"Michael McKelvy" wrote:



"Nousaine" wrote in message
...
"Robert Morein" wrote:




"John Atkinson" wrote in message
. com...
"Powell" asked:
"John Atkinson" wrote
I did a thorough study of this back in 1992. The results are
reprinted at http://www.stereophile.com/features/806/. Supporting
a speaker on spikes rather than a lossy interface of some kind
allows the cabinet resonances to develop fully.

June 1992? I don't see it in the index. Where in the 322 pages
is it?

324 pages (we start the folio numbering on the front cover, which is
page 1). The article on cabinet resonances starts on p.205 and
continues
through p.207. I then followed up the subject in September 1992. All
my discussion from both issues is included in the Web reprint that I
referenced above.

Great article!

As I write this, I'm having a telephone conversation with my friend Larry
Zeitz, who insists there must be an escape clause for downward pointing,
carpet piercing, floor spikes on floor standing speakers, and the
interface
between stands and floors.

I have responded that your article offers no implication that spikes are
useful for the floor interface.
Would you be so kind as to comment?

About this time I conducted a single blind listening test to find the
worth of
speaker support and sonics. I purchased 4 Snell KII loudspeakers with the
assurance that they had been sequentially manufactured and verified that
they
all met the specification of 0.5 dB frequency response relative to the
standard. Of course i didn't have that standard speaker but each of the 4
were
well within a half dB of each other.

I then borrowed a set of target speaker stands, filled each with 25# of
lead
shot, and affixed spikes to the bottom of the stands and the stand pads
that
would anchor the speakers.

Wishing to maximize the possible sonic differences I then devised a
competing
"stand" which was an empty 12-inch woofer carton, duct taped to a stamped
steel
Bose 901 stand whch was then brought to height with copies of the JAES and
paperback books.

I then conducted a number of single listener, single blind listening tests
with
randomized speaker location. The results were that listeners had a
preference
for speaker location BUT stand-interface wasn't a decision making element.

These were listener switched tests BTW. Each listener had control over all
aspects of program delivery.

And were you surprised at the results? :-)

No; frankly it didn't surprise me. But to my knowledge no one had (or otherwise
has to this date) conducted a bias-controlled listening test compensating for
speaker position as a variable examining the conventional wisdom.