sealed vs vented subwoofers

NO measurement or design principal has ever been
shown to give a close correlation with sound quality.

I agree in the sense that no single measurement does - it is usually a
combination of measurements and design principals that correlate to sound
quality. Often there is more than one way to skin a cat.

BTW - The student who ran these subwoofer tests in a parking lot has been
hired by Genelec - to
work in their R&D department. He will have access to anechoic chambers
now. I suspect the parking lot is just a way to get raw baseline comparable
measurements under controlled conditions - for sure these need to be
interpreted carefully. As Duke points out - a roll off is probably more
desirable than a flat response to 20 Hz - due to the in room wall boundary
boost effect.

Genelec is not well known to audio consumers but they have a strong
following in the music recording business.

Stanwal, I almost agree with your statement about the failure of measurements to correlate with subjective preference. I think it's fair to say that the audio industry has been measuring distortions that are easy to measure instead of measuring distortions that correlate well with human hearing. However, there has been progress made in this area recently which you are probably unaware of.

If you have access to the Audio Engineering Society's library, you might want to take a look at "Auditory Perception of Nonlinear Distortion - Theory" and "Auditory Perception of Nonlinear Distortion", both by Earl Geddes and Lydia Lee. Therein you will find described a rather complex metric that correlates well with subjective perception. The industry has largely ignored it, aside from this article from Audio Express magazine:

http://www.gedlee.com/downloads/THD_.pdf

See also this letter in response to the article:

http://www.gedlee.com/downloads/Comments%20on%20howard.pdf

Duke

Duke,

Thanks for the link - interesting reading. In simple terms, GedLee is saying
that higher order harmonic distortion is much worse than low order
distortion. This means 2nd order is better than 3rd order ...etc. etc. This jives
with everything we know - Class A amps sound better at low volumes - IMD
distortion is the worst - and that the "masking" effect means that
we may not hear nearby frequencies to a fundamental as easily as we might
hear a 9th harmonic (BAD).

This matches what Ralph has said so many times on these forums...high order
odd harmonics are bad - even in relatively much smaller amounts. To take
your example above, 30% second order harmonic distortion (barely or not
quite audible) may be akin or equivalent to 0.3% distortion in the 9th
harmonic. In that sense, an amplifier with THD of 1% all in the 9th harmonic
would likely sound much worse than an amplifier with 10% THD but all in the
2nd harmonic.

One could jump on this and say that all measurements are meaningless,
however, one must reflect that if an amplifier has a measured THD at full
power of less than 0.004% (vanishingly small) then it will likely sound good
anyway - irrespective of a GedLee higher weighting to the higher order
harmonic distortion (as, be it low order or higher order, the distortion is
simply very small).

Perhaps the problem (what listeners observe) begins when you hook up an
amplifier to a complex load and make the poor amp send bucketloads of
current to drive the woofer and then mere milli-amps to drive the delicate
little tweeter. When the rubber hits the road (in the real world and not a lab
test) the amp find itself being asked to perform two rather diametrically
opposing tasks: extreme butterfly wings delicacy and elephant brute force. A
case where IMD distortion seems inherently likely - so why does the industry
stick so vehemently to this design approach? And why is GedLee largely
ignored in manufacturer spec sheets?

Duke, thanks for those articals.

Shadorne, check out Nelson pass's 11/1/08 artical on his website entitled " Audio, Distortion and Feedback." I think he's right when he says that preference for 2nd or 3rd order harmonic distortion is listener dependent. IMHO it's not so simple to classify lower order distortion as uniformly preferable. - Jim