Dark Side Of The Moon on CD: a blind listening test

Hello there,
a frequent question on this board is which version of a particular album sounds the best.
I admit I'm obsessed with Dark Side Of The Moon, it seems I have different tastes and favourite mastering in different periods.
I'm an engineer and I enjoy the visual approach of frequency graphs, but the actual listening experience is what's really worth, in my opinion.
I'm also a psychologist and I think there could be a response bias by people who spent a lot of money on a Black Triangle or UDCD pressing and would encounter a cognitive dissonance if they found out it was not worth the money (I don't mean it's necessarily so, mind you).

So, if you're in the mood for a blind listening test, here you have six 1-minute excerpts of Money, from different CD masterings (A, B, C, D, E, F), in FLAC format. The tracks have been level adjusted in order to sound approximately at the same volume.

https://www.dropbox.com/sh/wd4dk98pbbzsbej/AAD1A2Dn_ruwhvKbnlKNm1EWa?dl=0

It would be great if you could listen and compare the samples, reporting which one sounds the best on your equipment and why, elaborating on what you hear. I'll reveal the sources at the end of the test.

I would consider extending the test to other songs from the same album (new poll), if it turns out there's interest in it.

Any suggestions are appreciated.
Look forward for your feedback.
G.

 

Not exactly a "golden ear" but I think my preference is for "C". It seems to me to have the crispest highs, and I tend to like crispy highs.

 

Whilst I agree with most of that, the Non-TO Black Triangle/Harvest has Pre-emphasis. You have not mentioned how you have handled that in your samples. If you have accounted for that via software still it will not be same result as when properly decoded by a circuit that correctly adheres to the Redbook standard. Also it is paramount to remember that the Pre-emphasis on the Black Triangle/Harvest is not in the TOC but in the Subcode. To best of my knowledge most software including iTunes [crappy pre-emph eq decode - makes stuff smiley IMHO] will be able to detect it as a result. In as far as hardware many players including Oppo wont detect it. As far as EAC to detect you need an ancient version and compatible drive [usually older]. When not decoded correctly the non-TO BT/Harvest will sound pretty darn bad. When done properly it then it brings out all it's high gen tape wonder

I understand what you are trying to do, and my apologies but due to the pre-emphasis I find your test/poll to be 100% invalid. I am attacking the method from a science|factual standpoint and not you the person. Hopefully as a professional you will take it in only that manner.

IMHO this thread should be deleted unless one's intention is to perpetuate false or rather incorrect data. If proving cognitive dissonance is the true goal, perhaps using another example where none of the discs has the pre-emph is in order

Pre-emphasis is a PITA

 

Here's a good list a few of us having been adding to over the years...
http://forums.stevehoffman.tv/threads/pre-emphasis-list.121188/page-26
Ricks is correct, Oppo cannot de-emph discs in which pre is found in the SUB, only TOC, Glad I had an older Denon player I kept as an emergency SACD player, as I have a few of them

 

Thank you for your comment.
I ripped the track directly from my CD player/amplifier stage, transferring the audio using a EMU 0202 sound card. I carefully checked the frequency plot after the rip, comparing it to the pre-emphasized track and I believe the de-emphasis is accurate enough: I can clearly see the 0-6dB linear boost in the 3.183-10.610 kHz range.

I'm aware that this method introduces potential noise due to the D/A/D conversion, but the de-emphasis should be performed in the analogue domain, so a FFT would not yield the same result. I read several threads in different forums and I gather that all audio programs, like Sox or WaveEmph, are inaccurate and introduce some deviation from the correct de-emphasis curve, linear phase distortion and pre-echo.
So the CD-rip method is the safest, in my opinion.

Having said that, I believe the CD-rip method will not invalidate the test, i.e. if the "holy grail" mastering is actually the most preferred one, this fact will emerge.

 

I didn't say which is the mastering I usually prefer (note: I own a Harvest blackface non-TO, and it cost me no more than €15, same price I paid for the 1992 Sax-mastered version). My only goal is trying to understand which mastering of DSOTM is the preferred one using one's ears. I believe only a blind test can reveal that.
If people actually think my method for handling PE is not accurate enough, I will remove the "holy-grail" mastering.

 

My favourite is the one I'm listening to...sorry

 

If you recorded from your Denon's analog out you should be fine, if my 1940ci can add the emph curve found in the SUB yours definitely should be able to, I think your good

 

That's what I did.

 

I prefer E and then A

 

Just reading this thread make mah brane asplode

 

Do not read. Just listen. It's all about listening

 

My first pick is E, followed by A, though A seems to have a little extra EQ boost in the highs. Looking forward to seeing which sample is which mastering.

 

So far I can only say I don't like D.
I'm keeping on listening.

 

While your method for the pre-emph is about the best that could have been done, the test IMHO is still too flawed for a valid scientific compare.

Bottom line, performing this kind of listening test from a true scientific standpoint needs to be done using one of these 3 methods to get true apples to apples without each person having the physical discs.

1) use samples only discs that do not feature pre-emph
2) use samples from discs that all feature pre-emph
3) for all entries in your compare list use samples captured from your CD player/amplifier stage as you did for the pre-emph discs; thus apples to apples.

A and B require a different album. Some nitpickers might still have issue with #3, but I won't.

BTW while I do own the BT "Holy Grail" and the Harvest non-TO, which of course sound exactly the same, I simply have no desire to ever listen to this overrated album. Honestly, I prefer Division Bell to any of Floyd's 70's output. Perhaps because it did not involve Waters?

 

"Still too flawed" sounds much better than "100% invalid" so I think I'll content myself, as some folks already voted. I believe the method, albeit sub-optimal, will work. Next time I'll possibly follow your "apples to apples" advice (I'd like to create a test on Wish You Were Here).

I'm sorry you won't take part in the game. By the way, I prefer Animals.

 

With my Pono and ATH-M50's I felt D was the best with smooth mids and spaciousness. Many of the others were a little bright, but that's sort of how those headphones are (a little bright).

 

Good idea to do the test thru a Pono !
I was listening thru my computer's headphone output. Not necessarily the best when you're looking to hear SQ.
I will upload the extracts in my Pono tomorrow

 

1-a
2-e
3- d
4-f
5-b
6- c

 

First of all they all sound great. I like a bassy sound so I've gone for D
A: Fuller sound
B: better separation slightly louder (then A)
C: similar to A
D: bassy
E: slightly different mix with rear separation
F: I was over the song by then

 

I liked 'D' the best.

'F' is the runner up.

 

I don't need to listen to any of them -- I've been well trained by this blog over the years: original cd release, black triangle!

 

D, is my choice, smooth and warm
Listened via Sqeezebox Touch through a Emotiva DC-1 DAC

 

It's been 43 years and I've let it go. SACD for stereo and original quad from the Immersion box.

 

I'd have to agree. E was nice for the clarity of the interplay of the bass, keyboard and guitar in the latter half of the track.