Looking at the variety of exponential, parabolic, conical, tractrix, hyperbolic and whatever ales horn profiles out there I was thinking that we probably incorrectly approach the entire subject of horn profile analyses.

We listen the horns and we feel it is indicative enough. We apply to the horns an abstract logic and pretend that our horns are noses of submarine, winds of airplane or the jet propulsion exits. Yes, the theories are bult around all those commercial applications but those commercial application target completely different goal then we do in audio. In those commercial application the behavior of the exposed curves is the only thing that is important. If those application pay attention to air pressure that passes along the curve then they care about it only in a reference to the body of the curve, and the do not recognize the pressure wave as a self contained precious entry. Also all those commercial application and all engendering thinking that associated with it deal with essential parallel source of air. This is never a case with our audio horns.

Consequently, the mentioned (and few other reasons) made me to review our thinking about “an ultimate horn profile” and about our ability to interpret the auditable result and project those interpretations to the horns.

It is undoubtedly that the major characteristic of sound reproduction is a shape (profile) of the front of pressure wave. The engineers do not like to talk about it because they can’t objectively measure or visualize it. However, let pretend that the front’s profile is the key. If so, then it would be easy to understand that there is nothing more devastating for the front of a pressure wave then steps within sequential horn opening. Any, deviation from constant explanation rate would format the front of pressure wave and presumably screw the sound (let presume that inside of compression chamber there is no reformation of the wave’s front, that is not exactly true)

Ok, that was kind of all well-understood things but what we do after we understood it? We take a “wining profiles horn” and attach it to a random driver and then we build our theories how those profiles behave. We “kind of” forget that the compression drivers already have the built-in beginnings of the exponential channel. Each driver has own build-in opening rate and own length of the wave-channel. The only thing that we actual use is the diameter of the throat but we even not care that before the exit the driver might have 5”-6” of horn sitting right inside. Did I mention that we not really know or care what rate this inner-horn?

The point that I am trying to make is following: when we take for instance the Altec 802 and palace it into 1” tractrix and then “compare” it with 2” JBL 2440 then…. beside the differents in the drivers do we also deal with the COMPLIANCE of the specific “inner-horn” with the profile of our “outer-horn”?

Dose it means that the “ultimate horn profile” would be a profile that would be a DIRECT AND FLAWLESS CONTINUATION of the GIVEN SPECIFIC DRIVER?

Rgs,
The caT

Romy the Cat wrote:

<P>Dose it means that the “ultimate horn profile” would be a profile that would be a DIRECT AND FLAWLESS CONTINUATION of the GIVEN SPECIFIC DRIVER?</P> <P>Rgs,<BR>The caT</P>

Ah Romy, did you not know that the old JBLs, RCAs (and possibly S2 as well) all have an internal flare rate of 160Hz? This is but one reason why I'm asking Martin Seddon to make me a pair of 160Hz horns...

...just had a look; it may have been 180:
http://www.audioheritage.org/html/projectmay/technology/435be.htm

Ah well, what's 20Hz between friends.

You're right of course - we want to preserve continuity. This is something Earl Geddes has suggested is easier to do with the newer generation of "pancake" drivers, if I remember rightly, as far as his waveguide design goes anyway (superficially, it's really just a conical with more horn like smooth throat transition).

cv wrote:

You're right of course - we want to preserve continuity. This is something Earl Geddes has suggested is easier to do with the newer generation of "pancake" drivers, if I remember rightly, as far as his waveguide design goes anyway (superficially, it's really just a conical with more horn like smooth throat transition).

cv wrote:

This is but one reason why I'm asking Martin Seddon to make me a pair of 160Hz horns...

Chris, be carefully with those large 160Hz horn and S2. The guys who make then have no instruments to center the mounting holes very precisely (wood-turning machines apparently do not furnish this level of precision, not to mention the concrete pouring machines :-) As the result, if you have a compression driver with own channel, and the driver is not exactly centered within your external horn (a fraction of millimeter does mater) then you would have a slight suck-off at middle. For instance my horn has a centering mechanism   (courtesy to John Hasquin) that allows centering the driver. I use a high resolution RTA (0.25db at 1/12 octave) running a sweep and moving the driver. Wherever I get a perfectly flat response at 4kH there I would consider the drive in the centered. Moving the driver within 2-3mm introduce an octave wide drop between 3kH and 5kH with minus 2-3 db at 4kH. Some of the drivers are less sensitive and particularly the drivers where the throats at the end of the phase plug but if you driver has own inner-horn than the centering it within the outer-horn become very important.

Rgs,
The caT

Very good point - I will direct Martin's attention to this. Visually, it looks like he did an excellent job on the 340 Hz horns, and the 160Hz horns will be rolled off at 1k anyway but nonetheless, will see what he reckons. Actually, no need - I just remembered: he got a machine shop to turn a threaded mould in ally, with a precise 1.5" plug protruding at the end; he then moulded the horn throat around that in resin. So in theory, it should be pretty damn precisely centred by design as the horn threads on to the driver (or is it the other way round?). Still, when I get some measurment equipment setup I'll check the 3-5k response to be sure.

JMLC has mentioned the need for precise centering in the past, and it's another thing that struck me about all the WE 555/594 users etc - they take what may or may not be wonderful drivers then attach them to some discontinuous *throat adaptor* using something that was salvaged from a shower hose...

JLH wrote:

In addition, the idea of an ultimate horn profile does not make any sense. Each and every horn profile has its own strengths and weakness. Depending on your goals, and the drivers you have, one profile will be "better" than the others. However, once you change your goals, or the driver, room, source, etc. a different profile will become the "best". There are no absolutes in Audio, only varying degrees of perfection.

I did not pitch the “ultimate horn profile” as a competition between the profiles and I rather called this thread like this satirically. The point was the we are considering the advantages of one curve over another but forget that we have the major  “step” between the internal and external curves of drivers with completely unpredictable sonic consequences.

The caT