Diana Ross - Upside Down

The 45 length is 3:38, the LP length is 4:04.

I noticed that both the 45 and LP versions have a glitch at 0:29, right after the syllable "up". It sounds a little like a record scratch, even though that's not really what it is. I confirmed from a needledrop that it's actually on the 45, and would assume from its presence on some LP-length CD versions that it's also on the vinyl LP.

45 length

The oldest CDs I have of the 45 length are three Rhino discs that all came out in 1992, all of which are digitally identical clones of one another:

• Rhino's Disco Years Vol. 4 (1992)
• Rhino's Billboard Top Hits 1980 (1992)
• Rhino's Billboard Top Dance Hits 1980 (1992)

All three of the above have an unusually high treble boost, which gives a monstrous presence to the snare drum and pushes the hi-hat way out in front. I noticed a lot of brightness on the Disco Years CDs, so it's possible that it was originally mastered for Disco Years, and digitally cloned for the Billboard discs.

There are a few CDs that use the same analog transfer as the above three:

• Razor & Tie's Back To The '80s (1994; mastered too loud and clips a lot)
• Time-Life's Sounds Of The Seventies Vol. 38 Celebration (1994; sounds extremely close to Back To The '80s and also clips a lot)
• Time-Life's Sounds Of The Eighties Vol. 14 1980-1982 (1995; differently EQ'd digital clone)
• Time-Life's Sounds Of The Seventies Vol. 47 '70s Dance Party 1979-1981 (1997; differently EQ'd digital clone)

The version on Motown's Hitville USA Vol. 2 (1993) seems to be a different analog transfer that tames the high end. It's most noticeable on the fade, where there's practically no hiss here, but plenty on all of the above discs.

The version on Motown's Year By Year 1980 tames the high end even more, and sounds like it's from a higher generation source tape than the above discs.

I think I prefer the Billboard/Disco Years mastering over the others. The Time-Life Dance Party disc is mastered by Dennis Drake, and he retains a lot of the sizzle in the EQ, so that Dance Party has more high end than Hitsville.

All of the above are the 45 length, and all have the glitch at 0:29.

LP length

For the LP length, it's a weird bunch of discs, none of which would rate as excellent for sound quality.

Motown's All The Great Hits sounds like it uses a high-generation source tape, as does Warner Special Products' 2-CD Good Times (1991).

EMI UK's 2-CD Now The Millennium Series 1980 (1999) has decent dynamic range, but shortens its fade by a few seconds compared to the others.

I actually prefer slightly compressed version on Priority's Mega-Hits Dance Classics Vol. 12 (1993).

The 2-CD Deluxe Edition of Diana is the only version that somehow removes the glitch at 0:29, but it's compressed and has a very aggressive mastering. It also includes the unreleased original Chic mix, which is superb and a great novelty to hear after all these years.

neither

One disc to avoid at all costs is Motown's 3-CD Motown Anniversary, which probably came out in 1983. The version here runs too slow, runs way short at 3:26, and sounds like it's from a very high-generation source tape. Awful.

Edited by crapfromthepast

My white label Motown promo 45 M1494F (65967) has a listed and actual time of 3:37. It has the guitar ending. Deadwax is M1494F-B-6597-08B M-1494-A It appears that some minor effort to obscure the "B" was made. Side 1 is both sides.

crapfromthepast wrote: The 45 length is 3:38, the LP length is 4:04. I noticed that both the 45 and LP versions have a glitch at 0:29, right after the syllable "up". It sounds a little like a record scratch, even though that's not really what it is. I confirmed from a needledrop that it's actually on the 45, and would assume from its presence on some LP-length CD versions that it's also on the vinyl LP.

OK, I must be nuts. Listening to my version on iTunes from the Billboard 1980 disc, I don't hear a glitch. (And my copy on iTunes is an M4A file, so I pulled it
in directly form the disc.)

Gene - I'm 99% sure that it's a homemade edit. It can be reconstructed using pieces from the LP version. If there was a real 12" version, I would have run across it by now (I've been a huge fan of this track since 1980).

Those "Dance Classics" CDs are great for what they are, but they're not striving for authenticity. I get the sense that they just wanted to put out cool versions of the songs that they DJed with, and the labeling (and possibly the licensing) were afterthoughts.

UPDATE: Well, while I haven't been able to get a 1987 / 1990 pressing of the "Diana" CD, I've gotten several others, and the results are interesting.

The 1999 "remastering" didn't exactly brickwall it, but it's pretty dense. Frequency response now goes all the way up to 22.05 kHz, but it has an average RMS power of -12 to -13 db. File times out to 4:04.578.

The 48 kHz / 24 bit (MQA) and 192 kHz . 24 bit remasters sound similar to each other, but amazingly less compressed than the 1999 CD. Stereo imagery, to my ears, seems more "stable" or "solid" than other versions. I don't know if this the greater bit depth / frequency response, or just being less compressed. These run about -16.5 Average RMS Power, and while they appear to have some peak control applied, I really think they sound the best of the ones I've heard so far. They run 4:05.562

I got one more copy of "Upside Down" from "Diana Ross - All The Best." It's a 96 kHz / 24 bit remaster... except the frequency response cuts off sharply at 21.78 kHz! I'd feel robbed if I paid extra for this. (Whether or not you could hear it is a discussion for another time.)

It runs an Average RMS Power of about -15, and looks dynamically controlled but not limited or clipped. Sound is a little flat.

Some may prefer this mastering but I think it sounds like it could have had noise reduction applied although the frequency display doesn't show it. I'm not sure if this was an early digital master or exactly what's going on. Seems a little gritty. Just not to my taste. It runs 4:05.421.

I';m still looking for a copy of the 1987 / 1990 original release. If I find it, I'll update this post.

--

Ron, you are SO right about the LP version. If it sounds decent, the wavefile shows it's been manipulated; if it's a good dynamic range, it sounds like either noise reduction has been applied (or maybe Dolby is switched on when it shouldn't be).

A little research on Discogs suggests it was first released on CD in 1987.

It was then re-released in 1999 with the following notes: "Digitally remastered using 24-bit technology."

The 2003 two-disc version is the one that I understand is limited to death, stomping out all of the dynamics.

Edited by NightAire