The attachments show the plan view of four side-by-side rectangular 15Hz exponential horns. Each horn has a mouth 8' wide x 14' high. Attachment 1 shows isophase wavefronts at 100Hz. Attachment 2 shows isophase wavefronts at 300Hz.
Attachments
I don't even know what you are talking about.
Two 16 foot wide horns sitting side by side do have two clear centers.
I told you that I'm not going to use any software for simming anymore except Direct (even though the other software agrees reasonably well with Direct) so you don't have to talk about point source sims anymore.
But don't confuse a subwoofer having an acoustic center as a point source sim. If Radalescu is right a Direct sims is not point source. But you still have to put the sub in the right spot based on it's acoustic center.
This illustrates my point quite clearly.
If the wavefront was completely flat as a pancake exiting the horn and stayed completely flat as it propagated past the mouth the wavefront would be completely in phase and it would be a solid color all the way across.
But it's not, there's considerable interference across the wavefront as a direct result of the curved wavefront even as low as 100 hz.
To be clear, the narrow beam of directivity is also a factor here, but between the two factors (curved wavefront and directivity) there's clearly hot spots across the wavefront as it propagates forward that you wouldn't have with a single sub.
I am still not convinced that there's a problem that needs fixing.
If I achieve a "wall" of stereo sound waves, that drives deep down to 15 Hz cleanly, then mission accomplished! A little unevenness within the sound field is not a deal breaker for me (this is a backyard DIY project). I'll have weatherproof outdoor horns, 100% free of room nodes, that will last for 50 years. I can play hide & seek with my kids in the horns, and they can pay hide & seek with their kids. And should I ever feel the need to sip tea, I'll simply place my chair within the one sweet-spot that has minimal lobing. . . . .
I figured you'd be happy Just-a-guy - my horns are now side-by-side, in lieu of 60 feet apart. My listening area is near-field, in lieu of far-field. My power requirements have dropped, and reductions in lobing have increased my overall system fidelity. You were an integral part in making these design changes.
The reflection of sound off a room wall adjacent to a speaker is, with some attenuation, the same as having a second sound source. Ditto for all stereo systems. And some math-compulsive people insist there is TERRIBLE comb filtering and lobing.
I don't hear a problem. Does anybody?
(The reflections in a room provide ambience which is essential to add to most recordings which are conjured to play in rooms. Without the goldilocks amount of ambience the sound is lousy. OP might want to introduce reverb to his outdoor system.)
Ben
Again you are not paying attention. There's no problem at mid and high frequencies. The notches are not deep and they are very narrow. You can hear it but it's not going to impact your listening pleasure.
At low frequencies in a small room it's hard to get your subs very far apart, which eliminates most of the deep two octave wide notches in the subwoofer frequencies.
But in small rooms we have a different issue, the standing waves. And you deal with them with treatments and/or the averaging effect of multiple subs.
Also please don't call me math compulsive. I hate math, I'm not good at it and I avoid it like the plague.
What destroys bass fidelity inside a room is standing waves (room nodes), and not necessarily comb-filtering. This is the prime reason why I'm building my horns outside.
And I simply will not have 2-octave wide bass nulls in my sound field, from (qty 4) 14' x 8' side-by-side 15 Hz bass horns. The horn mouth behaves more like a radiating surface, than it does a point-source. The sound wave angle-of-inclination between horns will be gentle, thus any destructive interference (lobing) will be gentle.
Additionally, 32 feet is a "room" dimension. . . .
Last edited:
This is literally exactly what I just said.
Is it qty 2 or qty 4 bass horn? You keep switching back and forth and it does make a difference.
Also, what are you using to determine what your coverage pattern is going to look like? Gut feeling?
Direct takes this into account as per the previous poster's description of how it simulates.
Regardless, by the time the wavefront of each individual horn gets to the hot tub it's going to be considerably curved. I'm not sure if you are understanding that sound wants to disperse in all directions in free space when it's not confined by a boundary.
I already told you it would probably be ok, but it's because of the extremely narrow directlivity, NOT because of the wavefront shape. With a narrower beam there won't be as much overlap at the higher frequencies, less overlap = less interference. Also with 4 sources there's considerable averaging creating a smoothing effect.
That's a pretty big room, I don't have any rooms that big.
Last edited:
I'd like to build 4 horns, 15-Hz straight exponential, 14' tall x 8' wide each, side-by-side.
This approach will require significantly less reinforcement to suspend (cast) 8 feet of overhead concrete - compared to 16 feet wide, and especially compared to 32 feet wide. I can build one set of forms, and use them 4 times - to get 4 identical horns. I can spread drivers across (qty 4) horn throats, in lieu of two, or just one throat (i.e. significantly less driver clutter).
So next design question: Should I build the bass horns with the mouths in-line with each other (i.e. a straight wall, with parallel horn axis), or should I angle the horns, so that each horn's line-axis converges on common focal point within the listening area (i.e. a concave wall)?
If I converge the bass horns to a common focal point, I'll maximize sonic performance at that listening point. But my question is, what happens outside of that common focal point. I.E. I do not want to make one location sound really good, by disproportionately sacrificing sound quality outside of the focal point.
Another potential problem with a curved wall, is that any sound reflections will converge right at my focal point also. . . . (I'll draw & post a sketch this evening)
Last edited:
A bit more - converging the horns to a common focal point in the near-field, will increase the angle-of-inclination of the combining waves, causing lobing to become more pronounced off-axis within the near-field.
My gut tells me I should keep the wall flat, or very gently concave at most - for a distant far-field focal point. Independent of the bass horns, the mains would obviously point towards the listening area focal point.
Or said another way - - - I'm not sure I'm ready to permanently "cast" a near-field focal point into my bass horn design. . . .
Opinions?
My gut tells me I should keep the wall flat, or very gently concave at most - for a distant far-field focal point. Independent of the bass horns, the mains would obviously point towards the listening area focal point.
Or said another way - - - I'm not sure I'm ready to permanently "cast" a near-field focal point into my bass horn design. . . .
Opinions?
Last edited:
Hi Entropy455,
If you delay the two horns on the outside, you don't have to curve the front 🙂
Regards,
Djim
If you delay the two horns on the outside, you don't have to curve the front 🙂
Regards,
Djim
1) Almost all multi-cell horns are simply a collection of exponential horn cells. My last experiment was a collection of conical horns, they "played poorly together" ;^) .
All exponential horns beam-width reduces as frequency increases, causing "fingering" in a multi-cell's response, each cell has it's own high frequency "ice-pick" surrounded by less HF, simply count the number of cells to arrive at the "ice-pick" count. Compromised designs have no "vintage", there were great horns made during the same "era" that multi-cell horns were used. My previous post explained why they were used back then, and why the compromise is just not acceptable compared to the many far better sounding alternatives available now.
2) "Too much" could also be argued as "just enough" if you have a wide "couch" to fill, or in outdoor situations, would like the sound to be good anywhere audience members may end up.
3) Your modern "high-fidelity" horn has absolutely no definition.
What Entropy Eric, you, Don Keele, Bill F, JAG, W. Geiger, Josh Ricci, and anyone else reading this post find to be "high-fidelity" is a changeable opinion.
Keele specifically advocates a completely different approach to "high-fidelity" now than when he was selling different versions of his "high-fidelity" "constant directivity" designs to Altec, JBL, and Electro-Voice.
Having heard, and compared Don's horn designs to his own various iterations, and those offered by other designers, including my own, over more than four decades of listening and measuring, I can say he has had some winners, and some losers.
The whole "trick" of sound system design is knowing what combination of elements make a winner, and what will be a loser, for the given application or client's taste. Like my dad used to say about the design and construction of apartment complexes, after he got out of the "race to the bottom" that was happening in single home construction in the early 1960's: "if it were easy, everybody would be doing it";^) .
4) Bill F. has given some good advice in a few posts, but neither his "logic" or "imagination" are evident well in the quote.
"Logically", it is always preferable to use a single point source to cover the intended listening area, multi-cell exponential horns are a "low-fidelity" approach to wide coverage demanding maximum sensitivity over "high-fidelity".
Keele's EV contant directivity horns did actually have a multi-sectoral throat section, but they were not a radial horn. A radial horn using an exponential flare specially can not meet the definition of "constant directivity".
The main reason multi-cell exponential horns are not done much anymore has nothing to do with cost- people are still paying fabulous amounts of money for antique multi-cells- but has everything to do with the fact that they offer no sonic advantages to the vast majority of the world's consumer base.
5) I stacked about a half ton of JBL radial horns on each gig from around 1976 until 1981 until I found horn alternatives that sounded better, had more usable polar response, were easier to build, and cost less. Radial exponential (or tractrix, or any other HF "beaming" flare) horns, like radial engines, have certain applications in which they "shine", but they are not the best solution to many applications, especially any application requiring more than one device used for vertical coverage.
6) Yes, I have plots. Lots and lots of plots of dozens of different iterations. It started to go "blah" in New Mexico a bit after the Christmas Holidays last year.
Like I said before, the whole "trick" of sound system design is knowing what combination of elements make a winner, and what will be a loser, for the given application.
I knew going in to the Maltese multi-cell project that it could not be an "ultimate" winner, and the two-cell horn array is not a loser, but the quad cell is only a winner in terms of SPL and "high-fidelity", it's polar response "sucks".
Polar response is often not high on people's priority, but in the intended application I planned to use the Maltese quads, it was nearly top priority- the lobes and fingering in the polar response can result in terrible feedback from downstage vocal microphones that are on the edge of the polar pattern.
At any rate, what is marginal for my application could work perfectly well in another, I'm sure I will find buyers for the "full range" Maltese horns after replacing them with four SynTripP Too© cabinets each loaded with 1) TC-9 and one 8" Celestion BG8-60, if I ever get around to it.
I just spent another eighty minutes typing instead of "getting to it"...
Art
Last edited:
Interesting. . . . So for near-field listening (assuming my horns are parallel - as shown in post #901), I could fire the outer horns a few milliseconds ahead of the inner horns, to "focus" the sound in the near field???
Hi Entropy455,
Indeed, with multiple horns/subs you can alter the pattern/sweet spot (incl. lobes) by using a delay. It gives much more freedom than arranging them in a (fixed) 'classic' arc.
In your "nearfield listening" situation, like you described, I would experiment with a lower Xover and 'some' negative gain (some means distance depending) for the inner basshorns. That way you have a smoother transition between outer left basshorn and left top versus outer right basshorn/ right top.
Anyway, you already seem to get the 'hidden' possibilities 🙂
Regards,
Djim
Indeed, with multiple horns/subs you can alter the pattern/sweet spot (incl. lobes) by using a delay. It gives much more freedom than arranging them in a (fixed) 'classic' arc.
In your "nearfield listening" situation, like you described, I would experiment with a lower Xover and 'some' negative gain (some means distance depending) for the inner basshorns. That way you have a smoother transition between outer left basshorn and left top versus outer right basshorn/ right top.
Anyway, you already seem to get the 'hidden' possibilities 🙂
Regards,
Djim
it will have no practical benefit.....
edit: sorry, should have added that as djim pointed out, delaying the two outer horns would widen the coverage pattern of the horn. delaying the inner horns would narrow the pattern and shorten "throw"
check this:
practical guide to bass arrays pdf
and this
www.electrovoice.com/downloadfile.php?i=8913
Last edited:
so these plots do contradict JAG's assumption that combing would occur in the listening area in front of Erics currently favored four-horn-array, don't they? If the wavefronts exiting the horns are coherent, there should be no comb filtering, right?
Djim,
For the size subs we typically use or hear at concerts, they can effectively be modeled as a series of hemispherically radiating sources of "X" spacing.
Although the position of lobes by use of delay certainly is possible, exponential subs of the huge dimensions Entropy Eric has stated he plans to build, with horn length equal or greater to the wavelengths of the frequency of the bass/low mid crossover points he has suggested, the horns have a distinct on axis DI increase with frequency. The horn's central "hot spot", pointed in whatever direction chosen for the multiple adjacently placed subs will be "beam steered" by delay differently from the torus wave surrounding the center "spike".
The larger the spacing between sources, and the less uniform the dispersion of said sources, the less effective "beam steering" is. EAW and Martin's recent beam steering products are able to effectively place the sound exactly where it is wanted, and "hard avoid" places it is not, because each pass band's transducers are located within 1/4 wavelength of each other. Entropy Eric's proposed adjacent locations are an order of magnitude larger than what is required for precise beam steering.
All is not lost with Eric's "concrete" plans, if he simply were to "toe in" his rather large pair of large bass enclosures, center of their mouths separated an equal center to center distance as his upper L/R cabinet(s), and let them "beam in" at the "sweet spot", a good time will be had by all at that central point, which will be +6 dB louder than a point equidistant to the L/R, but located in front of either array, rather than on axis in the center.
My above statement is based on actual measurements of many outdoor systems, but is dependent on proper time and phase alignment at the "sweet spot" of each different respective L/R pass band. The "good time" at that spot will also depend on how loud it is run, Eric's proposed LF system headroom is so far beyond the level I would find comfortable, I would not sit in the "sweet spot" unless I had the volume control at my fingertips !
Art
If the waves were fully coherent they would be a solid single color all the way across the wavefront of all four horns. But it's not, not a 100 hz and definitely not at 300 hz.
The narrow beams and comb filtering are what make the wavefront not a solid single color all the way across.
In Direct, zoom in the window 30-200Hz, and model several mic positions for a spaced pair of 'omni' sources. Then swap the omnis for a pair of SH50.
The position of the nulls is identical, but for the SH50, they are 3-6dB less deep.
A pair of SH50 boxes have two clear centres. The point source thing is their raison d'être. They are ~onmi at subwoofer frequencies, so in theory, their bass output should null exactly like the onmi point sources. But it doesn't*.
As you said "Direct is NOT modeling any horn as a point source", so presumably this difference represents how a big object works in the real world.
By analogy, when you've used spaced sources that are ~1% of the size of the thing you are trying to model, I think there is likely to be a similar amount of error.
*not meant as a comment on the quality of the SH50. I've never heard one, but I imagine they are grand.