Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/3/84; site teddy.UUCP
Path: utzoo!linus!philabs!cmcl2!seismo!harvard!talcott!panda!teddy!rdp
From: rdp@teddy.UUCP
Newsgroups: net.audio
Subject: Re: RAW SPEAKERS
Message-ID: <1134@teddy.UUCP>
Date: Tue, 13-Aug-85 13:48:01 EDT
Article-I.D.: teddy.1134
Posted: Tue Aug 13 13:48:01 1985
Date-Received: Thu, 15-Aug-85 08:24:57 EDT
References: <3177@decwrl.UUCP> <975@teddy.UUCP> <914@druxo.UUCP> <988@teddy.UUCP> <334@ttrdc.UUCP>
Reply-To: rdp@teddy.UUCP (Richard D. Pierce)
Organization: GenRad, Inc., Concord, Mass.
Lines: 129

[]

In article <334@ttrdc.UUCP> kyl@ttrdc.UUCP (Kwing Y. Lee) writes:
>Dick:
>
>    When I first purchased the Klipsch speakers time and phase coherence were
>just started to be the main design goals of speakers (Dalquist DQ10s). Back then
>the criticisms that I heard about the Klipsches are that they are too big, too
>expensive and they don't have the high end response that they should have....
>and also the so called "horn sound". Now nearly 10 years later, I still have not
>outgrown by babies and I still think they sound fantastic. I would really 
>appreciated if you can summarized all the criticisms of Klipsch horn speaker
>systems in one single comprehensive article. Thanks in advance.
>
>                                                  Kwing

Ok, I will try.

First of all, let me say that, as far a cabinetry goes, I cannot impune
the manufacturers at all. The Klipsch K-horns ARE solidly built, well fitted,
nicely finished, and, for something that large and awkward, are attractive
in there own occasionally endearing way. And I cannot find a single chink
in the armor of its progenitor, Paul Klipsch.

However...

The prime advantage to K-horns, and things of the horn genre is efficiency.
If that is the prime criteria, then I can have no objections.

Horns, realized in a practical manner, have real problems.

First, the K-horn bass driver (as I can recall) is a fairly old style EV
woofer, specifically designed for direct radiator use, albeit that alone
does not acount for most of the problems. The bass enclosure, a severely
folded horn, is, at best, a compromise. It is only a rough approximation
of an exponential taper, and therefore cannot provide a continuous match
along its path to the final load, the room. This causes problems such as
internal reflections, cancellations, what have you. The sharp bends also
cause reflection problems, as well as non-linear attenuations at the what
might be considered only moderately high frequencies for a straight horn.
At the the throat of the horn, measured particle velocities are high
enough to cause non-linearities in propogation characteristics. Add also
the fact that there is a many millisecond delay, which changes non-linearly
with frequency.

The net result is that the bass section has (as a measured fact) a very
ragged frequency response, which has been shown by some researchers to
level dependent. Mr. Klipsch continuously raises the spectre of "Doppler
distortion" as the bugaboo of all direct radiator designs, yet all of the
research papers I have been able to collect are, at best, confusingly
undecided in total about either the existance or the detectability of
such distortion. (H. D. Harwood, in the early 70's published an article
which both philosophically and experimentally dealt a severe blow to
Klipsch's assertions. I will try to resurect it if desired).

Note also that because K-horns need the entire room to attempt to couple
properly, they can be far more room dependent than direct radiator designs,
judiciously placed.

The mid-range and treble drivers are fairly mundane straight exponential
horn designs, nothing great, nothing terribly bad. However, Klipsch makes
the same, to me, very stupid mistake when it comes to orrientation. The
best, most uniform and frequency-independent dispersion occurs at right
angles to the long dimension of the horn. This means that in the Klipsch,
where the long dimension is horizontal, the best dispersion pattern is
vertical!. The measured dispersion characteristics of the speaker are
truly dismal in the 1000 Hz and up regions of tghe spectrum. I will concede,
however, the impracticality of re-orienting the drivers on such an already
huge cabinet.

The crossover is fairly simple-minded and straightforward, and that may,
in fact, be it's saving grace! But..

Because of the tremendous phase and delay anomolies at the various crosover
points, both the response characteristics and the dispersion characteristics
at those frequencies are truly bizzarre!

Since the indtroduction of more sophisticated measurement techniques, the
measured characteristics of the speakers have been even stranger than even
I would imagine. IMpulse response is very poor. The speaker has been measured
with incredible long decay times, complex rise characetristics, and truly
non-linear and non-minimal phase characteristics.

Now, be that all as it may, SO WHAT, you say. 

I am also a musician of sorts (more a listener, though), as well as a
reasonably accomplished instrument builder (I have built 11 harpsichords
or clavichords, 2 small portative organ, one positive organ, and have
repaired and restored many instruments ranging in age up to 300 years).

Simply stated, music played on instruments that I am familiar with are
reproduced very unfaithfully on Klipsh loudspeakers. Harpsichord music
looses completely any hint of clarity, inner voices are completely
muddled, and the higher registers are at once both very brittle and muddy.
Organ music, recorded properly looses all of the "pipe noises" that can be
easily heard in person and on other speakers of less technological
pretention. Yes, K-horns certainly do play very loud, but so what? The
kind of music I listen too, and the vast majority of classical music can
be more accurately reproduced on lower effeiciency speakers of demonstrably
higher fidelity. The efficiency and dynamic range arguments put forth by
many horn proponents are not supportable, given the REAL requirements of
reproducing music. The exception might be for rock, where sheer acoustic
power is needed to replicate the effect of a "live" concert, but here we
have no live analog to compare against anyway. Most proffesional PA systems
are so much worse than K-horns or any other home speaker as to be
laughable. (I once designed and built a set of high-power PA speakers for
a rock band. They were VERY good. They were smooth, low distortion, wide
bandwidth, etc. Nobody liked them.)

Now, do I object to you liking your Klipsches. No, absolutely not. What I
object to is their inability to faithfully reproduce musical sounds that
I know well, in comparison to many other loudspeakers. I object to statements,
such as one of the ones that followed your article, that "horn sound" is
smooth, uncolored, low distortion. On an objective basis, K horns and most
comercially available horns are neither smooth, uncolored or low distortion.
At least two of those quantities can be measured objectively, and the horns
fail both very miserably.

I will concede that K-horns are probably the best, most practical commercial
realization of a horn intended for home use. It's just that horn aren't
very good.

But then again, I like Harpsichord music by Francois Couperin, so I must
be nuts anyway.

Regards..


Dick Pierce