Yes, lots of systems used subs placed side by side, including yours. And physics says any sound sources playing the same frequencies and separated by more than 1/4 wavelength are going to destructively interfere with each other. None of this is news.
I mentioned in a previous post that your systems were usually only 1/2 the width of the 12' 8" you were boasting about.
The fact that you didn't use a central mono sub doesn't change anything, your blocks used side by side subs (also vertically stacked but side by side nonetheless).
You sure did, but this is what you said about the welter system contribution to that system -
"... WS equipment we brought to the stadium, and dB Sound's mains system, there were a total of 16 "blocks" of speakers ..."
16 "blocks" of speakers doesn't sound like a line array to me. If you meant line array you should have said line array.
I seriously doubt any of your subs could come anywhere close to this when placed side by side. Most "small" subs have little to no directivity at the bottom of their passband and only a bit at the top of the passband. For two subs placed side by side to have less than 10 degrees overlap, especially at the bottom of the passband, they would have to be EXTREMELY directional.
You talk about using WS blocks a lot, you've mentioned line arrays once, maybe twice, and at the end of your rambling stories, so towards the end of your career.
Well, if you would stay on topic and discuss OP's system instead of your resume this wouldn't happen, would it? Don't get me wrong, I find it all very interesting, but your are providing your resume in lieu of technical responses, and I think you know how I feel about that. That and withholding information for financial gain or other motives are two of my biggest annoyances.
Yes, it's a fundamental property of physics, not just horns. But as a side note it's always about distances and wavelengths - within 1/4 wavelength you get full mutual coupling, any longer than that and things change.
Using that example you are correct.
That is not at all the example I was expecting though. After mentioning Leach and Keele and the "On the specification of moving coil drivers for low frequency horns" paper and the Hornresp equivalent, the System Design tool, and talking about calculating ideal t/s parameters and the ideal horn and how low Mms, low Qts and high Bl are all about bandwidth and nothing else, pure exponential vs hyp/ex flares, I seriously thought you were designing your newer iterations of your horn in a more complex manner.
How did you come to these example horn dimensions? Are you just plugging semi random number into the simulator? (I say semi random because I know you are keeping an eye on compression ratio and you have an idea of what size mouth your want, and at this point it seems clear that you are completely set on a pure exponential flare for some reason, but is that the only design factors you are considering when plugging in the inputs? Because if so you are probably leaving some performance on the table.)
To recap, in this case, yes you are completely correct, after all, all you did was chop the horn in half, which is how I proved the concept of 1/2 space horns to you in the first place. To be more clear, if you had used Leach's math or the System Design tool, a dual side by side horn designed for 1/4 space loading will not be the same as a single horn designed for 1/2 space loading. That's kind of why I said that chopping a horn in half isn't the same as designing for 1/2 vs 1/4 space - and what you did in this example is chop it in half, you didn't really "design" it by any standard of using horn math equations.
Last edited:
JAG,
Again, you ignore nearly every salient point made for the benefit of the OP.
1) It appears to be "news" to the OP, who is proposing four side by side subs covering a wide pass band.
2)My 1980's side by side subs center to center distance was slightly over 1/4 wavelength at the crossover frequency. I learned about wavelength math around the turn of the century, was busy building systems and running a sound and lighting production company doing 800 shows a year through the early 1990s.
3) "Line Array" is a marketing term that became popular in the mid- 1990s that now is used to describe vertical arrays of any and all descriptions. It is no more descriptive than the term "car" is when writing about an automobile.
I described the vertical arrays WS and dB Sound used in 1980 in great detail, would you like me to also draw you a picture, as I do not have a photo of the event?
4) Obviously the bass horns I built and used in the distant past did not have a DI of 90 degrees other than in the upper pass band of 200 Hz we used in the "olden days".
JAG, please make an attempt to contextualize the information I present instead of continually showing off your ignorance of stated facts like it is an attribute that brings anyone closer to the light of the reality that side by side by side by side stacks of any audio gear without describing an arc from their acoustical center are a thing that should be relegated to the ash can of history, in spite of the fact that the practice has continued for decades.
That said, we are in the age where digital manipulation of the acoustic center of VLF can be effectively accomplished over a fairly wide distance, depending on the definition of VLF and "fairly wide".
Cheers,
Art
Last edited:
I'm sorry if my example came across as overly trivial. Nonetheless, the horn dimensions remain proper for a “generic” 15-Hz straight exponential bass horn. You have been pounding home the subject of lobing, and how it should be avoided – and I have been trying very hard to understand, and how to control the problem. Again, this was a generic example, intended only for conceptual understanding.
Permit me to be more specific with my question – Example 1 through 4 above have identical overall mouth dimensions (14 feet tall, 32 feet wide). I’ve already stated my rationale for example one – but I’ll state it again for the sake of clarity. Four quarter-space 15-Hz horns, with the left two playing the left channel – the right two playing the right channel - stereo. And assuming the bottom octaves are recorded mono anyway, the bulk of the sound will be a mono full-space horn. And with an 8 foot projected horn-throat spacing, they’ll sum nicely up to about 35 Hz. And at about 141-Hz & up, each horn will appear as a very discrete signal source.
Example 2 - aside from this setup being 10 times more difficult to cast out of concrete than example 1 (requiring a complex foundation), does this setup buy me anything? (still running stereo, similar to example 1 - just rotated 90 degrees). The 8 foot spacing remains, as well as the resultant lobing issues – however they’d be in the vertical plane, and not the horizontal.
Example 3 - this is just two half-space 15-Hz stereo horns, side-by-side. I liked the idea of example 1, over example 3, because I'd have more room for drivers in example 1 (less driver clutter)
Example 4 - a mono full-space 15-Hz horn. I acknowledge this could be a half-space horn, but then I'm back to a mono system, and with a VERY cluttered horn throat. . . .
I would like to know how I may get a pair of 15-Hz "stereo" bass horns to sound good outside - feeding a relatively small backyard listening area. Your simulations clearly showed that bringing the stereo horns together would be much better than a 60 foot spacing - and that is what I've done, conceptually. . .
Last edited:
Entropy -
No physical/acoustic separation, no stereo illusion. There's no "stereo sound" in any other sense and certainly none in lower freqs.
I sure wish you'd read up on comb filtering and lobing. I don't know why people with engineering training (as well as people with no training) think they MUST be terrible distortions and worth every compromise to combat them, when the empirical evidence is otherwise*.
Ben
*OK, I think I do know why. Umm, same people think stereopsis is the be-all and end-all visual cue to depth.
No physical/acoustic separation, no stereo illusion. There's no "stereo sound" in any other sense and certainly none in lower freqs.
I sure wish you'd read up on comb filtering and lobing. I don't know why people with engineering training (as well as people with no training) think they MUST be terrible distortions and worth every compromise to combat them, when the empirical evidence is otherwise*.
Ben
*OK, I think I do know why. Umm, same people think stereopsis is the be-all and end-all visual cue to depth.
Last edited:
1) The transition from a large source to multiple discrete sources with different path lengths to the listening position is a problem.
2) Vertical arrays eliminate horizontal lobing problems, but there is absolutely no sonic advantage, and considerable sonic disadvantage to using multiple point sources. If you need the headroom of multiple drivers, simply manifold couple them to a single horn, or two stereo horns at the distance apart that provides the mains separation you desire. By the way, using a separate compression chamber, you could start with one driver per horn, and add more chambers and a manifold later if you decided 130 dB SPL (at one meter) was not adequate.
3) Since separated 60Hz (old days) stereo bass horns sound good, stereo separated Keystone 35 Hz (this century) bass horns sound good, there is no reason to expect a different sonic outcome other than the bit over an octave deeper extension will do justice to any low pipe organ and synth bass lines that have been recorded, as well as the reverberation "cues" from very large halls.
There is one serious sonic disadvantage to a 15 foot straight horn vs. a folded horn (which can also be made of concrete) shown in the diagram below.
Although a folded horn is more complicated to build than a straight horn, it will use considerably less materials, and has a distinct sonic advantage for your (and my) preferred embodiment, stereo subs, located with the top cabinets directly above the horn center.
The diagram below shows a more distant listening position than I would prefer for optimal stereo sound, which would be at the apex of an equilateral triangle. At any rate, if built using folded bass horns, at any distance from the L/R speakers along the center line will sound great, and no position off axis will sound "bad".
OK, my work here is done ;^).
Art
Attachments
Last edited:
I am trying to contextualize the info you present, even though it is completely off topic. What I would like is for you to stop talking about yourself here and talk about OP's project. If you like you can start a thread on your adventures or maybe write a memoir but your ill described systems don't really have any place here other than to be place fillers for actual technical information about OP's project.
This is where the problem came from, it's semantics. Instead of calling it 4 quarter space horns, if you would call it one half space horn chopped into 4 that would be more accurate to the process you used to create it, more accurate to the environment it's working in and a lot easier to parse.
You can sim this stuff with Direct. Example 1 - 3 will have lobing issues, example 4 won't. The reason is that example 1 - 3 are well more than 1/4 wavelength apart (center to center) at the top of the passband. To be clear, the lobing issues will be a lot less severe than having the subs 60 feet apart, but there will be issues.
Like I've said a bunch of times, there's no stereo illusion at subwoofer frequencies unless the sound is panned hard to one side which is not really an effect worth recreating. This is the whole point of the non localization of subs - you can't locate where the sound is coming from. And as pictured, even if you could localize the subwoofer frequencies, and even if you did play them in stereo, you would have no stereo effect because they are simply too close together in these examples. It's like listening to a ghetto blaster from 12 feet back - sure there's two speakers and they are playing in stereo but they are very close together, so at 12 feet back do you really hear it in stereo?
If you absolutely have to have stereo in the sub frequencies you might as well just build your dual subs and place them 60 feet apart and just live with the comb filtering.
I've said it a dozen times, I'll say it again. You need to do some testing outdoors and determine which aspects are truly important to you.
Last edited:
Thanks. This is what I've been saying - dual subs will have severe comb filtering and cause considerable sonic disadvantage. (Yes I know very well this is not the point you were trying to make, but it is true at all frequencies.)
This last paragraph and the concept it describes is not going to be an issue.
If you draw it to scale and calculate the path length difference it isn't going to add up to much more than a couple of milliseconds at any point on the horizontal plane, basically inaudible.
If you draw out the folded horn to scale and do the same thing, this will indeed cut the problem roughly in half but since it's already basically an inaudible non issue it doesn't matter.
OP is going to make straight expo horns, this much has been clear since the first post of the first thread he started. So this issue doesn't matter, it is what it is and it certainly isn't any kind of deal breaker.
Hi 'just a guy',
Since you claim "physics says", I assume you can show a (valid) source or use the correct formula's from physics? 🙂
Regards,
Djim
Agree with a lot of weltersys's comments.
In the diagram, he properly points out difficulties with time alignment (ummm, didn't somebody mention that on Page One, ahem, ahem?). And all on an order of time-scale unbelievable to home music room listeners. Ballpark echo???
But Art's comment about ONE good seat, while correct in a purely math-compulsive way*, isn't quite the whole story. The intersection of the whole areas covered by decent speaker dispersion of the two speakers is the sweet area, providing time-alignment is OK. But as you can tell from his helpful diagram (at last), the separation of speakers is key.
While Eric seems dead set against putting the mains inside the horn(s), the might be the best way to go. However, the whole process is much more agreeable, if Eric stops sticking to a needlessly low crossover point. Otherwise, the mains can be pretty small, eh, or even horns for a spectacular all-horn system. And if the mains are inside the giant horn mouth, zero problem with stereo image, need I add, eh.
Glad to see we are addressing system design like grown-ups. Did I also raise that on Page One?
Ben
*and in that way, resembles a whole lot of math-compulsive nonsense about lobes and comb filtering
In the diagram, he properly points out difficulties with time alignment (ummm, didn't somebody mention that on Page One, ahem, ahem?). And all on an order of time-scale unbelievable to home music room listeners. Ballpark echo???
But Art's comment about ONE good seat, while correct in a purely math-compulsive way*, isn't quite the whole story. The intersection of the whole areas covered by decent speaker dispersion of the two speakers is the sweet area, providing time-alignment is OK. But as you can tell from his helpful diagram (at last), the separation of speakers is key.
While Eric seems dead set against putting the mains inside the horn(s), the might be the best way to go. However, the whole process is much more agreeable, if Eric stops sticking to a needlessly low crossover point. Otherwise, the mains can be pretty small, eh, or even horns for a spectacular all-horn system. And if the mains are inside the giant horn mouth, zero problem with stereo image, need I add, eh.
Glad to see we are addressing system design like grown-ups. Did I also raise that on Page One?
Ben
*and in that way, resembles a whole lot of math-compulsive nonsense about lobes and comb filtering
Last edited:
I've posted links and shown it visually with the passive crossover design software. If you want more than that you can look it up.
You were nowhere to be found on page one of this discussion, in fact you still didn't know this was an outside installation on page 50. And while OP has shown pics of his backyard space and all the relevant features in a nice diagram in the first week, here we are a couple months later with you asking for a napkin sketch of what he provided a long time ago.
Look back at Bob's frequency response charts. 20 db wide band dips in response at various locations is not benign comb filtering. You would have to be almost completely deaf to miss that. The variations in frequency response are as bad or worse than the sound in a small room with a single sub. And you don't even have to do any math to figure that out, just look at the information that's been put right in front of your face.
In post #16, I ask about how you time-align a giant horn. In post #36. I suggest time to starting thinking how to fudge a giant horn to make it feasible like according to the available setting.
You know what they say about arguing with a pig: even if you are right, the pig loves it and you are covered with muck.
B.
You know what they say about arguing with a pig: even if you are right, the pig loves it and you are covered with muck.
B.
Last edited:
Ben,
Absolutely correct, and the same applies when you try to polish a turd (a flawed design, hope I'm not to effing cryptic 🙄), the entire process stinks .
Having experienced so many steaming piles of dung over four + decades of work in venues of every possible configuration presented by "industry leaders", I just have to let the "muck" lie where it sits- the more polish applied, the worse the outcome.
Cheers,
Art
A wide horn array like example 1 looks good to me.
bob4 showed a sim of a wide row of horn subs that looked good at 80Hz - no nulls. If that's accurate, I don't see why your example 1 would be any different.
To sort of re-hash this + adapt what I said earlier, http://www.diyaudio.com/forums/subwoofers/297953-concrete-bass-horn-design-question-67.html
If it were my project, I'd be tempted to:
- Run the inner two horns purely as subs <60Hz (ish).
- Run the outer ones as part of the mains, e.g. from 20Hz-300Hz (as high as they can go, while sounding good) - so the mains don't have to be so big.*
- Stuff the mains inside the horn mouths.
As a long-term thing, if you give the horns a flat bottom (slight tilt for drainage), you could build a very short railway into them, so it would only take a shove to roll (even very heavy) mains in and out.
*Note: I hacked your example 1 illustration (without consideration for scale) to show how you might get a bit of toe-in + wider horizontal coverage. You'd probably want to walk around & sketch out your site to work out for yourself what horizontal pattern you'd want.
**Note: in the attached displacement plot, you'll see that a nominally 30Hz horn has quite low excursion above 60Hz. That is: the heavy work (where you would want all drivers operating) is in the bottom octave.
This also looks like an argument for making the horns huge. If the cutoff of the horn is low, relative to the tone being played, excursion is small.
**Note: in the attached displacement plot, you'll see that a nominally 30Hz horn has quite low excursion above 60Hz. That is: the heavy work (where you would want all drivers operating) is in the bottom octave.
This also looks like an argument for making the horns huge. If the cutoff of the horn is low, relative to the tone being played, excursion is small.
Attachments
You are arraying these horns so it may be time to look at them geometrically. Example 2, each of the four has straight side walls. At all frequencies when the wavefront reaches the mouth it will be the shape of a scoop of ice cream on the cone with the centre of the arc being the apex of the cone just behind the driver.
Example 1, the higher frequencies will beam somewhat in the horizontal, and the primary wavefront will be more narrow than the mouth. By the time it travels until it achieves the same width, the radius will be larger and the combined wavefront will be less like a 'w', but a little more straight.
Neither is necessarily perfect, just pointing out the more obvious concerns. Also, the lower walls in example 1 are pointing toward the ground. Either curved walls top and bottom to reduce this at the higher frequencies, or maybe better, remove the bottom half of them (the top wall could be optionally curved)
Just talking shape, not size.
Last edited:
Yes, this software is fully capable of any virtual mic position and any sound source position you like in 3 dimensions.
I don't get it, are we now arguing the fact that sound sources within 1/4 wavelength mutually sum to +6db and at distances further than 1/4 wavelength apart start to deviate from the perfect 6db summation and start to destructively interfere? This is really really basic theory and it's a road I don't want to go down. There is tons of info on this concept for anyone that cares to look for it.