crapfromthepast MusicFan
Joined: 14 September 2006 Location: United States
Online Status: Offline Posts: 2237
|
Posted: 26 June 2012 at 1:34pm | IP Logged
|
|
|
I'll start with my conclusion, then work backwards to explain:
I believe that the 1995 DCC CD version of Gerry Rafferty's City To City album is sourced from the '80s-era EMI UK CD pressing of the same title.
The DCC pressing is extremely sought-after by audiophiles and collectors, and routinely trades for more than $100 on the used market. The '80s-era UK pressing on EMI, in comparison, trades for about one-tenth that price.
In some circles, it's blasphemy to say that one sounds like the other. I will then further blaspheme in saying that the DCC is sourced from the '80s-era UK EMI disc.
To be clear, I'll show you the exact discs I'm comparing.
'80s-era UK EMI
Catalog number 7 46049 2, released 1986 or earlier.
As of March, 1986, this was the only CD version of City To City that was available. You can see it in this scanned page of the March, 1986 DiscList published by Infomedia out of Novi, Michigan. In case you have difficulty reading the scan, the relevant line is "RAFFERTY, GERRY - - - - - - - CITY TO CITY - - - - - - - - - - ENG [the "import" column"] - - EMI - - - CDP 746049". There was no US counterpart, as of March, 1986.
Here are scans of the front/back/disc of the UK EMI release:
If you bought the UK EMI disc, you got a great-sounding transfer of the UK analog two-track mixdown tapes. The EMI two-fer of City To City and Night Owl uses the same mastering as the older '80s-era UK EMI discs. I'm inclined to think that the two-fer of these two albums released on Collectibles also uses this mastering. As far as I know, the '80s-era mastering was the only mastering ever used on UK copies of the CD until EMI released a 2-CD version with a bonus disc in 2011.
Note that in the US, the vinyl releases on United Artists used sped-up versions of the songs on both the album and the 45s. I think that the US CD release also had speed errors, so I've avoided buying it.
DCC
Catalog number GZS 1075, released in 1995.
Scans of the outer cardboard sleeve, the back of the CD case and the disc:
If you bought the DCC release, you also got a great-sounding version of the album. It just happens to sound almost exactly like the earlier UK EMI CD version, which wasn't widely known at the time and still isn't readily acknowledged in some audiophile circles.
Comparisons
I compared all ten tracks in a variety of ways, and all do indeed sound similar. Aside from volume differences, I doubt that many of you would be able to tell the difference - I can't.
First comparison: Listen to them.
Both sound basically the same to my ears, including the amount of hiss on the fades, the points at which the fades cut to silence, the EQ on the fades, you name it. Qualities like "warmth", "soundstage", "definition", etc. - all the same on both. In an A/B test, I couldn't tell one from the other based only on my ears.
Second comparison: Look at the waveforms.
The EMI has its absolute polarity inverted in both channels, compared to the DCC. Meaning: the waveforms will appear upside-down if you look at them, but there's no audible change resulting from this. (Some may disagree, but this is not the thread to take issue with whether or not there's an "absolute" polarity.)
After you invert one of the versions, they track extremely closely with one another. They're not exact digital clones (i.e., they do fall slightly out of sync over time), but the speed difference is so small that I'd guess that the CD was played through an analog deck (for adjusting EQ), then re-encoded to digital.
To give you an idea of how small the speed difference is, I compared points near the beginning and end of "Baker Street". Over the course of the song (about 15,000,000 samples), the DCC and EMI UK discs fell out of sync by about 1500 samples. That's a speed accuracy to within 0.01%, held constant with no drifts and no wow/flutter.
If you've ever worked with analog tape machines, you'll know that it's generally impossible for two different tape machines to have their speeds match this closely. I know from my own experience, it's very difficult getting the same machine to have the same speed before and after lunch!
If DCC were to have done their own new analog transfer from the UK analog two-track mixdown tapes, at least nine years after EMI played them on a different machine, there is essentially no way for the speeds to match to within 0.01%.
In comparison, in my day-to-day experience comparing different masterings, I routinely see variations of about 0.5% to 1% for different playings of the same two-track mixdown tape, even for the same mastering engineer.
If two different CD copies of the same song differ in speed by 0.01%, that tells me that the analog two-track mixdown tape was played once to make the first CD, and that the seocnd CD uses the first CD as its source. In the terminology I usually use, I say that both CDs use the same analog transfer.
Third comparison: Do spectral analyses on both, subtract one from the other.
So to explain the graph below, I opened all 20 tracks in Audacity and ran a spectral analysis on each, or at least the first 237 seconds of each (that's all that Audacity can handle), using a Hanning window and a size of 1024 (I'll be the first to admit I don't really know what I'm doing when it comes to these settings). I exported all 20 spectra, imported everything into Excel, and made the plot that you see below.
The results, track-by-track, using the EMI UK as a source to reproduce the DCC:
1. "The Ark": raise levels by about 0.2 dB.
2. "Baker Street": raise levels by about 2 dB.
3. "Right Down The Line": boost 2 kHz by 0.5 dB and cut 8 kHz by 0.5 dB, then drop the level by 0.5 dB.
4. "City To City": boost 2 kHz by 0.75 dB and cut 10 kHz by 0.25 dB, then drop the level by 1 dB.
5. "Stealin' Time": raise levels by about 0.5 dB.
6. "Mattie's Rag": : raise levels by about 1.5 dB.
7. "Whatever's Written In Your Heart": raise levels by about 0.7 dB.
8. "Home And Dry": raise levels by about 0.6 dB.
9. "Island": raise levels by about 1.2 dB.
10. "Waiting For The Day": raise levels by about 1.1 dB.
This tells me that on eight of the tracks, there are level (volume) shifts from the EMI UK to the DCC, but no tinkering with the EQ. On "Right Down The Line" and "City To City", there are small EQ changes. That should be all you need to reverse-engineer the DCC from the EMI UK.
As an aside, the little wiggles in my plot are obviously from calculation errors, made from my naive assumptions in the settings for the spectrum analysis. No surprise there.
But an effect that I think may be real is the slow, downward drift of the curves from the left side to the right side of the plot. I think that may have arisen from playing back the original EMI UK disc on equipment that didn't quite have a truly flat spectral response. Don't know for sure.
One more little piece of evidence
For many of the DCC releases I've owned, I noticed a credit like the one on DCC's Steve Miller Band - Greatest Hits 1974-78, which reads "Re-mastered for this compact disc by STEVE HOFFMAN from the original two-track master mixes". The City To City just says, "Remastered for this compact disc by STEVE HOFFMAN, January 1995". No mention of the original two-track master tapes.
DCC never claimed to to do a fresh analog transfer of the two-track mixdown tapes for City To City, and I don't believe they did one. Instead, I think they used the EMI UK CD as the source, played it through an analog deck so they could do EQ adjustments, and re-encoded to digital.
Conclusion
The 1995 DCC disc sounds just like the 1986-or-earlier EMI UK disc. So much so, and in so many ways, that I believe that the DCC uses the same analog transfer as the older EMI UK disc.
I'm aware that the Steve Hoffman board would be a more appropriate place for discussion of this, but my posts on this topic get deleted immediately. You can draw your own conclusions.
|