Thanks Romy and Paul for your comments.

Romy wrote : "...I personally would probably go for a monopod frame above the upper bass horn (merging it in middle) as I feel it make the architecture more elegant, more airy and visually less demanding..."

Yes, perhaps a frame with a single mast would have been more orderly. It certainly would have been possible. My logic in going with parallel masts is as follows:

Option 1 : Single mast - Frame passes up under the 180Hz horn.
In this case, the horizontal member of the support would have eaten into valuable vertical space for the tweeter.

Option 2 : Single mast - Frame passing behind the driver and up over the 180Hz horn.
I now know what the 180Hz horn does (and really love it), but at the time I did not, and wanted to leave space for both deeper and more shallow horns; I did not want for example, the forward end of a horizontal member dictating horn geometry or placement (yes, there are ways around this problem) and by extension placement of all other horns of that channel.

Of the two options, this is the better choice. However, it must also be born in mind that the 180Hz horn + S2 is nearly as deep as the upper bass horn, and weighs 135 lbs, most of which are concentrated toward the big end of the horn, and that passing a central mast behind the S2 (leaving space for driver removal and time arriving adjustment) would mean holding all the horns weight in pure cantilever.

To summarize, because it is less restrictive with regard to horn placement, and because it is less solicited in terms of load (cantilever) I elected to go with a parallel mast configuration.

If however I were to put these things into production, I would nail down the variables and reconsider using a single mast, as the advantages it represents in simplicity of fabrication alone are enough to justify such an approach.

"...Horn installations are ugly from female perspective..."

Yes, and this type of woman is, from my perspective, as interesting as your average cabbage. However, last night I had a female visitor who commented (without being prompted) that she found the installation to be "beau"... I said "What?... Say that again please... Does this make you a lesbian?"

Anyway, the horns are not the only objects in the place which others might consider as aesthetically offensive elements. Most likely due to professional deformation (my DAY JOB), I have developed an aversion to all styled objects; I won't buy them, and can't stand living with them. By contrast, I find horn installations to be visually fascinating, because like weapons, they are so expressive of their function.

Paul S wrote :

"...So, you used to be:

tool and die guy

mold maker

brain surgeon..."

My background is in automotive styling (cringe!). Plaster working skills were learned by watching the French and Italians make automotive prototypes... I was lucky enough to witness the last of their breed in action... The best were veritable artists... We now do everything via CAD and CNC milling. Along with waiting tables, motorcycle mechanics and metal fabrication skills were all acquired in past lives.

About fifteen years ago, as a remedy to the effects of said DAY JOB, I undertook an ambitious personal project (its finally done!) which called for, among other things, a basic knowledge of structural engineering and metallurgy; I this from books.

BTW, one of my favorite books ever written happens to address the subject of structural engineering (along with quite a few other subjects... Not what you'd expect).

It is : "Structures, or Why Things Don't Fall Down", by J.E. gordon.
http://www.amazon.com/Structures-Things-Dont-Fall-Down/dp/0306812835

jd*

Hmmm, are you sure that you do not do a mistake?

Jessie, I do not know how to say it but looking at your last picture something does not work for me. The hone that you made are elegant and graceful but painted in glossy black with that very rough structure I feel they are less attractive. The texture surface is good but I think that the sophistication of the epoxy-sand mix is not good enough for the appearance of your fine horns. It is of cause personal preferences but If I have your wonderfully-smooth horns I would probably keep them white and then used different shadows of gray structured paint to color them. Like upper bass horn might be dark gray, almost black and the upper it goes the lighter gray it become – of something like this depends of the light condition of your listening room. I think at this point, what you do not know where and how they will be installed it make sense to keep them white unfinished. Then what you found the place you might take the drivers off, load a big car-painting mashie and compressor with paint and spay the horns in accordance to how you wish.  I would juts draw a macro-image of the room right on the top of the installed horns. Then juts repaint the room.  You never have this change again if you finish the horns right now.

Romy wrote : "...The horns that you made are elegant and graceful but painted in glossy black with that very rough structure I feel they are less attractive. The texture surface is good but I think that the sophistication of the epoxy-sand mix is not good enough for the appearance of your fine horns..."

Actually the photo was taken while the resin was still not cured... Once cured it goes more mat. This is then followed by a coat of mat black paint (what used to be called oil-based paint as opposed to water soluble latex) applied to both inner and outer surfaces.

Why mat black? Simple, ease of retouch. At one point the upper bass horns were done in dark bronze, and just as you suggest, I was going paint the smaller horns in progressively lighter shades... In the beginning I thought they looked great, but in the end I came to hate the color and especially the idea of being forced to keep on hand a correct mix of each color in the event of a retouch.

Other than dust, I have no problem with the black.

Regarding the scale of the texture: Like an idiot, I applied this texture at the time I moved (there were reasons to get it done at the old place) so the first time I listened to such an aggressively textured horn was once setup in the new place; meaning not only a different room, and different electricity, but also, finally with the upper bass horns in the same room.

Conclusion: Without conducting a back to back test against a smooth horn (which I would have to make), it is not possible to say, with any degree of certainty, what effect this texture has on sound.

I'll be sure to put one of the small horns in my suitcase if ever traveling to Boston.

jd*

jessie.dazzle wrote:

Though the frames have wheels and roll very easily, my room is cursed with an elevated platform smack in the middle of everything (about 14 inches higher than the main floor), and to move the horns up on to or down off of this platform requires taking everything apart. Over 1600lbs of horn and frame; I would not try pushing it up or down a ramp... so, about 5 hours to dissamble, move, and reassemble once off the platform.

Yep, welcome to the club. To move this thing over a room is very troublesome, I usually if I willing to do so take the upperbass horn out of frame. The rest becomes extremely blocky but more manageable… The good/bad part is that you might not need to move anything else besides the upperbass horn… :-)

jessie.dazzle wrote:

My lower mid horn is a 180Hz tractrix, band passed from 800 - 3200Hz, its output is not attenuated. Romy calls this the "Fundamentals Channel"; its contribution is truly fundamental, to the point that if I had to reduce my system down to a single horn system and could keep only one horn, this would be it (maybe I can start a movement!). The horn sits up on top of the stack, which when seated in the near field (less than 10 feet) means that you do notice sound coming from above... I would not say its a problem, but it is not the same as a perception where everything originates from around ear-level.  Anyway, you can probably guess what's coming... With the horns set up on the main level, I decided to put the couch up on the platform, which puts one's ears about in line with the tweeter. Results... Good idea!

OK, Jessie, not you enter the realm of fine tuning of your system. Your solution with platform is elegant but it cures symptoms not the courses. I personally do not feel that you use Fundamentals Channel properly. You desired to run your Fundamentals Channel up to 3200Hz but it is too high. Take a look, you have your MF channel I presumed crossed at 3200Hz with 3uF cap, at least it is what I have. However it is juts electrical crossover point. Do not forget that the secondary resonance of S2 driver pushes a large boost at 1250Hz making the S2 to response flat down to 1000Hz. Your Fundamentals Channel does not die at 1000Hz but run all the way to 3200Hz. The 3200Hz with first order is a nice and opened MF driver with a lot of HF extension in it. The HF extension makes the channel to be too identifiable height-wise. Also, the HF extension from Fundamentals Channel crates some lobbing problem in regards to MF channel. So, what you need to do is to found a right balance between the Fundamentals Channel and MF Channel.

When you said that your Fundamentals Channel is the single horn that you would keep I imidetaly sensed a problems. The in my few the Fundamentals Channel should not push recognizable Sound but rather the unpleasant boom noise. Portend a hypothetic guitar. The sounds of strings shell come from MF channel. The sound of guitar’s deck shall come from Fundamentals Channel. The sound of the room where the guitar plays shell come from the Upperbass Channel.  In your case I feel the Fundamentals Channel work too high.

Do the following – put two channels ONLY in operation: the MF and Fundamentals Channel. Run the RTA with MF channel and found out where the MF will roll off. Set the Fundamentals Channel at the very same point and listen both of the channels, running the low-pass crossover point on the Fundamentals Channel you will soon or later found the one you feel is right. I ma conversed that it will be lower then what you have now. When I did those experiments I detected that adding the 100-200Hz to low-pass crossover point on the Fundamentals Channel is auditable. Anyhow, when you will be lowering the Fundamentals Channel low-pass crossover point you will reduces the Fundamentals Channel’s HF out and it will reduce the ability of the Fundamentals Channel to lift up the imaging of the MF channel. Generally what I feel that implementation of Fundamentals Channel shell not lift up system vertical imaging. If it still does do then you might use a trick: lowering the Fundamentals’ low-pass crossover point but increasing the Fundamentals’ amplitude (I have an attenuator that allows me to fine-tune the Fundamental’s channel output).

jessie.dazzle wrote:

(I once left my door open as he passed by; he stuck his head in and, upon seeing the horns, stuck a finger to his temple, rotated it a quarter turn while looking me in the eye and doing his best to express sincere pity)

Oh, Jessie I was laughing for 5 minutes upon reading it. The sad part that I very much agree with that guy….

jessie.dazzle wrote:

The discovery is how simple it is to recenter the listener. The other nice thing is that when listening while standing (off the platform), one's ear is maintained at the correct altitude.

If you lower the crossover point at your Fundamentals Channel then when a listener will be standing and the Fundamentals Channel will be shooting right into the listener’s ears then you will lose the neutrality of balance.

jessie.dazzle wrote:

And Ok, maybe I don't have all the filters dialed in, and the DPOLS are like a dream to me now….

And with 1600 pound per channel you might forget about the DPOLS … -

jessie.dazzle wrote:

Here is an RTA plot of both the upper and lower mid horns, as they are currently set up

OK, Jessie, first thing first and I am not kidding – change the way you use RTA. You must not look at 80dB window. You need to found for yourself a reasonable resolution and stick to it. I personally use 1dB you might go higher or lower, but you need to see your .5-1dB very deistically. Use 10dB-16dB window. Now, it is very important to found a convent RTA window and resolution and never change it. The reason is that when you begin to listen it and interpret your measurement you will develop visional-audible relation. So, in order the RTA sweep has any practical since for yourself you need to develop in your head some sensitive of looking at the RTA’s output. So, go for 12dB window and .5dB horizontal markers.

Regarding to the graph itself. Everything looks fine, some adjustments you might consider. I will make general observation encompassing your comments

1)    Your must to select for you’re a metal reference level – you must not balance the system abstractedly. The reference level in your case would be the main plot of your MF channel. I marked it Red on your TRA

2)    The MF is fine but it has quite strong bottom output down to 500Hz and rolls off too soon at HF. What distance you measured it and how much off axis the mic was? Was Mic calibrated at HF?

3)    Since you have a dedicated Fundamental Channels I would cut the bottom response of MF Channels. Go for 2.7-3uF cap making the electrical crossover point around 4K. This will bring the acoustic crossover point of MF to ~ 1000Hz and would save a lot of necessary excursion at MF driver.

4)    Bring the Fundamental Channel 3-8dB lower – I do not see at your resolution how much – until the 500-900Hz region hit the Reference Level.  Make the band-pass more narrow, decreasing the HF extension from the Fundamental Channel. It will be a coil around  2.7mH that would make 900Hz. But 3 or 4 of then, the cheapest one and then listen the MF and Fundamental Channels combined, while taking the turn out of the coil. Approximately around 2.3mH or 1000-1100Hz you might found it works better. The precession is not too important in there – you can always tine-tine the Fundamental Channel with changing volume of the Channel – a good 16R LPAD would help a lot. Do not forget that Fundamental Channels is the body of sound – you will fine adjust it later on. The proposed ideal response of the Fundamental Channels I drew in yellows on the graph. In practice it might be even 5-4dB lover then this – and not to me measurable at all (hidden behind MF and upperbass output). Sonically, however that not-measurable weight will be there and do its duty. I use disconnect the upperbass  and MF to get the  Fundamental Channel profile.

5)    Find what to cross your Fundamental Channel at the bottom. It will depend how much your horn will tolerate without being choked with honk. Do not forget that at the range of the Fundamental Channel’s bottom knee the room maters.

6)    The prospect upper bass woe reaps is show with green dashes…

jessie.dazzle wrote:

If I get you correctly, the graph should display a range of around 12dB from top to bottom, in .5dB steps... Yes?

jessie.dazzle wrote:

I noticed that you use a non-logarithmic graph

jessie.dazzle wrote:

"...Your must to select for you’re a metal reference level – you must not balance the system abstractedly. The reference level in your case would be the main plot of your MF channel. I marked it Red on your TRA..." I understand this to mean an average…

jessie.dazzle wrote:

Measurement distance was about 6 feet back from the mouths of upper-bass horns and centered between the L and R stacks of horns (this is a plot of both upper-mid horns playing together... You will probably tell me that I should have tested a single horn, which is what I usually do). The mic is supposed to be calibrated, its not very old, and it has not been abused. The HF rolling off could be due to the fact that I was not on axis, or that both horns were playing; I also suspect the tone generator... I've been using a little computer program which seems to do ok until it hits the HF range, at which point I can hear it going a bit nuts. The computer sends the signal to the DAC, the same as if it were a music file. I also have a Tektronix tone generator, but have not dared connect it to either the preamp or directly to the power amps. I guess its just a question of measuring the strength of the output signal from the tone generator to make sure I don't fry something in the audio gear (any advice here would be very appreciated).

jessie.dazzle wrote:

Up to now I have been superstitious about using anything with resistors directly in front of a driver; but I could maybe mix up a double martini and give it a try.

jessie.dazzle wrote:

Yes, make that a tripple martini if I am going below 500Hz, 1st order...

jessie.dazzle wrote:

Its still not perfect, and I had the mic 1m from the membrane of the driver (had not yet read Romy's last post where he recommends putting the mic back at listening point).

I would like to stress it again that that positioning of microphones for measurement of en entire acoustic system is a very complicated subject. In some instances the static positioning would not work and the mic needs to be moved. Anyhow, I would like to keep the subject of mic positioning out of this thread. What Jessie you do are relatively simple measurements and in your case the microphone might be at any position – even at the mouth of the horns. The measurements themselves are not important. The important is how your auditable how the experiences correlates with your auditable experiences. So, the data that you collect via measurement will be juts to interpreted differently in case you measure from your listening position vs. you measure at 1 M from a given horn. It is like if you horn axis shot in your shoulder+1’ (like in my case) then where to position microphone for HF measurement: at the location of your head or at the MF horn’s axis. The answer would be – it does not matter for what you measure. Give yours another month and you will not be caring about the absolute response anymore but you will be concerned about the very specific rates of response change. The rate of change might be observed by many means and in many microphone locations. So, pick the mic position that you feel comfortable and stick with it. It will in time give you a sense of “how the given measurement sound” when you just looking at the graph of the given channel will be able to predict sonic consequence of the given solution.

jessie.dazzle wrote:

Yes, it seems you are correct. Arriving at this conclusion alone would have probably eventually happened (well I like to think so!), but I was reluctant to try such a thing because, to be honest, I had always thought you under-used your Fundamentals horn due to the difficulties you had in implementing it. I have often wondered if you were not supplementing the output of that horn via your Injection Channel. I now see (hear) for myself that neither is the case.

I do have difficulties with Fundamentals Channel but not those that you mentioned. My desire do not let the Fundamental Channel to run high is purely deliberate. Running the Fundamentals higher reduces MF texture and vertically offset imaging. The imaging might be brought back by running the upperbass higher but then we have some imagine diffusion (lobbing) + the HF because too heavy and loose lightness. With three channels having that large overlap some music will to onto sonic saturation. It is like having a very heavy meal on thirsty mouth and be not able to drink water… Try to make this setting and listen some symphonic poems like Liszt, Strauss and etc. it will be intolerably- stuffed. I do have the difficulties with my Fundamentals Channel’s lover knee. Here I would like to have 180Hz horns, lie you have instead of my 250Hz horn, but it is what it is. The 180Hz horn would be larger than my current horn however, and I am at a very strict nearfiled environment… so, it is not clear to me if I would go for 180Hz horn in my case. Anyhow, the Injection Channel is not in the picture regarding this matter.

jessie.dazzle wrote:

I took a look at some of the sweeps you've posted for the S2 and Water Drop tweeter; it seems you tend to use a 40dB window... This I think I could manage. I find reading your RTA sweeps seem a bit coarse to me for example when trying to read HF rolloff of the S2, but this may be because you were using something like a 1/3 octave setting at the time (?)

The HF rolloff of the S2 is a tricky subject. It is 12.5kHz in the horn axis with a slight +2dB infliction at 10kHz. I do not use S2 on axis and at the location my head it is around 10kz flat. However, here is where the S2 shows itself off. The S2 has own poisoned caustic resonances at very HF that I found very interesting to use  and that subjectively make the S2’s 10kHz to sound more satisfying then TAD/JBL/EV/ALTEC’s 15kHz. The S2 at the very top has that “sonic thorn” that in my view set it apart from other drivers – I kind of got addicted to it though the “S2’s thorn” in my case (single stage amp and heavy loading) is very much reduced. So, you can’t say that it is juts 10kHz - it is rather the S2’s 10kHz - the very different animal then any other 10kHz.  

Yep, it is a lot of work and it might look like not well organized work but there is good part in all of it – when you will be do you will be done and you would never do it again. Practice shows that those custom tailored systems are made and then used for tens years.  Sure you might move to different rooms in future that would lead for some minor crossover adjustments or you might try different derivers to get desired qualities of a given channel but the bone of this type acoustic system is pretty mush fixed and has not a lot of room for any further improvements. Well, unless Mr. Le Cleach invents some kind of new horn profile that would be better but will not be compatible with anything else… OK, it was a joke…

Anyhow, I do not know at which stage your project is what I think you are somewhere in a process of integrating the drivers. Do not forget that any audio project is a wonderful opportunity to buy another set of stupid audio tools – your case is not an exception. I have two large boxes - one with coils and one with caps; the indictor and capacitance meters. They are the tools I used to search for my crossovers – I trust the calculations and the predicted response but I still always would like to have the tangible feeling what is going on. This pain might be optimized. You might get yourself Pioneer D-23 crossover. It is it a very nice and flexible 4 way unit:

http://www.thevintageknob.org/PIONEER/SERIES20/S20D23-specs.html

You might run one channel opened and then with the D-23 to observe the level of it’s none-transparency and then extrapolate the observed non-transparency in your judgment.  Do not take the digital crossovers out of consideration. The Behringer is $300 unit it means that you might rent it from you local pro shop for 10% per day. (My shop consider Saturdays to rent and Monday to return as one day). The digital crossover is also a very nice tool for prototyping. What I like about those prototyping tools - I like 2 things. First, using them you actually trying and listen instead of ordering the caps and fight with coils. Second, when you use the crossover prototyping tools you do not listening the actual sound of channel. Instead you practice the target listening – try to hear exactly what you need. I found it is self-educational.

Still, the fixed coils and caps might do – you can always parallel of serialize them… If I were you I would not think at this point about DSET.  The DSET has own complexities and before trying to assess those complexities you need to have a references how your acoustic system shell perform before it will be enriched or screwed by DSET. Your Lamm ML2.0 shall do wonderful job to drive your thing…

I would propose to spend a year or two living with a built and fully functional installation and then to see if you feel any needs for “other” amplification. If you decide that you need other amplification then it should be not the architectural decision but rather your very specific sonic demands that you flash out from a performance of you setup at that time.

I like the idea of the Poineer crossover... Regarding digital units; just looking at the face plate on the one from Behringer scares me (perfectly intuitive for someone "from the other side", which means it would take me two weeks to understand), but it is surely a very capable unit. It is no accident that I live right next to a shop selling really nice coils, and pretty decent caps; I don't mind the work of swapping components... It was just difficult to get things to move in the right direction with mic placed in the listening position.

"...when you use the crossover prototyping tools you do not listening the actual sound of channel. Instead you practice the target listening – try to hear exactly what you need. I found it is self-educational..."

I'll keep this in mind.

"...Still, the fixed coils and caps might do – you can always parallel of serialize them…"

Yes, I have been.

If I were you I would not think at this point about DSET.  The DSET has own complexities and before trying to assess those complexities you need to have a references how your acoustic system shell perform before it will be enriched or screwed by DSET. Your Lamm ML2.0 shall do wonderful job to drive your thing…

I am not for the moment thinking about DSET amps for myself, but looking at the RTA curves and trying to get what I want from the various drivers does tend to underline the advantages of this aproach. For example, killing the output from my upper-bass horn above 700Hz... Every time I stick a coil on the driver (any of the drivers I've tried in that horn), I end up pulling it back off let it go full-range.

Back in your "full-range amp/speaker-level crossover days", you wrote that a coil used with the upper-bass horn, tends to "talk" with the S2s... In my case, I don't know if this is what's happening... Need to do some target listening. I keep telling myself that I MUST use a coil with this horn... I mean, I have a lower-mid horn etc... But the system just sounds better with the upper-bass horns running full range.

In any case, again, I'd like to get the existing system to a stable state... Sort of a foundation to which I can add the mid-bass horns.

BTW, here are some photos showing one frame loaded with horns and drivers :









There is space, and the frames are strong enough to support an additional load, but for the moment, I don't imagine adding anything.

jd*

PS : I will most likely be going with a hyperbolic/exponential profile for the mid-bass horns... Anyone know of a good Hypex profile generator? Also, yes, I've sort of started to admit that rectangular horns are probably the way to go... Believe it or not, they will be more difficult and require more tooling than round horns. No matter, they offer the advantage of preserving access the kitchen.