AudioBanter.com - View Single Post - Why DBTs in audio do not deliver (was: Finally ... The Furutech CD-do-something)

(Stewart Pinkerton)
wrote:

On 15 Aug 2003 21:30:55 GMT, (randyb) wrote:

(Stewart Pinkerton) wrote in message
...
On 15 Aug 2003 14:34:45 GMT, (Nousaine) wrote:

My point isn't that every
amplifier used in every test hasn't been replicated but that many tests
had
been published but none of those had results that matched those claimed
by
Stewart.

That's OK; but until his results become documented and are replicated
they
remain anecdotal.

So, how soon can you get here? :-)

Ok, I have not been around this group that long, but I am pretty well
on the objective side of the fence. My question is that both you,
Stewart and Tom Nousaine seem to be objective, but apparently Stewart
has come up with a different set of results in a DBT. Could you both
elaborate without me searching the archives as to what was tested and
the results. Thanks.

We were burgled (grrr!) and I had to replace my TV sound system, so I
was in the market for a new amplifier. I had just purchased my Apogee
Duetta Signatures and Krell KSA-50mkII amp, so I had a good
'reference' amp and a tough speaker load with exceptional clarity and
'openness'. I had also settled on Tannoy 633 speakers for the TV
system (in my view, you buy speakers you like, then you buy an amp
which can drive them).

I assembled a 4-pole switch-box with relay connections, and a 3-way
switch enabling me to select A, B or 'X', with X being set by a hidden
toggle switch on the relay box. The 'X' switch had a central 'off'
position, so that each reset used two clicks of the switch. I also
made up a 4-channel attenuator, enabling me to set the gains of each
channel of a pair of stereo amps to be matched within +/- 0.1dB.

Next, I entered negotiations with some local hi-fi dealers, and over a
period of a couple of weeks or so, I was able to borrow a pair of
Denon POA6600 monoblocks, a Hafler XL600, an Arcam Xeta One, a Yamaha
2090, a Yamaha AX-570, A Rega Elex, A Musical Fidelity E600, an
Audiolab 8000A, and an Audiolab 8000P.

I set up a series of 20 trials of each amp, with my wife doing the
switching. She flipped a coin 20 times, and noted on a score sheet
whether it was heads or tails. She set the 'X' switch to A for heads
and B for tails after each test. I'd satisfy myself as to the identity
of X and mark it on a score sheet, my wife would enter the room, reset
the identity of 'X' and leave the room while I conducted the next
test.

To we terminally tweaky/freaky we'd have you leave the room when the switch was
made as well. 20 trials is a good number.

The end result was that the Denons (fave rave reviews in all the rags
at the time) and the Rega were instantly tossed for rough treble, the
MF was instantly tossed for a 'veiled' sound with recessed treble, the
Yamaha 2090 and Arcam were tossed for slightly bright treble, while
the remaining amps were too close to call.

So you have results for all these comparisons? The next obvious question is
what was the reference for any given device? Krell? What was the reference
"toss'' point for the *******, as it were

But it seems that you didn't confirm frequency response into the load either.
Not a critical issue but one that I generally care for by checking level match
at 100, 1000 and 10,000 Hz.

Retesting of the Audiolab
8000A revealed a *slight* veiling of the sound, but it just crept into
the '16 out of 20' target score, while the Yamaha AX-570, the cheapest
amp on test, was a real surprise. I got 15/20 correct, but only
revealed by a slight treble brightness on viloin solos. The Hafler and
the Audiolab 8000P were sonically identical to the Krell, no better
than 12/20 on original or retest. The Hafler, probably the best
all-rounder as it was very powerful as well as quiet and sonically
impeccable, was rejected on the grounds of its horrendously noisy
cooling fan,

This is a REAL reason for rejection and needs no corollary tests. But if it
was sonically impeccable (I would guess that would be transparent to whatever
amplifier which was the comparative reference?) and could be placed in a sealed
cabinet or adjacent room it would then be acceptable.

leaving the Audiolab 8000P as the 'winner'.

Interesting that a sonically impeccable product gets rejected for having a
noisy fan

Good reason. Better than most.

Truth to tell, on the Tannoy speakers I couldn't hear any significant
deterioration with the AX-570, but I liked the compact size and 'less
is more' approach of the Audiolab, so I stumped up the extra, and it's
still serving me well after about 8 years, as is the venerable Krell.
I should perhaps mention that I'd actually traded in an older 8000P
against the Krell when I bought the Duettas, as the Audiolab got
*very* hot after playing some heavy rock music through these
insensitive 3-ohm monsters. The Krell didn't put out any more power,
but being designed to play continuously into 1 one-ohm load, it wasn't
at all bothered about driving the Apogees.

As a bizarre aside to the above, I published these test results just a
few weeks before the notorious 'Sunshine Trials', where a very vocal
(sadly now deceased) 'audiophile' dealer failed to distinguish his
favourite exotic Pass Labs Aleph 1.2 monoblocs from an old integrated
Yamaha amp (an A-700, IIRC), on his own reference sound system. Steve
Maki and Tom Nousaine proctored this test, and Tom was later accused
of bringing in the Yamaha as a 'ringer', in view of the exceptional
performance of the AX-570 in my own tests. Obviously a gross cheat,
eh, sneaking in a good-sounding $500 integrated amp to beat up a
$12,000 pair of 'designer label' power amps!
--

Stewart Pinkerton | Music is Art - Audio is Engineering

I have a bunch of "ringers"! At most price levels and even ages.