  6. bmoregnr

    bmoregnr Forum Rezident Thread Starter

    Location:
    1060 W. Addison
    Hey, welcome aboard. I look forward to hearing more from you.
     
  7. Keith V

    Keith V Forum Resident

    Location:
    Secaucus, NJ
    Which version of Gaucho plays slower?
     
  8. bmoregnr

    bmoregnr Forum Rezident Thread Starter

    Location:
    1060 W. Addison
    I think it's the MFSL (and it seems some late 80's-early 90's remasters that got out before Nichols noticed?) that is/are sharper. Go to the Snap, Crackle and Pop Music post by Roger Nichols here: Metal Leg 18 – October 1991
     
    Last edited: Mar 4, 2022
    Keith V likes this.
  9. JoshM

    JoshM Forum Resident

    At some point, I plan to do a TBVO on Gaucho, but I’ve done some preliminary research, and the MFSL is the same speed as the original vinyl. So either the original vinyl was wrong and not caught for decades, or the MFSL is correct and the faster/sharper CDs are incorrect.
     
    fatwad666, bmoregnr and Keith V like this.
  10. Plan9

    Plan9 Mastering Engineer

    Location:
    Toulouse, France
    Well Nichols heard the MFSL and immediatly realized it was at the wrong speed.
     
    StingRay5, Musicisthebest and Keith V like this.
  11. JoshM

    JoshM Forum Resident

    There are few people in music and audio I admire more than Nichols. But we're all human. (In the Metal Leg, he calls the MFSL sharper/faster, but it's actually slower.) I'm just reporting what I've found.

    Lining up the tracks in Audacity, here are the approximate times for the same drum hit near the end of "Babylon Sisters":

    5:31.852 '80 US vinyl rip #1
    5:32.247 '80 US vinyl rip #2
    5:30.194 '90 Japan vinyl rip
    5:32.192 37220
    5:32.171 MFSL
    5:32.220 1991 Japan CD
    5:32.122 DVD
    5:29.781 2000 Remaster
    5:32.104 2003 SACD
    5:29.653 2010 SACD

    I expect some minor discrepancies from vinyl due to slight table speed variations. But, given that, the overall pattern here is that the original Japanese vinyl, the 2000 remaster, and the 2010 SACD are the faster ones. Given that the 2010 SACD was billed as a new transfer from the Japanese master tape, maybe it's not a surprise that it's similar to the Japan vinyl, assuming the Japanese vinyl was sourced from the same tape. The 2000 remaster, as far as I know, is the first one in this list sourced from Nichols's early-'80s digital tape transfer.

    I honestly don't know what to make of the above. It's possible either that, up until 2000, the only release to get Gaucho's speed right(ish) was the original Japanese vinyl release. However, it would seem surprising that no one -- including Nichols, Scheiner, Becker, and Fagen -- noticed that all of the U.S. releases up until 2000 were incorrect. On the other hand, maybe the digital tape made by Nichols in the early-'80s was recorded with the analog master running a little too fast. Perhaps that became his personal reference for the album. So when the MFSL was released, he believed the MFSL was too slow, even though it's actually the same speed as all of the previous releases besides the Japanese vinyl.

    Which is the correct speed? I don't know. All I can say is that the MFSL speed is close to the original vinyl and was the common speed for all digital versions up until 2000, at which point the faster 2000 remaster was released.
     
  12. Keith V

    Keith V Forum Resident

    Location:
    Secaucus, NJ
    This is very good. Thank you.
    Would the Citizen SD version be the same speed as the 2000?
     
    JoshM likes this.
  13. rjp

    rjp Senior Member

    Location:
    Ohio
    the MFSL 'aja' and 'gaucho' sound so nice on my system though.........i can't imagine not having them. and i have compared them to the box.
     
  14. JoshM

    JoshM Forum Resident

    Good call. That's one I forgot to include in my analysis. It's based on the early-'80s digital tape, so I'm guessing so. But I want to confirm it.

    If the Citizen version is the same speed as the 2000, then it's the first digital one at that speed. So ignore that part of my post.
     
    Keith V likes this.
  15. princesskiki

    princesskiki Kiki's Mom

    You are claiming that you or Nichols could hear a difference in pitch/speed between a track that runs at 5:32 and the same track that runs at 5:30 on the radio or car stereo (whatever that article said)? Hmmm.
     
  16. JoshM

    JoshM Forum Resident

    Both Stevie Wonder and Nichols supposedly heard the minute difference between the first Nightfly CD that used an analog tape and the corrected one that used the digital master. I'm not doubting Nichols's abilities. Could I hear the difference were I not lining them up and A/B-ing them in Audacity? I'm make no claims about that!
     
    princesskiki and Keith V like this.
  17. Keith V

    Keith V Forum Resident

    Location:
    Secaucus, NJ
    It might not be the first. They supposedly were using the proper digital transfer after the 1992 “scandal” with “faulty” analog masters.
     
    Plan9 likes this.
  18. princesskiki

    princesskiki Kiki's Mom

    Okay, hopefully, I will shut up about this subject after this post:

    As far as I am concerned, Robert Ludwig was not only one of the best LP mastering engineers from the late 60's to and throughout the 80's but his work also evidences that he was very meticulous about his work. Also, I read somewhere that Nichols insisted that he gives his final okay before anything got released (at least initially released). To me, that means that the ORIGINAL promo US LP of Gaucho with Robert Ludwig's initials and the earliest matrices on the dead wax would have been approved by Nichols and would be at the "correct speed."

    The JVC Japan for US CD pressings are pretty much a "flat transfer" of the original US promo LP of Gaucho and I can't tell any speed or pitch differences between the two.

    While it is true that many engineers in the music business in the first half of the 80's preferred the 3M technology over the 1600/1610 PCM technology in terms of sound, PCM technology won out (probably because Sony, Philips and Denon all championed it for their own self interest). Nichols may have used the better 3M technology when he did the original digital transfers of Steely Dan albums, but by the time "his" Aja and Gaucho came out on CD, there was compression and possibly noise reduction/shaping added and they just do not sound as good to me as Steve Hoffman's Aja or CSR or JVC Japan Gaucho.

    But then again, many people here seem to like compression and noise reduction/shaping.
     
  19. edenofflowers

    edenofflowers A New Stereophonic Sound Spectacular!

    Location:
    UK
    Is there a reason that the songs from Gaucho can't be checked against standard tuning to see if they run sharp or flat?
     
  20. shakesomeaction

    shakesomeaction ‘s what i need

    Location:
    couch
    Just wanna jump in here to defend the 2014 SHM-SACD of Gaucho. I’ve heard other versions and this one does it for me. The detail and sharpness lend itself quite well to the material, provided you have a decent SACD player.
     
    Plan9 and Keith V like this.
  21. fast'n'bulbous

    fast'n'bulbous tight also

    Location:
    New York, NY
    2.5 seconds out of 5 min 30 sec is 0.7%. That's very audible.
     
    Keith V likes this.
  22. Keith V

    Keith V Forum Resident

    Location:
    Secaucus, NJ
    My Gaucho is 38:09.
    MCAD-37220 DIDX 56
     
  23. StingRay5

    StingRay5 Important Impresario

    Location:
    California
    It's noticeable if you're comparing discs side by side, but it's small enough that not many people would notice it otherwise. In terms of tempo, it's less than a beat per minute's difference, and in terms of pitch, it's much less than the quarter tone that Nichols claimed.
     
    princesskiki likes this.
  24. Plan9

    Plan9 Mastering Engineer

    Location:
    Toulouse, France
    Frankly this mastering is my favourite Gaucho. I have heard all other digital versions and a couple of analogue versions too.
     
    shakesomeaction and albertop like this.
  25. JoshM

    JoshM Forum Resident

    Both the Citizen box and the VDP-26 '84 Japanese CD are the faster version, clocking in around 5:29.795 for that drum hit.
     
    Keith V likes this.

Share This Page

molar-endocrine