| Author |
|
crapfromthepast
MusicFan
Joined: 14 September 2006
Location: United States
Online Status: Offline
Posts: 2196
|
| Posted: 28 April 2015 at 7:44pm | IP Logged
|
|
|
The oldest version that I have on CD is on Mystic Music/Cema's 2-CD Mellow Gold (1991). It has pretty severe noise reduction. Listen as the high end completely disappears on the fade. Avoid at all costs.
There's a better, but still muffled, version on Rhino's Have A Nice Day Vol. 25 (1996). There are differently EQ'd digital clones on:- Time-Life's 2-CD Body Talk Vol. 12 By Candlelight (1997)
- Eric's Hard To Find 45s On CD Vol. 8 (2002; has added noise reduction)
There's a far, far better version on JCI's 18 Disco Superhits (1996), which sounds like it's from better tapes than the Rhino, and has no artifacts due to noise reduction. The same analog transfer is used for:- Starland Music/Warner Special Products' 2-CD Reflections Of Love (1996; truncated fade)
- Madacy's Rock On 1978 (1996)
- Simitar's Number Ones Smooth Love (1998)
My recommendation:
Go with JCI's 18 Disco Superhits (1996), although Rock On 1978 will work just fine.
Edited by crapfromthepast on 29 April 2015 at 7:01am
__________________
There's a lot of crap on the radio, but there's only one Crap From The Past.
|
| Back to Top |
|
| |
The Hits Man
MusicFan
Joined: 04 February 2007
Location: United States
Online Status: Offline
Posts: 665
|
| Posted: 29 April 2015 at 12:44am | IP Logged
|
|
|
The song may have been brightened up in some CD mastering
because the 45 and album was somewhat dull too.
__________________
|
| Back to Top |
|
| |
EdisonLite
MusicFan
Joined: 18 October 2004
Location: United States
Online Status: Offline
Posts: 2236
|
| Posted: 29 April 2015 at 12:57pm | IP Logged
|
|
|
crapfromthepast wrote:
| There's a better, but still muffled, version on Rhino's Have A Nice Day Vol. 25 (1996). There are differently EQ'd digital clones on... |
|
|
I've been meaning to ask you for the longest time ... When a recording is different (e.g. EQ-wise), how do you determine if it comes from another existing version on another CD (and then Re-EQ'd or compressed and altered) vs. a brand new transfer from the original master tapes?
Even two versions from the same transfer could have two different blank-space times at the beginning of the song if one version shortens or lengthens the pre-intro space (so that can't always be a clue). And if someone takes an existing transfer and compresses it, the shape of the WAV will look quite different from the original WAV (meaning both could come from the same transfer yet look radically different.)
So I've always wondered how you can decipher between a "digital clone" that's been modified (EQ, compression, blank space at the intro, faded earlier than the other version) vs. a newly transferred version.
I find it fascinating because I'd never be able to determine what comes from the same source if the two files don't match up exactly.
Edited by EdisonLite on 29 April 2015 at 12:58pm
|
| Back to Top |
|
| |
crapfromthepast
MusicFan
Joined: 14 September 2006
Location: United States
Online Status: Offline
Posts: 2196
|
| Posted: 29 April 2015 at 8:47pm | IP Logged
|
|
|
Good question - I'll describe what I do for these analyses.
First, I load everything into a multi-track editor. I use an ancient version of Cool Edit Pro (v. 1.1), but Audacity should work fine as well. I adjust the positions of the files to line up some particular feature at the beginning of the song. I then look at the end of the song to see if any of the files still line up.
If two files line up exactly, then they're from the same digital source. One can be a digital clone of the other, or both can be digital clones of another file.
If two files line up exactly, I invert one of them, then adjust the volume level of one to try and get cancellation. If the files cancel down to dithering noise (typically -70 dB or lower), then I can say that one is digitally exactly a particular number of dB louder or quieter than the other. If the files cancel to a particular level (typically around -40 dB) but still leave some residual signal, and the residual signal has some odd EQ, then I can say that one is a differently-EQ'd digital clone of the other. If the residual signal shows some artifacts on the loud portions, like on the drum beats, then that can look like the artifacts of additional compression/limiting - you can see from the waveform if there's been additional compression/limiting.
I listen very carefully to the fades on all the tracks with a volume boost of 40 dB. You can learn a lot from the fades. If the high end disappears on the fade, that's an artifact of noise reduction. If the EQ stays the same over the course of the fade, that's good - no additional NR. With the 40 dB boost, you can also clearly hear if the fade is truncated or shortened.
To tell which came first chronologically, I use the copyright date on the disc. If I need to do a tie-breaker, I listen to the fade. The disc with the longer fade usually came first.
If two files don't stay synched perfectly, but drift apart by just a few samples, then zoom out and look at the waveform. If they show the same shape (and it will be obvious if they're the same shape), with peaks and valleys of roughly the same magnitude at the same locations, then I'd venture that they're from the same analog transfer.
In my experience, if two tracks stay closely in synch, say within 10 or 100 samples over a 3-minute song, then they're most likely from the same analog transfer. That's speed accuracy of something like 0.01% or 0.001% (I'm being approximate here - just going for an order of magnitude.) In practice, there's really no way to get your analog tape decks to have a speed accuracy of better than 0.01% or 0.001%; it just doesn't happen. I worked with analog tape for years, and the speed drifts were very real, and much larger than 0.01%. You could play a tape before lunch, and the same deck playing the same tape after lunch would be off a little. I'd venture that even if you wanted to, you couldn't synch two analog playbacks of the same tape to be that close in tempo.
And there are other tools you can use. For songs played to a click track or with a drum machine, you can use software to analyze the file and extract a tempo in BPM. I use a free piece of software called BPM Analyzer, which gives pretty accurate values to the hundredth of a BPM. (ex: 95.25 BPM) If I see two files with the same BPM down to the hundredth of a BPM, then that's another clue that it might be the same analog transfer.
When I talk about high-generation/low-generation source tapes, you can sometimes tell the difference in a side-by-side comparison wearing headphones. An instrument with a high-frequency content, like a shaky-percussion instrument or cymbals, may appear to be more clear for low-generation source tapes, and may appear lost or obscured for higher-generation source tapes.
The "preferences" that I call out in my posts are based on a few things. I detest noise reduction, so I usually avoid discs that have NR. I listen for low-generation source tapes, which usually points me toward the Rhino discs. I listen carefully to the fade, and if I can, I usually go for the disc with the longest tail on the fade. EQ variation doesn't bother me so much, so I don't mind the slight top-end boost on the Rhino discs. I'm not a big fan of compression/limiting, so I usually avoid modern-day remasters if I can. These aren't all hard and fast rules, just rules of thumb.
After a while, you can see patterns. Rhino discs usually use original analog transfers, or digital clones from earlier Rhino discs. Time-Life discs mastered by Dennis Drake are usually differently-EQ'd digital clones of tracks from earlier discs. A lot of the Warner Special Products discs use the same analog transfers as earlier discs. The Madacy discs (Rock On series) are usually digital clones from earlier discs.
This is a gigantic A/B comparison of 21 versions of "My Sharona" from the Steve Hoffman board. It has pictures, which should be helpful.
Feel free to ask if you have more questions. I haven't written any of this up in any formal document, so this post may be a good start.
Edited by crapfromthepast on 30 April 2015 at 7:02am
__________________
There's a lot of crap on the radio, but there's only one Crap From The Past.
|
| Back to Top |
|
| |
EdisonLite
MusicFan
Joined: 18 October 2004
Location: United States
Online Status: Offline
Posts: 2236
|
| Posted: 30 April 2015 at 11:00am | IP Logged
|
|
|
Thanks for your analysis of your analyses!
That all makes sense but one question:
<I then look at the end of the song to see if any of the files still line up. If two files line up exactly, then they're from the same digital source.>
I assume if you look at the end of the song files and one is a second or two shorter than another file ... you still consider that they could be from the same analog transfer, right? We all know on faded songs that someone can take a CD, and for whatever reason, fade it earlier (maybe they hear hiss and feel an earlier fade is the easiest way to get rid of it.) So I'm not sure you gave an example where two recordings are apart by one or two seconds, but you determine they're from the same original transfer. But I believe, from past descriptions, I have seen you make that comment. So I assume when versions differ by a second or two (or more), you still then determine if they're from the same source. Is that the case?
Edited by EdisonLite on 30 April 2015 at 11:01am
|
| Back to Top |
|
| |
crapfromthepast
MusicFan
Joined: 14 September 2006
Location: United States
Online Status: Offline
Posts: 2196
|
| Posted: 30 April 2015 at 12:37pm | IP Logged
|
|
|
That's correct.
Specifically, I look near the end of the song and see if a particular drumbeat near the fade lines up. If it does, that's one more clue that the two may use the same analog transfer.
If one fades a little earlier than the other (giving the two tracks different lengths), I can use that as a data point to determine which one was the "parent" and which was the "child", to stay with my cloning terminology.
One example that I can think of off the top of my head: Sniff 'N' The Tears' "Driver's Seat" on Eric's Hard To Find 45s On CD Vol. 8 is digitally identical to the version on Rhino's Have A Nice Day Vol. 22, but with an additional fade from 3:31.5 to 3:41.7 that recreates the 45 fade.
__________________
There's a lot of crap on the radio, but there's only one Crap From The Past.
|
| Back to Top |
|
| |
|
|
Active Topics 
Memberlist 
Search 
Help
Register 
Login
Top 40 Music on Compact Disc : Chat Board
Topic: Samantha Sang - Emotion
