In the passing in Netherlands European Triode Festival

http://www.triodefestival.net/index.php?page=etf-2007-photos

... there was presented horn system by Marc Henri  and Andre Klein

 (Information and image courtesy to Raymond Koonce at http://www.timebanditaudio.com/etf2007/etf2007.html)

This 3-way installation is simple, it is elegant and it is “properly done” – time aligned (look like) and made without any other typical-stupid horn mistakes. With a proposer selection of drivers, crossovering and etc this configuration is a very good endeavor into horn universe. There are some controversial moments in this design but they are NOT WRONG moments, they are rather juts controversial.

I visitor of my site shat me an email informing that that above depicted horn installation by Marc Henri and Andre Klein was made French Musique-Concrete company. I do not know a lot about those people, beside that they hand around Jean Michel Le Cléac'h. The Musique-Concrete has web site with English section:

http://www.musique-concrete.com/Eindex.htm

Do not miss their “Material selection” page:

http://www.musique-concrete.com/Emat.htm

Where they very rational and very sensibly go over the horns material and finish. The Musique-Concrete has popped up at this site: they build hors for Rony Weissman:

http://www.goodsoundclub.com/Forums/ShowPost.aspx?postID=4593#4593

Anyhow, my comment in this article is about the above depicted installation Musique-Concrete presented at European Triode Festival. Among many “issues” that I see in this system (like use of 1kW push-poll amp to drive the horns, like the way tweeter is used, and few othersthings) I would like to concentrate on one the Musique-Concrete bass solution.  As I said in my previous post what Musique-Concrete does is not a mistake but rather a disputable direction. I happens very disagree with entire Musique-Concrete bass direction and the purpose of this post is to voice my view on the subject as it is a very common false direction for many horns installations.

If your look at the page with various systems where Musique-Concrete used:

http://www.musique-concrete.com/Eexemples.htm

then you will see people employ many solution to compliment the Musique-Concrete with bass: sealed enclosures, ported, bass horns, Onkens. Musique-Concrete offers own solution in a way of bass horn (image is courtesy to Jürgen Ewert at http://ircaudio.thor.prohosting.com/etf07.htm):

I never like that type of the horns and I consider them fundamentally faulty. Let look deeper. The Musique-Concrete horn looks like 250 Hz -300Hz horn with 2” throat. Let presume that it crossed at 500-650Hz, let do it 500Hz for a love or round numbers. I personally do not feel that it is possible to have one front radiating channel that would cover from bottom to 500Hz but my lock of believe should not be impediment to others to try.  The Musique-Concrete use, it looks like, 55Hz-60Hz horn loaded with two 15” or 18” drivers, used (once again: it looks like) with open back. It looks sexy and it rocks the boat of people who “love horns” but … I feel that those horns are very poor and they are in fact are the bottle neck this Musique-Concrete installation.

In a horn the most important category is the ratio between the mouth and throat. This ration of the Musique-Concrete’s bass horn is truly pitifully. In this configuration, with such a huge throat the drivers act as a pure open baffle direct radiators and then “see” the horn as no lower than (in my estimate) 250Hz. So, the Musique-Concrete bass horns works fine (I hope they parallel the drivers) down to 250Hz, then the horn EQ stops and the drivers drive the room as direct radiators. To serve the 250Hz horn the driver should have primary resonance around 150 Hz -200Hz but with this resonance they will be worthless as direct radiators. So, people usually stick in that type of “horns” the horns drivers with resonance of 50-60 or even the 515B-type of driver with resonance or 25hz. They presume that with lower bass from the driver then will have more bass out of the horn. However, the horn can’t traduce lower bass then the horn rate. So at 55Hz-60Hz the horn want to cut the response but it cannot as the throat is stressed with excessive pressure of pounding LF drivers. So, the horn got choked with LF pressure that destroyed bass quality tremendously.  I have heard a number of those attempts – all of them sound horrendously bad, screwing with barbaric upperbass pressure the much higher frequency region. (It is what some of the cretins that I know call the “reputed vintage sound of moves theaters form 30s”)

Hi Romy and all,

first of all, my name is Henry, with a "Y". due to an Irish grand-grand-grand father called "O'Henry". the O' was deleted when my grand-grand father was born, due to a silly servant of registry office. :-) all my horns uses the calculation formulas developped by my friend Jean Michel Le Cleac'h.
my colleagues at work are Dominique Mafrand (www.Madotec.fr)
Hugues Borsarello and Olivier Roblin from Borsaudio (www.Borsaudio.com)
and Anaël Rousseau from nAr Cables (http://www.nar-cables.com/)

i met André Klein for the first time during this triode festival and never heard about him before. he built the big "steamer", an incredible audio sculpture which won the thoriated tungsten amplifier contest.

i do not love too much classical bass horn like Jbl 4550. the bass reflex low end do not match the over-100Hz horn bass.

i have made several bass horn prototypes, straits or curved, for 10" and 15" drivers, with several cutoff from 100 to 40Hz. some with closed cabinet or with open baffle.

For this Musique-Concrete horn system called "Grande Castine" (it was developped in Saint Cast Le Guildo, Britanny, France, from were females inhabitants are called "Castines" ) i prefered a modified 60Hz Le Cleac'h cylindrical horn, with a "decompression" ratio of 1/1,4.
the goal is : no cabinet sound, no bass reflex sound, high efficiency, a cutoff low enough to use an active infra-bass cabinet for the first octave, and a resonable size (1m*1m*80cm)
it is fully usable from 60 to 250Hz.
actual drivers are ALtec 515-16G 

the mid is a 220Hz 2", with modified Radian 950Pb high-mid-treeble is a modified BMS 4540

crossover : active 18db, 250Hz for the bass
passive, 300Hz 6db for mid
passive, 4000Hz, 6db for high.

The problem with this bass horn is a great sensitivity to room acoustic.  not very surprising for an open baffle ! if there is any resonance in the room, bass become less "slamy", and fusion with other register is less good.

several evolutions are coming : custom made drivers, an "open-cabinet" for the bass horn, a quite heavy all passive-all symetrical crossover...and an active bass cabinet for the 25-60Hz area.

 i am quite proud of this system : very good fusion, no sweet spot problem, usable at high sound levels with a single watt, light-weight and beautiful for a horn system...any audio anthousiast can come home with his own cd's. Saint Cast le Guildo is near "Le mont Saint Michel", west of France : http://www.saintcastleguildo.fr/

the system at home : http://blogs.yahoo.co.jp/yokotakeuchi1968/6637163.html  (the double 18" infra-bass is not in use actually)

best regards,
Marco ---------------------------------

Marc HENRY wrote:

For this Musique-Concrete horn system called "Grande Castine" (it was developped in Saint Cast Le Guildo, Britanny, France, from were females inhabitants are called "Castines" ) i prefered a modified 60Hz Le Cleac'h cylindrical horn, with a "decompression" ratio of 1/1,4. the goal is : no cabinet sound, no bass reflex sound, high efficiency, a cutoff low enough to use an active infra-bass cabinet for the first octave, and a resonable size (1m*1m*80cm) it is fully usable from 60 to 250Hz. actual drivers are ALtec 515-16G

Well, I did not know that you have a second bass channel. Surely it is better direction to build bass. Are your first octave cabinets sealed? If so then you are in “open bottom” environment and it is, as I fell, is the best way to go. Or course if you have a dedicated first octave channel then for you upperbass you did not go, as most of the Morons would, for 515B or 515E type of driver buy you smartly went for 65Hz 515G. The 515G in 60Hz horn would do very nice… but… there are a few “buts”:

You need to unload the unnecessary bass out of this 515G driver, use 25Hz -30Hz high pass

Your 60Hz horn with that very large throat doe not load the driver down to 60Hz. You might disagree with me but make an experiment. Put 515G in an open baffle and RTA response. Then ran the same RTA from the same position with 515G horn-loaded. Now you will see where you horn will have equalization. Wherever plus 6dB boost stops from horn it will be the threshold what your 515G will start to run as a direct radiator.

I understand why you went for double driver in bas horn – to keep the thing shorter but I found it too much comprise to pay.

Marc HENRY wrote:

the mid is a 220Hz 2", with modified Radian 950Pb high-mid-treeble is a modified BMS 4540

Hm, the BMS drivers. I dealt with them in past, was not successful.

http://www.romythecat.com/GetPost.aspx?PostID=2855

There is guy who run a little own ass-kissing sweatshop http://www.bd-design.nl and he claimed that he used “modified BMS”. When I popped up at his site once and asked him about the modifications he was extremely not responsive. Ironically I was not ask what he modified (and I was under heavy suspicions that HIS modification worth nothing) I just asked him to name what dissatisfied him in the sound of the default BMS driver. He become very irritated, banded me from his site and if I am not mistaken he deleted my questions… become I “behaved badly at other forums that he read”. Interesting that I did not do anything bad at his site beside juts asking him why (sonically) he decided to “modify” his BMS drivers. Apparently the little sweatshop-running, frighten piece of shit felt that the questions were too highlighting the BS that he would like to keep secretive from his idiotic customers. Anyhow, I do not insist but if you wish to share your thighs about the BMS drivers modifications (and I do not encourage you to spill you secrets if you have any) then feel free to do so.

Marc HENRY wrote:

crossover : active 18db, 250Hz for the bass passive, 300Hz 6db for mid passive, 4000Hz, 6db for high.

Very good. May I suggest you something? You are not the subject of “Stop having a fear of horn vignetteing” rule:

http://www.romythecat.com/Forums/ShowPost.aspx?postID=6147#6147

as you have large PARALLEL surface ahead of your tweeter. Why don’t you try to do what I did in my Macondo – use tall ribbon. With all other arguable advantage and disadvantage of Ribbons a tall ribbon would act as a line-array dramatically decrease the vertical radiation pattern – it is exactly what you might  need with your “Grande Castine”.  If you are interested you might read about the line-arrays:

http://www.romythecat.com/Forums/ShowPost.aspx?postID=3013#3013

 … the type of the tweeter I use:

http://www.romythecat.com/Forums/ShowPost.aspx?postID=2974#2974

 … and the thoughts that lead me to end up with what I ended up with:

http://www.romythecat.com/Forums/ShowPost.aspx?postID=3766#3766

Marc HENRY wrote:

The problem with this bass horn is a great sensitivity to room acoustic.  not very surprising for an open baffle ! if there is any resonance in the room, bass become less "slamy", and fusion with other register is less good.

Yes, you are right. Open baffle is superbly affective tool to fight the room problems, however the dipole has on intrinsic problems – the type of the bass it produces:

http://www.romythecat.com/Forums/ShowPost.aspx?PostID=1045

Also, I disagree conceptually that you try “curing” room by topology of speaker.  I am pitching very different philosophy. If you interested then you might read it here:

http://www.romythecat.com/Forums/ShowPost.aspx?postID=4421#4421

Marc HENRY wrote:

several evolutions are coming : custom made drivers, an "open-cabinet" for the bass horn, a quite heavy all passive-all symetrical crossover...and an active bass cabinet for the 25-60Hz area.

Interesting, keep showing what you will end up with. I will promise  to keep criticizing everything – it is looks like my specially :-) BTW, why the crossover should be “quite heavy”? Why do not do passive line level and forget about all problems with crossovers. Do not repeat my mistakes when I orders expansive air-core 8ga coils. There are many reason what it did not and should not work.

Marco, sorry for many links but I did not want to write again that was said formerly…

Paul, l I have written about it somewhere: the passive speaker-level crossovers on bass channels are no good. Marko has 250Hz that with 3rd order would be around 10mH and 2mH. With air core then would be very large coils with huge amount of DCR – very negative for bass where .5R of series resistance is well- auditable. If Marko goes for 10ga (I do not think that he will be able to go for 8ga on 20mH) then it will be lover DCR but still substantial. Nevertheless, the 10mH air core with 8ga wire it will be around 50-60 pounds and around $000 each. Is it necessary? The ERSE meal-core coils have very low DCR and for channels that care no HF they will do equally good with air-core. At least in my sub 600Hz channel I saw no difference.  Not to mention that the super large air-cores cones, connected to the driver act a perfect resonance contour and begin to suck AM waves – with high sensitivity speakers all of it will be playing right in your face. If you even built an EMI/RFI shield around 12” air coil then you know how much problems it is.  Even if Marko go for regular first order filter (as he shell do with a single driver) then it will be 5mH against 8R – it will be more manageable but still it will be too large and unnecessary in my view.

There is an additional bitch on passive speaker-level crossovers for bass – it is impossible to calculate them. The driver impedance wares and the Q of the filter is never stable. In addition the coils of the crossover begin to talk with driver inductance and it minimizes the effectiveness of the filter (it would greatly depend from the type of the drivers)…

I was wondering: if Marko has been using active crossover on his bass horns then he has a dedicated amp for that channel. All that he need to do it just to build-in a passive like level crossover into the amp. He would also need to revise that 3rd order filer as I do not think it was right. He has Radian driver sitting in 220Hz horn and he crosses it at 300Hz, first order? That is VERY aggressive!!! I am quite confident that 220Hz Le Cleach horns must not be used with 300Hz 6dB – they should “choke”: (read the quoted section)

http://www.goodsoundclub.com/Forums/ShowPost.aspx?postID=5317

I think Marko eventually will move the crossover point to ~500Hz and then drop the sharpens of the low-pass filter on his bass horn. BTW, it would be fun THEN to talk about the modification he did on his Radian driver as the driver will be working in completely different and much less stressful mode.

And there is no substitute for listening to it, no matter how careful the "calculations".  Anyway, I don't have it "figured out", that's for sure. Some guys figure in and use the inductor DCR in lieu of resistors (apropos), which seems clever.

I actually think there is more "slop" available at speaker level, and this may be the main "attraction" for me; but I like at least the idea of smaller parts at line level, albeit more need for precision.

It's been quite a while since I used a variable active X-over, so no recent frame of reference/aural memory, since back in the days of "flat".

If Marko "needs" 3rd order, is somerthing else amiss?

Best regards,
Paul S

Paul S wrote:

If Marko "needs" 3rd order, is somerthing else amiss?

Marko high order low-pass filter on his basshorn is perfectly understandable and very predictable. To make it “right” it should go fist order but then the entire system should be rearranged.

1)    The MF crossover must move up. Come on! The 300Hz crossover in 250Hz horn – it is not serious.  If the MF crossover goes up then the filter on upperbass will be able to use 3rd order instead of 2rd order

2)    The notorious double drivers in upperbass horn. If it use one driver, and preferable with not 15” throat then the 2rd order might become a firs or the firs and a half order. The sampler throat = the longer horn. The longer horn = more LF EQ and more HF attenuation. The smaller throat = the larger front chamber in front of 515G. The larger front chamber = more HF attenuation. So it is perfectly possible to kill HF from 515G by natural means.

Sure, it would be very different speaker but I do not propose how Marko "should" change his speaker, I rather I advise how a conceptually better speaker might be made pointing out to the  various typical compromises and problems in different horns acoustic systems. So, I hope Marko would not take it personally and would not turn suicidal…. :-)

Markus wrote:

Very interesting discussion. I have taken a close look at Marco's horns and am thinking about using them in a horn installation. I plan to use the TAD 4003 from, say, 1 to 8 kHz, and a tweeter above that. I would love to use horns below 1 kHz but will prbably have to do without for space reasons. I'm thinking about the TAD 1102 between 200 Hz and 1 kHz and the JBL 1500AL below that. Doing the 200 Hz crossover properly in passive is almost impossible, so I'm pretty sure I'll use an active crossover, but I haven't decided about the 1 kHz point. Anyone have experience with passive versus active at that frequency?

Be careful between that 200Hz-1kHz–it is VERY critical region. You might try also JBL LE8 - it is proven phenomenally sounding driver for your range.

About the crossover: I do not know what would mean “almost impossible”. I also do not understand why people question “passive versus active”. Active crossover implies multi-amplification and if so then the question should ne not passive versus active but multi-amping of not. If a person has selected to go multi-amping then there is no doubts in my mind that “proper” crossovering should be implemented at line-level passive. There is absolutely no need to introduce in signal chain another active element juts become you need to introduce a slope for a given channel. Any amplifier provides plenty opportunity to stick a filter in it: in input, in feedback, on plate, between stages and in many other locations… In many instances those filters improve the amp itself.

Markus,

If it is your task to stay with a single amp then all bids about line-level crossover are off. I would like to stress that although there are zillion theoretical and applied reasons to prefer line-level but still it does not mean that regular passive crossovers at speaker level could not work out, regardless the frequency where they are implemented. It is all about your choice, intentions and desire to scope out the application within a specific scale and to project the result to the reference points with which you are feel comfortable. I am not against passive crossovers at speaker level. I am against active crossovers as they, I feel, should be replaced with line-level passive.

nl wrote:

However, these sorts of basshorns with twin 15" drivers and a very large mouth (larger than the cones of the drivers) have a long history. The Altec 210 cabinet, as used in the Altec A2 systems with two 515-type woofers, is a prime example. The engineers of these designs did actually know what they were doing, and made a series of engineering compromises. I can imagine what they are -- shorter horn length, smaller size, lower manufacturing costs and easier installation in theaters etc. Remember, these engineers were the very same ones that also made the compression drivers we use today like Altec 288, which typically have a 2:1 diaphragm/exit ratio. Some have an even smaller mouth, like the WE 555 which uses a 2" diaphragm and a 5/8" exit. While these are not "proper" horns in the fashion of midrange horns, nevertheless they have many horn properties such as increased sensitivity. So, they are not just funny-looking bass-reflex boxes either.

In many ways they are, nl. It is very simple to figure it out – just see where the horns stop EQ, let say 3dB, of the original driver response. If disregard the horn profiles and room loading then the delta between the sq inch of mouth and throat are important. The Altec 210 might not be considered in this case as they are also act like open buffers and therefore they (nether Knagfilm high-pass their horns)

nl wrote:

I think an interesting question would be: what are the various compromises made, and what are the results of those compromises? In some cases, the "compromise" might not be that important, for example reduced high-frequency extension might actually be an advantage overall. Thus, maybe we could come to a conclusion that "this is a good Altec 210-type design for those who must make those compromises, while this (other) design is best for those who can accept a ten-foot-long horn in their room."

Yes, it is trueand I see you point. However, I kind of always think more about sound reproduction then about loudspeakers making. Pretend somebody has special health need –100% oxygen to breath. You might feed the person with 50% oxygen and the person would not die, but your will also will not get out if this person the expected from him results. The same I feel with balance between compromises and design objectives. Sure, everything is compromise but let do not take 3 cub. inch enclosure and try to pull out of it 10Hz bass. We for sure can verbalize that it was good 10Hz bass for 3 cub. Inch enclosure but… something does not work for me in this logic. Can we name the multi-channel time-misaligned horns as Good Sound for random phase sources? Sound is Sound and has own needs, own inters and own agenda. I feel that it makes sense to talk about sound without mentioning the topology of sound reproduction. It sounds perhaps slightly minimalistic but in fact it is not.

nl wrote:

Within that horn-loaded range, the 210/515 type combo does reasonably well, and produces rather more inspirational midbass than most direct-radiator type strategies. There is a reason why these designs persisted for roughly forty years, until replaced by smaller/cheaper/inferior sound systems (but with better low bass) to match the smaller/cheaper/inferior theaters built since 1985 or so.

Well, yeas and no. Although I agree that 210/515 type combo does reasonably well in lower midrange and upper bass but it has in my view a different upper bass then I would like to be. The 210/515 type combo never was able to play soft, nether dynamically nor tonally. Also, I personally never saw a successful integration of 210/515 type combos with lower frequency channel. All those open baffles as soon they reach the region of acoustical shortage do something to sound that poisons me. 

nl wrote:

So, as I said, it would be interesting to understand what one gains or loses by this method instead of something very difficult to implement such as a 10 foot long 70hz horn with proper compression loading. Much easier to time align as well, which is one of the reasons for their original development in the 1940s.

OK, let look at the typical 210-like horn with 50-70Hz and dual 15” throat - we understand that it is short and wide horn, very similar to what Cressaro and many others used for their bass bin. I call it the “big throat approach”… I have writhe about it many times, look in my “Problems with horns: upper bass “thread linked to this thread. When I see the 210-like horns I always ask myself what kind of driver I need to use in this configuration. Let pretend that it is 50Hz horn, then what driver? The common rules suggest that I would need a driver with let say 45Hz resonance frequency but the Wide Horns do not load LF. We take the driver in open baffles and measure it response. Let pretend that it has 100dB at 90Hz. Then we load the driver in the horn. The end of the 210-like horn working as a horn will be not what the circulation of mouth suggest but where the driver stops hitting 106dB (depends of profile it might be more or less but 6dB is a good average). The 210-like horn will not have a lot of horn gain. The response will be speeding lower than in open baffles but it will not be the horn EQ but rather the driver acting as a direct radiator. Then we have more problems. The driver is capable to do now 50Hz as a direct radiator but the mouth is 70Hz. Some say not a big deal, in fact many idiots stick 20Hz driver into 50Hz horn but that destroys everything sonically. So, my point is that in a properly sounding horn-driver configuration the bottom of the channel response shell be the sound of horn not the sound of driver. It is imposable in 210-like horns unless you high-pass the horn at I would say 100Hz-120Hz. Then you have correct sound out of the horn but at colossus waste of the horn mouth. You can get the same effect if you use a driver with free air resonance frequency of 80-100Hz but again: why would you need 50-70Hz mouth for such a driver if the horn can’t not yield the LF gain. In my view the direction is to minimize the size of the throat and unfortunately brings the length of the horn all the way up.

So, how I found the 210-like horn with good driver sound?  It is important to understand that from my perspective there is no good drivers for 50Hz horn and 2x15” throat, there is no good hypothetical driver as the shell be cut off at 120Hz to serve inters of sound instead of the interest of  acoustic pressure. Among the few 210-like horns that I heard they all sound absolutely in the way I am describing: the excessive LF on the woofer chocks the horn’s bottom knee, making sound very hard and very stiff. The last my exposure to the mistakes I describe above was last year when I visited the Oswalds Mill guy. He has the bottom born of the same idea: modified RCA ~40Hz mouth with a pair of 22Hz Altec 515E loaded into the horn.

I usually do not comment about my private visits publicly but in the case of that dude my rules off the limit and the Oswalds Mill guy fully endorsed my attitude. So, that version of the 210-like implementation sounds spectacularly horrible and it is not an accident. Anything below 200Hz was literally shattered by the high pressure-boom coming from those horns. it sound like you have month fill of water and you are trying to sing - a horrible feeling. The Oswalds Mill listening space is huge and it dissipates a LOT of low frequency, so Jonathan was forces to crank up power at bass but the more powers this choked horn endure the more horrible it sounds as the choking in effect rise with volume of the horn.  As the result anything under the lower midrange sounded like wet dog and was truly revolting. I did not explain to the Oswalds Mill guy the details – he is not particularly smart man in audio and he not particularly knows, cares or express any interest about sound, We just listing some records I brought and it was extremely indicative. We were playing the Tchaikovsky’s Pathetique symphony and I asked the Oswalds Mill’ wife what she thinks about this music. Thanks God they do not know is and her reaction was very interesting and natural. She said that she has a difficult time to understand it because all orchestra sounds the same and there was not a lot of thing doing on between instruments. I have to add that we were listening Mravinsky from 1961 – the most articulated Pathetique even committed to record media to my knowledge. Sure, I did not explain to her that it was not Tchaikovsky or Mravinsky faults but the sonic deserters her husband implemented in his bass horn.  However, her very natural reaction was a very good depicture of what I am trying to say – the bass horn are dangers and it is better do not do it if they are not done properly. I am sure if the same Oswalds Mill guy stick 4 drivers on sealed/ported box then he would avoid many horn mistakes he made.

Markus wrote:

Some rather astonishing features. I'm very interested to hear it.