Steely Dan-"Peg"

My commercial 45 has a listed time of (3:58), but an actual time of (3:52).

The short answer: I don't think there are any bad masterings of "Peg". (I haven't heard the Aja remaster that came out around 1999, though.)

There are endless threads on the Steve Hoffman board about the many different masterings of Aja. It's pretty brutal sorting them out, and frankly not worth your effort.

My copy of Aja has matrix number MCAD-37214-A8E, and I believe that it's a made in Japan-for-US disc released in 1984. (The peak level for "Peg", track 4 on the disc, is 88.7% - that helps tell the masterings apart) Almost all of the subsequent masterings I found use the same analog transfer as this mastering of Aja:

• MCA's (Canadian?) Greatest Hits (copyright 1978, unsure when or where this was actually released) is digitally exactly 0.2 dB quieter
• Time-Life's Sounds Of The Seventies Vol. 10 1978 (1990) is a digital clone
• MCA's original pressings of A Decade Of (mastered by Bob Ludwig and Roger Nichols)

All of the above sound basically the same - same great dynamic range, same EQ, no evidence of noise reduction, same reassuring hiss on the fade. On Aja, the fade is technically truncated, but it's insignificant; the level drops abruptly from -70 dB to zero. The other discs make the transition to zero a little more gradual.

The version on the 4-CD Citizen box set (1993) uses the same analog transfer as my copy of Aja (not a digital clone), but with a significantly brighter EQ and evidence of a tiny, tiny amount of noise reduction (I hear a small whooshy effect around 3:51 on the fade.) "Peg" sounds pretty shiny and wonderful here - the brighter EQ really brings out the shaky percussion during the chorus. There's a digitally identical clone of Citizen on later pressings of A Decade Of (mastered by Glenn Meadows).

So to sum up, all of the above use the same analog transfer from the same two-track mixdown tape. Everything pre-1993 has no noise reduction. Everything 1993 and after has a brighter EQ that brings out the high end, but has a tiny amount of noise reduction. None sound bad. (All run at 117.0 BPM throughout with no drift - it's a live drummer playing to a click track.)

The only mastering of "Peg" that I have that uses a different analog transfer from the above is MFSL's gold disc of Aja (1988; UDCD 515). It sounds even brighter than the 1993 discs, and shows no evidence of noise reduction. I thoroughly approve. (Here, it runs at 117.3 BPM throughout.)

If you can't afford the MFSL gold disc, just know that the others also sound great and you won't be disappointed by any of them. (Again, I haven't heard the 1999 remaster of Aja.)

Edited by crapfromthepast

As someone who has amassed a large chunk of Steely Dan
CDs, I have to say that my favorite sounding disc for
most of their hits is 2006's Steely Dan: The Definitive
Collection.

This set was remastered by Sony's Joseph M. Palmaccio and
it's a night and day difference between all other CD
appearances of their best known songs. Best of all, no
noise reduction at all.

Andy - Happy to explain.

When I talk about a particular analog transfer, I mean a specific playing of a two-track mixdown tape, and from that specific playing, capturing the analog audio in the digital domain with an A-to-D converter.

We can tell that the same analog transfer is used on various masterings by comparing the waveforms.

If we can get them to line up and remain in sync throughout, and we can get partial or total cancellation by inverting one of them, then we can say that one is a digital clone of the other (maybe with a level change, an EQ adjustment, and/or added compression/limiting).

If we can get them to line up but they drift slightly over the course of the song (and slightly means just a few samples), then we can say that both use the same analog transfer. This case occurs when the track is played from a digital source in the analog domain, maybe to adjust EQ etc., and captured back into the digital domain with another A-to-D converter. The clocks of the playback device and the A-to-D converter aren't perfectly synchronized, so there's always a small drift between them, hence the misalignment of a few samples.

Hope that helps.

Why? I have two ideas, neither of which is based on any facts.

First, I think the extra-analog-step cases may have predated digital workstations. It could be that a particular engineer wanted to mix on a particular console, and the only way to do that was to send the source material though it in the analog domain.

Second, I think some of the extra-analog-transfer cases may have been done inside the record company of the source material. For example, a tape librarian at the source record company might have done an analog dub of material to be sent to the compilation record company.

Just guesses. I've wondered about that quite a bit myself. I noticed that it also depends on who is mastering the compilation. Some mastering engineers work in the digital domain, so I see a lot of digital clones on their compilations. Others (likely) work in analog, so I see the same-analog-transfer tracks on theirs.

I guess I can defer to Mark, or any of the other engineers here who have worked on compilations. Any thoughts?