I agree with you. If a loudspeaker is placed beyond a foot but within a few feet of any boundary, then a self-intereference notch similar to the floor bounce notch results.
I like extending the distributed sound source approach up into the lower midrange. The multisub approach works great up to about 100Hz, but then you have to worry about localization. However, putting sound sources closer to the mains allows a similar approach to be used up closer to the Schroeder frequency. So you can use multisubs down low, and a 2.5-way or overlapping woofer/mid to help up high in the modal region.
I first noticed this with array speakers. They have no floor bounce. I prefer constant directivity point sources above the Schroeder frequency, but below that arrays have their advantages. That's what the whole multisub concept is about. It's a low frequency array.
I like extending the distributed sound source approach up into the lower midrange. The multisub approach works great up to about 100Hz, but then you have to worry about localization. However, putting sound sources closer to the mains allows a similar approach to be used up closer to the Schroeder frequency. So you can use multisubs down low, and a 2.5-way or overlapping woofer/mid to help up high in the modal region.
I first noticed this with array speakers. They have no floor bounce. I prefer constant directivity point sources above the Schroeder frequency, but below that arrays have their advantages. That's what the whole multisub concept is about. It's a low frequency array.
Wayne
A little misleading or I don't agree.
I use my subs as high as 180 Hz in some cases (which covers the modal region completely) and localization is not an issue. I would not take a sub that is behind you that high, but one that is in the frontal half, preferably behind the mains can easily go that high and it does help to smooth some of the 100-200 Hz dip that is so common.
I would not consider distributed subs to be "a low frequency array" in any strict sense, other than any use of multiple sources is an "array". But a line of woofers is not the same thing as distributed woofers.
A little misleading or I don't agree.
I use my subs as high as 180 Hz in some cases (which covers the modal region completely) and localization is not an issue. I would not take a sub that is behind you that high, but one that is in the frontal half, preferably behind the mains can easily go that high and it does help to smooth some of the 100-200 Hz dip that is so common.
I would not consider distributed subs to be "a low frequency array" in any strict sense, other than any use of multiple sources is an "array". But a line of woofers is not the same thing as distributed woofers.
markus76 said:
Slew rate and HF response are the same thing, so the answer is 1 / 180 Hz I guess. And yes, this is pretty high and I don't always do this and try not to in general, All I am saying is that I have not found it to be a problem. In my system one sub goes to about 120 Hz the rest are below 100 Hz. But my room is very well damped and the Schroeder Freq. would be quite low.
Markus - GET IT RIGHT!! 
Just a second order LP in the plate amps.
One of the key distinctions in what I recommend and the previous researchers is the use of overlaped mains and subs. This is unique but I find (and Wayne seems to agree) that this works well. It adds more "subs" to the mix, while perhaps requiring a little more capabilities from the mains. In my systems this is never an issue, in others it will be.
Just a second order LP in the plate amps.
One of the key distinctions in what I recommend and the previous researchers is the use of overlaped mains and subs. This is unique but I find (and Wayne seems to agree) that this works well. It adds more "subs" to the mix, while perhaps requiring a little more capabilities from the mains. In my systems this is never an issue, in others it will be.
gedlee said:
Without knowing much about sub, 12db is what I would prefer and expect to be best, unfortunately not many of those around, and when, they are mostly low quality
Theres a company, API or something, making subs with "cascaded" individual adjustable filters, which I find interesting
I dont understand why nobody makes more adjustable sub filters, which shouldnt be too hard
To me it seems like those who make subs and plateamps really doesnt know much about speakers, crossovers or have even the simplest basic knowledge about how it functions
I think most plate amps have the basic 12db variable filter, a phase switch and a volume control. I haven't seen many plate amps that don't have those features. For those who can work with a soldering iron and don't want to use plate amps there are some excellent kits on the market. One example:
http://jaycar.com.au/productView.as...d2=&pageNumber=&priceMin=&priceMax=&SUBCATID=
there are others as well ESP, Rod Elliot's site has some excellent bass controller projects.
col.
http://jaycar.com.au/productView.as...d2=&pageNumber=&priceMin=&priceMax=&SUBCATID=
there are others as well ESP, Rod Elliot's site has some excellent bass controller projects.
col.
This is a more general waveguide question.
A waveguide has a cut-in frequency where constant directivity begins. Let’s assume our driver and waveguide combination has a flat power response within our bandwidth of interest. In the lower frequencies before cut-in begins, wouldn’t the narrowing directivity cause a rise in the on-axis frequency response? So, would it be ideal to design for a slightly rising power response with frequency?
If the above is correct thinking, here is my real question. Is there a correlation between how much a waveguide’s directivity narrows before cut-in that would allow us to accurately predict how much the on-axis response would be boosted? For example, we have a waveguide where cut-in begins at 1.2KHz and we are using this waveguide down to 600Hz. Let’s say we have a directivity of 120 degrees at 600Hz and it narrows to 60 degrees at cut-in. Is there a way to predict how much the frequencies before cut-in will be boosted? Does a halving of directivity over a one octave range result in a 6dB boost at 600Hz in our example? The goal of doing this would be to get a nice flat frequency response without resorting to complicated crossovers, filters or attenuation schemes.
Rgs, JLH
A waveguide has a cut-in frequency where constant directivity begins. Let’s assume our driver and waveguide combination has a flat power response within our bandwidth of interest. In the lower frequencies before cut-in begins, wouldn’t the narrowing directivity cause a rise in the on-axis frequency response? So, would it be ideal to design for a slightly rising power response with frequency?
If the above is correct thinking, here is my real question. Is there a correlation between how much a waveguide’s directivity narrows before cut-in that would allow us to accurately predict how much the on-axis response would be boosted? For example, we have a waveguide where cut-in begins at 1.2KHz and we are using this waveguide down to 600Hz. Let’s say we have a directivity of 120 degrees at 600Hz and it narrows to 60 degrees at cut-in. Is there a way to predict how much the frequencies before cut-in will be boosted? Does a halving of directivity over a one octave range result in a 6dB boost at 600Hz in our example? The goal of doing this would be to get a nice flat frequency response without resorting to complicated crossovers, filters or attenuation schemes.
Rgs, JLH
JLH
I am not sure that your assumption is correct. The waveguide narrowing is a function of its width and hence length. It does seem logical to assume that at the lower frequencies where the directivity is changing with frequency that the relationship between axial SPL and the power response would also have to change. But I suspect that it might be more complicated than you suggest.
I really don't know the answer to your question because I have never done this and its too complicated a situation to guess at. The thing is that the drivers that I use mostly die below the point where the directivity narrows. And also rememebr that as the frequency goes lower, the directivity narrows then widens again as the waveguide becomes more like a small piston. This relatonship is complex and strongly dependent on the mouth size. There is a lot going on in this region. Clearly nothing that couldn't be modeled, and SPEAK may even give you some insights - it should be fairly accurate as the frequency goes down, its going to be less accurate at the higher frequencies when HOM and mouth reflections become an issue.
I am not sure that your assumption is correct. The waveguide narrowing is a function of its width and hence length. It does seem logical to assume that at the lower frequencies where the directivity is changing with frequency that the relationship between axial SPL and the power response would also have to change. But I suspect that it might be more complicated than you suggest.
I really don't know the answer to your question because I have never done this and its too complicated a situation to guess at. The thing is that the drivers that I use mostly die below the point where the directivity narrows. And also rememebr that as the frequency goes lower, the directivity narrows then widens again as the waveguide becomes more like a small piston. This relatonship is complex and strongly dependent on the mouth size. There is a lot going on in this region. Clearly nothing that couldn't be modeled, and SPEAK may even give you some insights - it should be fairly accurate as the frequency goes down, its going to be less accurate at the higher frequencies when HOM and mouth reflections become an issue.
Sensitivity is not a problem in the Summas at all. They are among the most sensitive speakers that I have seen. As to MTM, this has its problems too as the vertical polar response tends to be too narrow and then the waveguide is fairly wide. Redesigning the system just to get a narrower vertical response may have some advantages, but there are a whole lot of disadvantages too. Mostly larger size and cost, neither of which is very attractive.
"As to MTM, this has its problems too as the vertical polar response tends to be too narrow "
Wouldn't vertical coverage need to be *really* narrow for that to be an issue?
Also, why then follow with "Redesigning the system just to get a narrower vertical response may have some advantages"?
Seems like an 8" WG MTM would be an interesting alternative to a 12" WG MT, giving similarly improved efficiency and output capabilities over a single 8", where the 12" won't work because of width or depth constraints.
Wouldn't vertical coverage need to be *really* narrow for that to be an issue?
Also, why then follow with "Redesigning the system just to get a narrower vertical response may have some advantages"?
Seems like an 8" WG MTM would be an interesting alternative to a 12" WG MT, giving similarly improved efficiency and output capabilities over a single 8", where the 12" won't work because of width or depth constraints.
noah katz said:
The bigger issue is the need for a non-axi-symmetric waveguide. MTM doesn't work with a round waveguide - the distance between the two woofers is too large. With an elliptical waveguide this MIGHT work out OK, but that's certainly not guaranteed. Two 8" drivers would not have the low end that I'd look for in a good main speaker. Two 10" MIGHT, two 12's of course.
Again, without an elliptical waveguide this is all academic and an elliptical waveguide involves enough complications that its not a short term project and not guaranteed to be effective either. Research into these alternatives is what is needed and I don't have the time to do this research.
You have to look at cost versus benefit and I've said many many times that I don't consider vertical response to be a critical issue. Its NOT on par with the horizontal response - I go to great pains to get the horizontal right and would not do anything to improve the vertical response if this degraded the horizontal response.