Blind Shoot-out in San Diego -- 5 CD Players


On Saturday, February 24, a few members of the San Diego, Los Angeles and Palm Springs audio communities conducted a blind shoot-out at the home of one of the members of the San Diego Music and Audio Guild. The five CD Players selected for evaluation were: 1) a Resolution Audio Opus 21 (modified by Great Northern Sound), 2) the dcs standalone player, 3) a Meridian 808 Signature, 4) a EMM Labs Signature configuration (CDSD/DCC2 combo), and 5) an APL NWO 2.5T (the 2.5T is a 2.5 featuring a redesigned tube output stage and other improvements).

The ground rules for the shoot-out specified that two randomly draw players would be compared head-to-head, and the winner would then be compared against the next randomly drawn player, until only one unit survived (the so-called King-of-the-Hill method). One of our most knowledgeable members would set up each of the two competing pairs behind a curtain, adjust for volume, etc. and would not participate in the voting. Alex Peychev was the only manufacturer present, and he agreed to express no opinion until the completion of the formal process, and he also did not participate in the voting. The five of us who did the voting did so by an immediate and simultaneous show of hands after each pairing after each selection. Two pieces of well-recorded classical music on Red Book CDs were chosen because they offered a range of instrumental and vocal sonic charactistics. And since each participant voted for each piece separately, there was a total of 10 votes up for grabs at each head-to-head audition. Finally, although we all took informal notes, there was no attempt at detailed analysis recorded -- just the raw vote tally.

And now for the results:

In pairing number 1, the dcs won handily over the modified Opus 21, 9 votes to 1.

In pairing number 2, the dcs again came out on top, this time against the Meridian 808, 9 votes to 1.

In pairing number 3, the Meitner Signature was preferred over the dcs, by a closer but consistent margin (we repeated some of the head-to-head tests at the requests of the participants). The vote was 6 to 4.

Finally, in pairing number 5, the APL 2.5T bested the Meitner, 7 votes to 3.

In the interest of configuration consistance, all these auditions involved the use of a power regenerator supplying power to each of the players and involved going through a pre-amp.

This concluded the blind portion of the shoot-out. All expressed the view that the comparisons had been fairly conducted, and that even though one of the comparisons was close, the rankings overall represented a true consensus of the group's feelings.

Thereafter, without the use blind listening, we tried certain variations at the request of various of the particiapans. These involved the Meitner and the APL units exclusively, and may be summarized as follows:

First, when the APL 2.5T was removed from the power regenerator and plugged into the wall, its performance improved significantly. (Alex attributed this to the fact that the 2.5T features a linear power supply). When the Meitner unit(which utilizes a switching power supply) was plugged into the wall, its sonics deteriorated, and so it was restored to the power regenerator.

Second, when we auditioned a limited number of SACDs, the performance on both units was even better, but the improvement on the APL was unanimously felt to be dramatic.
The group concluded we had just experienced "an SACD blowout".

The above concludes the agreed-to results on the blind shoot-out. What follows is an overview of my own personal assessment of the qualitative differences I observed in the top three performers.

First of all the dcs and the Meitner are both clearly state of the art players. That the dcs scored as well as it did in its standalone implementation is in my opinion very significant. And for those of us who have auditioned prior implementations of the Meitner in previous shoot-outs, this unit is truly at the top of its game, and although it was close, had the edge on the dcs. Both the dcs and the Meitner showed all the traits one would expect on a Class A player -- excellent tonality, imaging, soundstaging, bass extension, transparency, resolution, delineation, etc.

But from my point of view, the APL 2.5T had all of the above, plus two deminsions that I feel make it truly unique. First of all, the life-like quality of the tonality across the spectrum was spot-on on all forms of instruments and voice. An second, and more difficult to describe, I had the uncany feeling that I was in the presence of real music -- lots or "air", spatial cues, etc. that simply add up to a sense of realism that I have never experienced before. When I closed my eyes, I truly felt that I was in the room with live music. What can I say.

Obviously, I invite others of the participants to express their views on-line.

Pete

petewatt
Sabai: Sandra Day O'Connor once said: one key lesson She learned throughout her career is that (and I am paraphrasing here) "once you made your point just shut up and go away"!"
the problem with such a test, is that the results cannot be extrapolated onto one's own stereo system, i.e., one would best listen to the same digital hardware in the context of one's own system--true for any component, to make a meaningful judgment.
Once upon a time, Donald Rumsfeld wrote the following verses:

"As we know,
There are known knowns.
There are things we know we know.
We also know
There are known unknowns.
That is to say
We know there are some things
We do not know.
But there are also unknown unknowns,
The ones we don't know
We don't know."
11-09-10: Mrtennis
the problem with such a test, is that the results cannot be extrapolated onto one's own stereo system, i.e., one would best listen to the same digital hardware in the context of one's own system--true for any component, to make a meaningful judgment.

100% true, that isn't going to happen for most, if not all of us. So can we extrapolate, not perfectly, but better than nothing, which is what we would be left with if we had to test all five units in our own system - unless, I really had nothing to do all day. I want to see more shoot outs with feedback from the groups, and I'll add my own grain of salt to the results, for I think they are far from being meaningless, as long as we don't take these things as definitive - they are not, and mowst of us know and accept that.

OP, how about 5 power cords next?:)
hi pubul57:

i have a question for you:

suppose you hear of a comparison between 5 similar components, e.g., cd players, and the number of listeners exceeds 100. suppose one of the players is favored by more than half of the listeners. do you think you would favor that cd player over another chosen at random within the same price range ?

in other words, what is the value of the results of a "shoot out" consisting of 5 cd players when there exist over 1000 digital soiurces (must include dacs and transports)?

it would seem to me the results are a microcosm of what is possible.

i would propose a different experimental design, impractical as it is, which would require more listening and a greater period of time to accomplish the results.

here it is:

select over 100 cd players or dacs/transports, listening to 5 at a time. from the first 5 select the "favorite.

from the remaining candidates, slecet another five and pick a favorite. at the end, there will be 20 favorites.

narrow down the twenty to five and report on those 5 out of one hundred.

my point is that 5 is too small a sample size to be very useful. there are many other digital "players" which could have been selected..

however, i will agree, the exercise is not useful, although the idea could be implemented to include more candidates, even if it requires months to complete the listening.

perhaps , there are some serious audiophiles who could perform a valuable service by comparing a large number of , say tube and solid state amplifiers, capable of producing a reasonable spl, in a given stereo system.

let's not try to get crazy here as there are many variables, but the idea of experienced listeners comparing equipment is of value if implemented using a larger sample size than 5. there may be some short cuts to reduce total listening time. it takes some clever people to design such an experiment.