Jean Michel on LeCleac'h horns

[Pano]

Good questions Bear.

Big Lip vs baffle blend. What's the diff?

[soongsc]

I personally would like to blend into something like and egg or sphere. But I wouldn't debate on it.

[Pano]

LOL! OK. I've been following your sims and they made me curious about what best to do behind the horn.

If the horn rolls over all the way, is that enough?
If the horn mouuth blends into a flat baffle, is that enough?
(how should it blend?)


All of that is going to sound different. What is a good approach?

[soongsc]

I think most people will probably not hear the difference due to listening environment or personal perception.

If the horn were build into a wall, it probably makes sense to blend with the baffle. I think it makes sense to guide the wave so that back/sides will have contiguouse rolloff like the off-axis does to the extent practical. The exact shape probably would not make too much difference, but I really don't have enough data to be certain. Wanted to do more sims, but the time it takes to do it is really too long. The computer just get messed up after 3 days non-stop sim.

[sqlkev]

bear and paro
i've been meaning to *hear* the differences in different lip profiles/curvatures too
you can see a bit of simulation in the hornresp software but i don't think that's entirely too accurate since i dont think the software factors in the linearity of the frequency response of the compression/dome/cone driver being simulated

i think our best bet here is to go with what Jean-Michel suggests and hope for the best (the reason i'm saying this is because personally...i do not have the luxury to make a few different sets of horns to play around with)

[Jmmlc]

Hello Bear,

From what you wrote I can see that there something that probably I didn't clearly explained.

In fact the Le Cléac'h horn doesn't possess any lips! I mean there is no lips "added" to the mouth as it is necessary to do in a waveguide or a conical horn in order to reduce diffraction at mouth.

What you and other call lips are in the case of the Le CLéac'h horn is purely the result of the natural combination of:
1) the expansion law of the area of the wavefronts (e.g.: hypex, exponential)
2) the necessary parallelism (= equistance) of the curved wavefronts (defined as equiphase surfaces).

When calculating a Le Cléac'h horn, there is nothing special in my spreadsheets or software related to the curvature of the mouth. The curved back mouth is not intended nor provoked, it is the result of the "design by nature" of the horn.

Now on the point of baffling against rolled back mouth. If we stop the profile of a Le Cléac'h horn when the opening angle becomes 180 degrees and if from this point we decide to add an infinite baffle, we introduce a perturbation of the expansion law. This perturbation induces a partial reflection of the wave energy back in the horn and have effects on the acoustic impedance curve, and therefore on both the response curve and the group delay curve and on the pulse response (waterfall, CSD...). Those effects will be more efficient in the low frequency range of the response of the horn.

That's why I prefer to use unbaffled Le Cléac'h horns the mouth of which open at 360° minimum (Hornresp allows a max profile tangential angle of 180° so an opening angle of 360°).

You may see in attached file a theorical graph of the wavefronts I calculated. Elsewhere in the thread you can compare to the simulated pressure field at low frequency.

Best regards from Paris, France

Jean-Michel Le Cléac'h

Originally posted by bear
J-M,

Been following along with this thread.

I am unsure of some of the things I am seeing and reading.

Regarding a "lip" or "radius" that extends past the point of being perpendicular to the main axis of the horn...

Are you saying that the lip of the horn is best terminated by a large fat lip of considerable size? Did I see a curve where the outer edge of the lip is brought all the way back around to the rear of the horn?

If this is correct, what role does are area that intuitively appears to be unrelated physically to the wave that is launched have?

I think I am seeing that there is a suggestion that terminating the horn to a flat large ("infinite"?) baffle has benefits... are we talking after a curve/radius of some dimension has reached the point of being tangent to the plane that is perpendicular to the main horn axis? Does a larger flat baffle "work" with a "lip" that extends back past this plane (that being the "front of the mouth" where the curving horn mouth meets it)?

And, are we then saying that almost any horn expansion might then benefit from an arbitrary "lip" of sufficient dimension? And, then what is the relationship between the "lip" dimension and frequency/wavelength?

Hope I am being clear enough...

_-_-bear

[Jmmlc]

Hello Panomaniac

You wrote:


It's easy to see that the big, round horn beams more than the more complex JBL & Beyma.
I've noticed that very "pure form" horns tend to beam badly.

In the past the increase of directivity with the frequecny rise was not a problem...

Beaming became a topic of discussion between audiophiles quite recently. I am not alone to think that there was marketing issues behind that. Many audiophiles have been foolished by such claims as control of the directivity being the most important parameter for good reproducing of music. IMHO (but my opinion is what it is...) the claim of the necessity of having the same spectral balance for the reverberated field and the direct field is pure marketing and doesn't fit with my own experience (and my own experience is the country where I want to live)

Also because the Le Cléac'h horn delivers a very flat frequency response on axis, many audiophiles inclined to use a single Le Cléac'h horn to cover both the medium and the high frequency ranges. But if you need a larger "sweet spot" in the high frequency range when listening to music then the solution is simple: add a tweeter.

Lynn Olson found an argument I am very jaelous to have not found myself before as I am totally in accordance with (as it fits perfectly with my experience). The Quad ESL57 which is still one of the most musical reproducer possess a pretty high directivity (which one increases when frequency rises). But, when we listen to a pair of ESL57 in a normal room we don't feel anything wrong with a theorical tonal disequilibrium between the reverberated field and the direct field.

Please read Lynn's excellent post:

http://www.diyaudio.com/forums/showt...39#post1787639

Best regards from Paris, France

Jean-Michel Le Cléac'h

[bear]

Originally posted by Jmmlc
Hello Bear,

From what you wrote I can see that there something that probably I didn't clearly explained.

In fact the Le Cléac'h horn doesn't possess any lips! I mean there is no lips "added" to the mouth as it is necessary to do in a waveguide or a conical horn in order to reduce diffraction at mouth.

What you and other call lips are in the case of the Le CLéac'h horn is purely the result of the natural combination of:
1) the expansion law of the area of the wavefronts (e.g.: hypex, exponential)
2) the necessary parallelism (= equistance) of the curved wavefronts (defined as equiphase surfaces).

When calculating a Le Cléac'h horn, there is nothing special in my spreadsheets or software related to the curvature of the mouth. The curved back mouth is not intended nor provoked, it is the result of the "design by nature" of the horn.

Now on the point of baffling against rolled back mouth. If we stop the profile of a Le Cléac'h horn when the opening angle becomes 180 degrees and if from this point we decide to add an infinite baffle, we introduce a perturbation of the expansion law. This perturbation induces a partial reflection of the wave energy back in the horn and have effects on the acoustic impedance curve, and therefore on both the response curve and the group delay curve and on the pulse response (waterfall, CSD...). Those effects will be more efficient in the low frequency range of the response of the horn.

That's why I prefer to use unbaffled Le Cléac'h horns the mouth of which open at 360° minimum (Hornresp allows a max profile tangential angle of 180° so an opening angle of 360°).

You may see in attached file a theorical graph of the wavefronts I calculated. Elsewhere in the thread you can compare to the simulated pressure field at low frequency.

Best regards from Paris, France

Jean-Michel Le Cléac'h

Starting with the last first.
It appears from the diagram that the mouth terminates in a spiral, effectively going more than 360 degrees around.

I was inquiring IF when measured that there could possibly be a reflection back into the horn or out to the world from the edge or termination of the mouth that is 180 degrees or greater?? It would intuitively appear that such energy has already "turned" and is heading away, never to be reflected or diffracted in a direction that would effect the wavefront again??

You are saying that the effects are more pronounced as the wavelength increases WRT the dimension of the mouth's curvature?

You seemed to show a curve of another horn design that was similar in expansion (earlier post) but apparently had a very large mouth curvature to the rear of the horn. What is your opinion of this sort of approach?

Personally I am not terribly concerned about the "beaming" aspect for my room (longer than wide) where it is a benefit not to have significant HF energy sent wide or high.

My interest is in terms of a "coherent" subjective sound, where the apparent source of all energy is from the same point - yielding a properly presented stereo image.

The idea of a broader HF - or the idea of the HF being essentially of equal spread vs. frequency has some merit in my mind, but may or may not be desireable or beneficial depending on other factors in the way it actually works and sounds. I don't know if that works in the real world or not (not yet anyhow).

Btw, the term "CD" is an oxymoron to me: Contant Dispersion or Constant Directivity? Bad terminology, imho. Same stupid abbreviation two different things. Drives me nuts!!

Fwiw, horns like the Mantaray (Altec's version) fail in the nearfield because the acoustic take off point for the wave varies in distance along the length of the horn with frequency - and it is very very apparent. Sounds just plain bad. The diffraction throat is a problem as well - although one would think that converting the wave front to the equivalent of a "ribbon" shape would actually be a benefit, it seems not.

Another question, J-M, it seems that you have recommended using your expansion starting at or above an octave up from the mouth freq? What does the "native" response (no xover) look like from the horn run wide open? Is there a problem running closer to the low freq cutoff?

Put this another way, if I want to run down to 250-300Hz (assume the driver will go there), do I need to design to <125-150Hz in order to make that happen?

Regards,

_-_-bear

[angeloitacare]

hi Jean Michel

i guess you've overseen my last post, nr.113.

what i'm most like to know, i've asked there :

i am curious to know your listening impressions of the setup, you will be listening this weekend.

( with big LeCleac'h horn and Radian 950pb driver )

I've used in the past a tractrix horn with 23" , and a BMS 4592nd driver, covering from 300hz, all the way up. What most disturbed me, was beaming in the treble, and some sort of sound , which bothered me, which might be HOM's, but i'm not shure about this.

With the actual configuration, LeCleac'h/S2, i observe a much improved 3D imaging. And with the separate horn tweeter and wider dispertion, sound opens up - no complains anymore....

I had last weekend a friend of mine here, which uses to listen live classical concerts, and he brought a classic cd with orchestral music, he listens live. He was impressed how close my system sounded , compared to the live event.

Angelo

[soongsc]

Originally posted by Jmmlc
Hello,

According to Earl Geddes the production of HOMs is related to the derivative of the profile of the waveguide (or horn) and more precisely to the second derivative.

You'll find in attached file compared curves for the OS waveguide and a Le Cléac'h horn showing along the axial distance, the evolution of the first and second derivative of the slope angle of the profile versus the curvilinear distance to throat.

As we can see both 1st and 2nd derivative have greater absolute maximum values near the throat with the OS waveguide when for the Le Cléac'h it is at the mouth.

For what its worth... ;-)

Best regards from Paris, France

Jean-Michel Le Cléac'h

Hello Jean-Michel ,
I think you did not consider that OS wave guides have a lip radius that blends with the baffle. This would change things a bit, wouldn't it?