Hello, Ulf.

Interesting. The TAD1201H is 12-incher with resonant frequency of 50Hz.  It has relatively low resonant frequency for 100Hz Tractrix midbass but you most like used 10” thought and consequentially shorter horn. The shorter horns beam less. I tried the phase plugs in my 100Hz Tractrix midbass (4” throat, 8” driver, and 85Hz free air resonance). It defiantly extended unnecessary in my case HF, but I did not detect any changes in the beamnees...Probably the reason was because I was trying to kill HF in my horn as much as I could but you are trying to use HF. Well, it is all depends of what you want to get out of your horn…

In any case, your experiments with phase plugs for upperbass should be freakishly interesting for the flaks who try to push their horns high. The upperbasses have relatively large size and shape of the plugs, much different then we expect from the conventional plugs of the MF drivers. Do you have any pictures or sketches of your various plugs along with some measurable data? If you upload them then I will with great pleasure will map them to the site’s Knowledge Tree as it might be very educational. Particularly I will be queries to learn how the size and the shape of the phase’s “bulge” affect the frequency response.  When I made my experiments with upperbass plugs I was heavily (and perhaps mistakably) invested into the “water drop” shape.

Interesting, Ulf.

If your horn runs up-to 500-700Hz (somewhere like mine) then what kind “beaming” you meant? Frankly speaking I did not detect any beaming at those frequencies, quite in contrary – it is very difficult to make anything to beam at under 500Hz (unless you have a very large room). Perhaps we are talking about the different things? In term of sound – could describe what you feel is going on with sound when you feel the beaming kicks in? Is it a change of radiation pattern of actually the beaming?

Regarding the pictures – it is very simple and you can inject them into your post, hosting them on my server. On my home pate, on the left menu should be an entry: Announcements. It will have an article describing how to inset images into posts.

A mix of conctrete and steel skeleton is a very good mixture, particularly for the upperbass horn. BTW, go not be “disappointed” that you did not go for Tractrix. The Tractrix has many advantages but for the range where you operate I hardly see any advantages (besides the fact that Tractrix is shorter) that would be important. Also, the finishing of this thing is important. The smooth, shiny and sexy is not necessary better...

As I understood, since you said that it was “not a measured response” you did not actually experienced the sonic benefit of the upperbass horn phase plug but you rather projected them theoretically? If so, then I might propose that when you actually listen the plugged horns then you “might” not find the upperbass phase plugs as a worthy idea. Well, the phase might or might not do something for widening the horn’s radiation - I feel it is irrelevant. What I believe is relevant is that a phase plug greatly extend HF response and it is NOT desirable for this type of the horns, not to mention that your driver has too heavy diaphragm to handle MF at more or less acceptable level quality. Perhaps I’m wrong and it all will be depending from the quality of the MF driver that you will me mating the upperbass horns with …

BTW, how many times did you accidentally hit those horn’s corners with your hips?

ulf wrote:

A few years ago we discovered the tractrix horn for MF use. It was really the most important improvement so far. No more honking exponential horn sound. The improvement was so clear that A/B testing was not needed.

For whatever reasons tractrix rules only MF horns. For upperbass horns, I feel, tractrix it is not necessarily the only way to go, although my personal upperbass horn is religiously tractrix – call me a hippocratic Cat. It is important to understand that it might not be a define decision in the exponential vs. tractrix debate in case of a horn that works under 500Hz.  It might not be a final decision even in case if a person made a few different horns of different profile for the same installations.  The upperbass horn’ performances vary from so many different factors that it is very difficult to say where the specific sonic attributes came from a profile of the horn or from many other reasons (including the room coupling).

ulf wrote:

My friend made a pair oval 350Hz horns of concrete and I made a pair of round 240Hz horns of MDF. We compared these horns and were not able to detect any audible problems with the oval horn.

Interesting. An oval horn would have less the off-axes imaging skewing of a rectangular horn but at the same time it will have mode gain form the boundary coupling. I do not see any benefits of MF oval, since MF horn hardly ever sit near floor but for an upperbass I think oval should be very good idea.

ulf wrote:

We settled on a design that would be close to tractrix, use phase plug and back cavity and the selected driver was the TAD 1201H. The 1201 is really suitable for midbass with its low Q, superlow mms and high Fs and BL. In the listening test it sounded cleaner than the other drivers we tested. Although I'm not a fan of composite diaphragms, I really liked the 1201 in this application.

I never tried this driver. BTW, if your throat allows and if you have some free time then fry my Fane Studio 8M driver. It would be interesting to learn what you think about it’s sound.

ulf wrote:

Making the mould was a labour intense task!! Pouring the concrete and removing the moulds was nice and really exiting. You just have to be careful! The horn weighs about 300kg and the mould maybe 100 so watch your toes!

… and you did a phenomenal job. The way in wich you’ve done it is exactly how it should be done.

ulf wrote:

The preliminary results of the horn without the phase plug was rather disappointing, but once the plug was in place we were quite happy with the result.

Hm, interesting. Can you expend on it? What exactly “was rather disappointing”? What changed subjectively with Sound when you applied the plug? I would NOT expect that a phase plug for a horn that woks under 500Hz might change Sound from “rather disappointing” to “acceptable”.  Would it be possible that the plug changed the throat resistance and “accidentally” changed the dumping of the driver’s cone? I need to ask” when you was tuning the size of your back chamber then you did it with the phase plug installed? One more thing I would like to dump on the table” the reactance of this specific driver to the plug induced throat resistance. Usually the upperbass drivers in horn disregard the presents of front chamber and juts “shot through”. If what you described was so then is anything in the TAD 1201 driver the makes it overly sensitive? I would particularly pay attention to the way in witch the dust cap is mounted. The dust cap generally is a big bitch…. Anyhow, I still would like to proposed that the diagram that you posted above might be true for higher frequencies but not for sub 500Hz. This why I asked you about your subjective sonic observations with and without the plug….

ulf wrote:

Romy, yes the horn will transfer more HF POWER. But the phase plug helps distribute the HF so the level at the sweet spot may actually be lower (I think).

I am sure that the upper bass horn with and without phase plug should have different low pass crossover. I would very much expect that if you bring via crossover the horn response with and without then plug to a common denominator then you might not see the “sweet spot” difference.

ulf wrote:

A comment about the directivity pattern that I included: Although it is not measured you easily can hear it when moving around your head close to the horn moth. With this phase plug the response is more uniform compared to the same horn without the plug. Without the plug the HF get more attenuated when you move from the middle of the mouth.

Ulf, looking what you said I might presume that something is not kosher in your installation in the region of upperbass horn and MF horn integration. Regardless the phase plug is installed or not you must not be able to hear any HF attenuation FROM UPPERBASS CHANNEL when you move off axis that would be different from MF CHANNEL. The HF attenuation of course will take place but with upperbass and MF working TOGETHER should keep this “togetherness” off axis…

ulf wrote:

Throughout this process we have had many "nerd meetings" when a group of people of similar interests have gathered and listened and compared different solutions. Even though not all have been Horn Fundamentalists we were able to reach consensus on all the issues.

Well, I do not question that the “group of people” heard change of sound. However, as I said above: by just sticking a plug in/out DOES NOT demonstrate the effect of the plug. When you apply the plug you need to change crossover and size of your back chamber in order to observe ONLY the contribution of the plug. Make simple experiments by measuring the changes in response of the horn and driver’s impedance when you apply the plug. I presume is I what it was what the “group of people” were hearing what you played with phase plugs. Though I might be mistaken of course but I have my reasons as this point do not ”buy” the benefits of phase plug for “under 500Hz”….

ulf wrote:

All of us share an interest for music and technology. Frequent visits to live performances are most important. Any visit to the local High-End dealers are only beneficial for getting a good laugh and it makes you wonder how people can spend so much money on crappy (but usually good looking) equipment.

Dear Romy,

OK, some more about the midbass.

Well, it was not really a surprise to us that the horn without the phase plug was dissapointing. Without the plug we had a horn that was a bit too short, had too large front cavity and less compression than we desired.

So it is nor really fair to say that the plug always is better, but in our design the plug is crucial since it is a part of the horn and helps connetcing a rather large driver to a small throat.

The goal was not to investigate about plugs but to make the horn. It would require a good deal of work to extend the horn to the same length as with the plug and reducing the front cavity to the same volume which would be needed if we should really test the plug. I do not think I will do this since I am happy with the result.

Under 500 Hz? With slow rolloff filters we still have some energy rather high upp in the MF range. We wanted to the directivity in the upper region of the midbass be close to what we have in the lower region of the MF. I think that the phase plug could be a way to achieve this.

As an example take a typical High End system with a crappy 8" woofer and a crappy dome tweeter crossed at 2-3k. The woofer will clearly beam at 2k and the tweeter will have a very wide distribution. The crossover from the woofer to the tweeter is easy to hear and can not be considered seamless. Of course we want something better than this.

I think that directivity in the crossover region matters quite a bit if you want to make it sound seamless.

On an other more disturbing matter: How can we use MF up to the region of 10k with a driver that has a 200Hz phase plug?? This will give a large amount of distorsion!

One more worrying issue: How much overlap should we have in the crossover region. I made a 1100Hz tractrix for 1" drivers and tried it as tweeter HP at 9k. It was easy to hear that it had better bottom than the normal tweeter. So if we want to use MF from 500Hz would a 250Hz horn be enough?? Or should we have to use a 100Hz. This issue gets even more disturbing when we consider the midbass!

I have not tried the Fane 8M. On paper it does not look that attractive, but it still might work well as you said. Quite heavy diaphragm though.

I tried some Fane tweeters which I found to be aggresive and metallic. Which reminds me....tweeters. Still a problem!

Best Regads: Ulf

ulf wrote:

Well, it was not really a surprise to us that the horn without the phase plug was dissapointing. Without the plug we had a horn that was a bit too short, had too large front cavity and less compression than we desired.

It is hard to say. I find that it is never possible, at least in my experience, to say what compression would be desirerable. The horns programs and projected compression rate will indicate only pressure level but never indication about quality Sound. I uses never target any compression objectives and always try to make the thing to sound “correct” and then whatever compression it lead to I let it be. As long I have compression as a desire I always ended up with inferior sonic result.

ulf wrote:

So it is nor really fair to say that the plug always is better, but in our design the plug is crucial since it is a part of the horn and helps connetcing a rather large driver to a small throat.

This is why the start pussy is using a sampler driver… :-)

ulf wrote:

Under 500 Hz? With slow rolloff filters we still have some energy rather high upp in the MF range. We wanted to the directivity in the upper region of the midbass be close to what we have in the lower region of the MF. I think that the phase plug could be a way to achieve this.

OK, let forget about the directivity as I fie it is invented problem. Mostly likely you MF horn used a first order crossover and most likely it is crossed hear one octave form the horn rate. If so then when MF driver at it’s minus 6dB then the roll-off from the horn’s bell kicks in. At near 12dB the roll-off from the driver come to the picture and the decay goes even sharper.  So, a typical MF horn dies with second to thong order. Contrary to it an upperbass horn with plug and first order has a loooooong tail all the way up to 4K-5K. Usually the tail has quite a few dBs down and it is good integration tool (the MF and Upperbass works as alone array in this case). I was using this method for years but then I stopped because the “quality” of this tail was not as good as the lower knee of my MF diver.  There is another problem with the long Upperbass tails: it is very default to manage it. The Upperbass driver has inductance that “talks” with inductance of the coil that used for the Upperbass low pass filter. In real world it forces to use coil dot 3-4 times more value then what would be predicted by calculations. Measuring response of Upperbass horn drivers in real time I never was able to get the correct roll off and it always forced me to go for very large coils. (I use air coils with 8 ga wire… so you might imagine how big it was for 6-8mH). Still, the larger coil I went the more it did not sound correct to me. The only solution that worked for me was decoupling the filter’s coil from the Upperbass driver coil – I did it by implementing a crossover inside my dedicated upperbass amplifier. The deprecated toady version is here:

http://www.goodsoundclub.com/TreeItem.aspx?PostID=2762

ulf wrote:

On an other more disturbing matter: How can we use MF up to the region of 10k with a driver that has a 200Hz phase plug?? This will give a large amount of distorsion!

I do not believe in such drivers. A horn channel by the nature of topology must not be a wide range channel.

One more worrying issue: How much overlap should we have in the crossover region. I made a 1100Hz tractrix for 1" drivers and tried it as tweeter HP at 9k. It was easy to hear that it had better bottom than the normal tweeter. So if we want to use MF from 500Hz would a 250Hz horn be enough?? Or should we have to use a 100Hz.

It all depends what driver does , what horn does, what crossover does, where the horn sits in the room, not it positioned relive to other channels and from many other factors. Usually 1 octave between crossover point and horn rate is enough but in some cased it might be more or less. It is very hard to say generally. Look for instance my frustration with my lower MF channel in the “Adding one more spherical to Macondo.” thread:

http://www.goodsoundclub.com/TreeItem.aspx?PostID=2433

I think it should be asses in each specific situation differently….

ulf wrote:

This issue gets even more disturbing when we consider the midbass!

Exactly! It imposable to waste one octave of bottom horn opening juts for crossover compliance. With midbass it usably works differently, although I purely undusted that it is compromise.

ulf wrote:

I have not tried the Fane 8M. On paper it does not look that attractive, but it still might work well as you said. Quite heavy diaphragm though. I tried some Fane tweeters which I found to be aggresive and metallic. Which reminds me....tweeters. Still a problem!

Yes, I know that in the paper Fane 8M does not look impressive thought I do like very little excision, 1.6T in gap and 103dB sensitivity. The T/S parameters I find are irrelevant for horn-loading. The good sound of Fane 8M is mystery to me as well. I have the Studio 5M and Studio 10M driver the sound nothing like the Studio8M. The Studio 8M is some kind of very strange mutant, thanks God! BTW, I also used Fane tweeters and also find them horrible, however what whatever reasons I did head them perfume remarkably good in one (just one) Bruce Elgar’s installation… So, go figure…

Rgs,
Romy the Cat

PS: BTW, Ulf, here is an idea for you how you can minimize the side of your front chamber and get more compression but without extending the unnecessary HF with a phase plug.

ulf wrote:

One advantage of the use of a phase plug in a midbass horn is that it will beam less at higher frequenzies than a horn without phase plug. If the phase plug is designed for the specific horn and driver it will help "steer" the HF. Me and a friend designed (and built) a 100Hz Tractrix midbass horn for a TAD1201H driver and we used a phase plug that was about 15cm long. It followed the tractrix curve all the way to the diaphragm. The basic idea was to get a horn that would have a wide distribution at it's upper region. Practical tests have shown that it works rather well. Compared to a similar horn without a phaseplug it beams much less. Ulf

Your midbass horn philosophy sounds very interesting (I’d like to build one having a 80/100-800/1000 Hz range).

I agree -as it was said in "midbass horn: the problems"- that seeking a speaker/horn/room combination response without significant irregularities in the two octaves beyond the intended crossover with the midrange horn is a necessary design constraint, allowing use of 1st order crossover.

As I understand it and in order to achieve this, one would look for a speaker having among other characteristics a high efficiency-bandwidth product, a limited voice coil inductance, a diaphragm sufficiently rigid and airtight to withstand a moderate to medium (3 to 5) compression ratio and a smooth response up to 4-5 kHz. I selected a few 8 to 12 inches models sharing such characteristics: Electro Voice Dl 10X, B & C (10 MD 26, 8PE 21), Beyma (102 Nd), Mc Cauley 2326, PHL Audio 3860, RCF L10/750 YK, Precision Devices (PD 107, PDN.10 MH 25), Supravox Alnico 285-2000 (interesting but Fs may be a bit low for this application), TAD TM 1201 H.

But as you mentioned, this won’t be sufficient to predict accurately the sound of the driver once loaded by the horn. The Fane studio 8M you mentioned is out of production and old models not easy to find here.

Ulf, I would really appreciate to know which drivers you tested before selecting the TAD 1201 H (real tests? with different or identical compression ratios?). May you please detail your findings and listening impressions, as it might be very educative? May you also describe more accurately your rectangular tractrix midbass horns: flare rate, length, mouth and throat cross section area, mouth width and height, measured response you mentioned? As well as the integration of the midbass with the rest of the system?

I thought it may be interesting to load the back of the driver with a progressively damped (stuffed) quarter or even half-wave pipe, tuned to (say) 30 or 40 Hz and having cross section area equal to throat area, having enough place to play with it. This may be another way to cancel throat reactance at Fc; it would also provide a resistive load down in the LF region and may allow use of interesting drivers having in other respects limited Xmax and higher Fs, such as the Mc Cauley 2326. I would appreciate your comments about these suppositions.

Thanks for your answers,

Luc,

I am not ‘lucfm’ and I hope he will reply to you but I would like to pass some of my observation that you might find useful.

In my past when I discovered Fane Studio 8M I went over a number of 5-8 inches. At that time I used AG’s Trio upperbass horn and I was looking for something more interesting – the default AG’s bass driver was quite poor. I had very specific requirements to the driver and tried quite a few of them, predominately vintage drivers, as the today’s driver did not necessary comply with my prejudges about “proper cone structure” and “proper suspension”, not to mention their low sensitively and their aiming for the port-like exertion….

I hardly see how T/S characteristics affect the actual sound of a driver loaded in horn. People do modeling, circulations, spared to each other “smart” theories about the driver-enclosure interaction. I know all of it but I care less about it as I feel it in one way or other work for boxes but as soon you load the damn driver into a horn, them all the rules are off and only God knows how it will behave.

I mean - the behavior of the driver is still predictable IN TERM OF PRESSURE IT WILL DEVELOP but nothing know in term HOW it will Sound as soon the horn loads the driver. There are so many none algorithmable variables in horns, drivers and rooms that from my point of view as soon a diver hits belong 500Hz all calculation or theories should be tossed away and the channel should be actually heard and assessed with proper (objective) listening evaluation teachings.

So, why the upperbass driver in a horn does not comply with any theories we know. Now let go to the ugly part: I feel that the answer is because all our upper bass horns are eventually a half-ass crap by the nature of our design because we are greedy dudes….

Make an experiment, and I will simplify the case quite assertively. Take a typical compression driver, cross it at 800Hz, second order and load it into a proper contemporary horn (Tratrix or JMLC) of 300Hz. Listen that horn, you will get some sort of sound that let accept as OK Sound. Now begin to very slow lower the crossover point and keep listening the channel. While you lowering the crossover point, somewhere around 550Hz (I took this number purely arbitrarily as it would depends from VERY MANY circumstances) the horn will begin to demonstrate what I call “choked sound”. The “choked sound” is HOW HORNS SOUND IN 99% OF ALL HORN INSTALLATIONS OUT THERE – people just too damn to deal with it. The “choked sound” is the satiation when Sound produced by a driver can’t be “processed” by horn. In this “choked mode” a horn produces the “sonic boom” that was made by the horn’s bell and that “sonic boom” screw up the enter band-pass of the channel - the game is over.  Increasing of the crossover point for ¼ octave (for instance) will fix the situation - so we have approximately one octave between horn’s rate and mix crossover point…

Now, apply the very same excrement to the upperbass horn. You spent a half of your vocation, pile of money, huge amount  of useful livable space, gallons of blood and weeks of arguing with your homemates, your played the real-estate acrobatics to ingrate all installation and eventually you defeated everything and made your damn-proper, upperbass horn. You did not cheat and your made it with a proper small throat, perfect driver and a full-profile opening. Not you need to decide where to cross it. If to do it PROPERLY then you need to cross it at ~180Hz. You look at mirror and ask yourself: “I am out of my mind to make 600 pound 50-inch horn and to use it ONLY down to 180Hz!!!!????!!!??!?!?!?!?” Well, we all were there and we all know the answer:  no, you will cross it much lower to get the “horn bass” under 100Hz. If you very smart then you even build a back chamber to get come bass boost hear the horn’s mouth rate. You are not a fool – you can recognize the signature of the “choking sound of the sonic boom” in your horn that I described above but you juts close your eyes to it and declare that it should be this way….

I can go on and on but you get the picture – our uperbass horns are conventionally are crap (I do not even mention the bass horns) and when we talk about the sound of one or other better drivers loaded upperbass horn we are talking about the sound of a given driver under the barbarian condition of the “choked sound”

Well, the described effect is not as horrible as I would like it to portray but the “choking sound” of out badly made horn installations should not be taken out of consideration.

Dear Malinowski,

Here are the input data used in the HornResp simulation for the tractrix midbass.

This is a 117cm long tractrix horn with the TD-1201. Throat area 84 cm2, mouth 10000cm2 with 5,6 liters back cavity.

Ang: 1Pi; FTA 89,80; S1 84; S2 10000; Tra 117,64; F12 97,04;
Sd 531; CMS 1,57E-4; MMD 53,63; RE 11; BL 36,8; RMS 13,63; LE 1,38; EG 4
VRC 5,6; FR 40; VTC 531;
LRC 5; TAL 2; ATC 531;

If you have no experience of HornResp I suggest that you download a copy and try it out. It is a very neat software that is useful when designing horn speakers. You can easily make simulations and get a decent result concerning frequenzy response and so on.

About how you should design your midbass?? Hard to say really. It all depends. But I would at least recommend that whatever design you choose you should make a full size horn.  Tractrix, Lecleach and Klangfilms kugelwellen profile all share a similar mouth termination. They all exit at  90 degrees (or more) from the axis. This seems to be the main feature that makes them different from hypex and other common types. I have only experience of the tractrix but I Kugelwellen and Lecleach could be as good or even better.

Lecleach has based his designs  on a new calculation method which seems to make sense. Many of the old designs semms to be based on just trying a curve and then see how it works or on methods that were manageble to calculate at the time.

We have tried a few oval and rectaungular tractrix horns. Yes, you're right, they are not real tractrix but sound fine anyway.

So I think the mouth termination has major impact on the sound quality. But the best way to get there from the throat is still a mystery to me.

Yes the midbass horns are still in use and sounds Grand!


Best Regards: ULF

Reading how you plan to go about making this horn, I offer the following (this first one you probably aready know):

Make the outside of the mold in quarters or thirds and hold it together using fabric straps (not bolted flanges)... If you make it in halves the mold will split due to expansion of the casting material when the chemical reaction causes it to heat up... The straps will flex just enough to allow expansion.

Also, make the rear chamber plug a separate part from the main "volcano" form... This will make the whole thing a lot easier to pop out of the mold, as you will not be forced to unmold a multi-hundred-pound part that will bind due to the double draft conditions (tested!!!!). It will also allow you to fit different diameter rear chamber plugs later on.

Also, when using a template to form the plaster, the key is to load the plaster on and turn or cut it while it is still very very  liquid... Just keep splashing it on and turning (really a two-man job). Obviously, with a horn this big you will want to rotate the template, and leave the assembly stationary... The challenge will be to make sure the rotation template is RIGIDLY mounted at both top and bottom, and is held in a RIGID frame.

I use polyester resin to seal the plaster surface, then floor wax (the real stuff... the kind intended for hardwood floors) as a mold release... it is way cheaper than buying "Mold Release Wax"... I used to complete the process with a coat of PVA (poly-vinyl-acetate, a commercial mold release agent), but this is just a waste of money... 2 or 3 heavy coats of floor wax does the trick.

Get on a high-carbohydrate diet one month before starting work!

I look forward to seeing the images!

jd*

jd wrote :

"...Obviously, with a horn this big you will want to rotate the template, and leave the assembly stationary... The challenge will be to make sure the rotation template is RIGIDLY mounted at both top and bottom, and is held in a RIGID frame..."

On second thought, because the horn must sit on something while you turn it, I can't see a simple way to rigidly locate the bottom of the template while still allowing it to freely rotate about the horns central axis... Probably better to bolt the template to a corner of your house and get three pre-carbo-loaded friends to help rotate the horn.

Alternatively you could do what I did... I rotated the master for my upper bass horn in halves, with the horn laying horizontally... Because the template did not have to rotate more than 180°, this allowed simple fixation of the template at both ends, to the central axis (the other reason for doing it in halves is getting it through the door).

jd*

ulf wrote:

Let me tell you a bit about how an why we designed the midbass horns the way we did. A few years ago we discovered the tractrix horn for MF use. It was really the most important improvement so far. No more honking exponential horn sound. The improvement was so clear that A/B testing was not needed.  My friend made a pair oval 350Hz horns of concrete and I made a pair of round 240Hz horns of MDF. We compared these horns and were not able to detect any audible problems with the oval horn. So we figured that the Tractrix curve may not be so important. Both horns have the flat termination of the tractrix and this seems to be the most important issue that differs from exponential horns. After a while he wanted to upgrade his 90Hz Hypex midbass horns and we had long discussions about the design. We made some tests with the existing 90Hz where we tested some different drivers, phase plug or not and back cavity. We settled on a design that would be close to tractrix, use phase plug and back cavity and the selected driver was the TAD 1201H. The 1201 is really suitable for midbass with its low Q, superlow mms and high Fs and BL. In the listening test it sounded cleaner than the other drivers we tested. Although I'm not a fan of composite diaphragms, I really liked the 1201 in this application. Given the size of the horn we wanted the mouth to rest on the floor. A true tractrix is designed to work in free space and a horn of this size placed in free air would really take a lot of space. So we used a modified tractrix that would better mate to the floor. On 3 of the lips the tractrix is clearly visible, the lip that rests on the floor is a bit flatter. After many simulations in Hornresp (thanks David) we had a design that we were satisfied with. Still we were not quite shure about how we should design the phase plug so we decided to not include it in the concrete horn. The plug and the first (about  15 cm) segment is bolted on the concrete horn. Making the mould was a labour intense task!! Pouring the concrete and removing the moulds was nice and really exiting. You just have to be careful! The horn weighs about 300kg and the mould maybe 100 so watch your toes! The preliminary results of the horn without the phase plug was rather disappointing, but once the plug was in place we were quite happy with the result. Compared to the old 90Hz hypex we got a cleaner, more controlled sound. Over all we only could find positive changes. The most amazing thing though is the soundstage. It really sounds large. A surprising amount of stereo information were revealed that we didn't expect. It mates really well with the MF horn and it is virtually impossible to hear the crossover. A measured response showed that we had some more output in the upper region than we wanted. Up at 800 it is about 3dB too much and this remains a problem. So there is still room for improvement. Romy, yes the horn will transfer more HF POWER. But the phase plug helps distribute the HF so the level at the sweet spot may actually be lower (I think). A comment about the directivity pattern that I included: Although it is not measured you easily can hear it when moving around your head close to the horn moth. With this phase plug the response is more uniform compared to the same horn without the plug. Without the plug the HF get more attenuated when you move from the middle of the mouth. An image of the plug. You can see how the HF from the left side of the diaphragm will beam to the right side of the horn. The plug follows the tangent of the horn side. Throughout this process we have had many "nerd meetings" when a group of people of similar interests have gathered and listened and compared different solutions. Even though not all have been Horn Fundamentalists we were able to reach consensus on all the issues. All of us share an interest for music and technology. Frequent visits to live performances are most important. Any visit to the local High-End dealers are only beneficial for getting a good laugh and it makes you wonder how people can spend so much money on crappy (but usually good looking) equipment. Best Regards: Ulf