Assuming that I'm following what you're saying, it sounds as though you're taking a line source and making it behave somewhat like a point source. The wavefront that launches from the middle will not be met by an equal strength pressure and will begin to bend upwards and downwards from the center.
I'm reminded of the way that Quads trigger (comparatively) large panels in order to make them approximate a point source. No, I'm not saying it's the same, just that it sounds as though you're taking one sort of speaker and making it behave like another.
This may or may not be a good thing, but I'm not sure that it qualifies as a line source any more. Perhaps new nomenclature is required.
Grey
I'm reminded of the way that Quads trigger (comparatively) large panels in order to make them approximate a point source. No, I'm not saying it's the same, just that it sounds as though you're taking one sort of speaker and making it behave like another.
This may or may not be a good thing, but I'm not sure that it qualifies as a line source any more. Perhaps new nomenclature is required.
Grey
I already took my best swing at new nomenclature in the long post above: Quasi-not-really line source. That's the best I can do
.
You wanna talk about a line source not behaving like a line source, have you ever checked out bessel array theory?
Craziness!
(Won, here's another idea to stoke your mental furnaces regarding short arrays.)
Here's an example--a five-driver bessel line array.
(numbers show relative power and polarity for each driver in the bessel row.)
1:2:2:-2:1
It gets even crazier for a seven-driver row:
1:2:2:0:-2:2:-1
Yes, the middle 0 means no driver there, just a blank (though necessary) space. I guess you could drop a tweeter in there for coincident sound.
Don't ask me how, but you can mathematically solve for a bessel row of any lenght, even for bessel matrices.
I've never played with bessel arrays, but the claim is that they approximate a point source. I think Philips holds the patent.
I hear that coherence isn't good till you move back to 3x-5x array hight, then it sounds smooth.
.You wanna talk about a line source not behaving like a line source, have you ever checked out bessel array theory?
Craziness!
(Won, here's another idea to stoke your mental furnaces regarding short arrays.)
Here's an example--a five-driver bessel line array.
(numbers show relative power and polarity for each driver in the bessel row.)
1:2:2:-2:1
It gets even crazier for a seven-driver row:
1:2:2:0:-2:2:-1
Yes, the middle 0 means no driver there, just a blank (though necessary) space. I guess you could drop a tweeter in there for coincident sound.
Don't ask me how, but you can mathematically solve for a bessel row of any lenght, even for bessel matrices.
I've never played with bessel arrays, but the claim is that they approximate a point source. I think Philips holds the patent.
I hear that coherence isn't good till you move back to 3x-5x array hight, then it sounds smooth.
Interesting. Would you mind telling me where you found that info on Bessel Arrays, Bill? Those multiples seem pretty easy to deal with, although now they seem rather arbitrary. The 7 speaker bessel (or really, 6 with a hole) seems to be perfect for my MMMTTMMM plans. I wonder what a bessel array of order 2 looks like, and how it relates to the D'Appolito (sp?)
Thanks, Won (off to Billy Joel/Elton John concert...)
Thanks, Won (off to Billy Joel/Elton John concert...)
Two links to get you started:
http://www.prosoundweb.com/install/synaudcon/tt25_4/tt25_4_p1.shtml
http://yu-ra.tripod.com/audio/array_wiring.gif
Bill
http://www.prosoundweb.com/install/synaudcon/tt25_4/tt25_4_p1.shtml
http://yu-ra.tripod.com/audio/array_wiring.gif
Bill
Oops, I see the second link is no longer good. Sorry, I didn't turn up much more on a google search.
Won, just a word of warning about your MMMTTMMM idea:
You might run into problems with your tweeter dispersion pattern.
One of the unavoidable characteristics of line arrays is that driver interactions cause cancellation that narrows dispersion along the axis of the line. For instance, vertical line arrays tend to have wide horizontal dispersion and narrow vertical dispersion. This is just fine if your head stays at a level between the array ends. In fact, it can be a plus in that it tends to limit floor/ceiling reflections.
However, in your case you have a tweeter array that's maybe three inches tall. As frequency increases, line array driver interactions become more pronounced as overall dispersion narrows. If you line your two tweeters up vertically, the vertical dispersion will be quite pinched. String them side by side and you'll rotate that pattern 90 degrees. Neither makes for a very big sweet spot.
It may work just fine for you depending on your expectations, it's just something to be aware of.
Bill
Won, just a word of warning about your MMMTTMMM idea:
You might run into problems with your tweeter dispersion pattern.
One of the unavoidable characteristics of line arrays is that driver interactions cause cancellation that narrows dispersion along the axis of the line. For instance, vertical line arrays tend to have wide horizontal dispersion and narrow vertical dispersion. This is just fine if your head stays at a level between the array ends. In fact, it can be a plus in that it tends to limit floor/ceiling reflections.
However, in your case you have a tweeter array that's maybe three inches tall. As frequency increases, line array driver interactions become more pronounced as overall dispersion narrows. If you line your two tweeters up vertically, the vertical dispersion will be quite pinched. String them side by side and you'll rotate that pattern 90 degrees. Neither makes for a very big sweet spot.
It may work just fine for you depending on your expectations, it's just something to be aware of.
Bill
Yes, it will.
But lobing should start higher in the band. I'd planned to cross the array over as low as possible to a high efficiency horn or bullet tweeter. I don't know how great a tonal match that would be, but it would be a fun experiment. I'm intrigued at the thought of 102 dB/W system efficiency.
I poked around at SpeakerBuilder for info about Wayne J's "insane" matrix array, but all I could find was his note that he was dropping the project. Anyone know if/where I could get the full scoop?
Bill
But lobing should start higher in the band. I'd planned to cross the array over as low as possible to a high efficiency horn or bullet tweeter. I don't know how great a tonal match that would be, but it would be a fun experiment. I'm intrigued at the thought of 102 dB/W system efficiency.
I poked around at SpeakerBuilder for info about Wayne J's "insane" matrix array, but all I could find was his note that he was dropping the project. Anyone know if/where I could get the full scoop?
Bill
Hello,
Will creating a curved baffle, with the center point of the radius for each baffle being at the listener's respective ears (puting each speaker the same distance from the listener's ears at one listenting location), reduce the problems of lobing, nodes, and such, at the listening location? I realize that this is only a solution for one position in the room (maybe), but the end is satisfatory to me if I could get the efficiency of the line array without the downfalls.
Thanks,
Dave
Will creating a curved baffle, with the center point of the radius for each baffle being at the listener's respective ears (puting each speaker the same distance from the listener's ears at one listenting location), reduce the problems of lobing, nodes, and such, at the listening location? I realize that this is only a solution for one position in the room (maybe), but the end is satisfatory to me if I could get the efficiency of the line array without the downfalls.
Thanks,
Dave
Good question.
All I can say is that the safest way to go would be seperate enclosed spaces. That way driver interactions will be purely electrical and the speaker will end up behaving closer to the T/S model.
By the way, I found another good line array link:
http://www.meyersound.com/products/concertseries/m3d/line_array_theory.htm
Bill
All I can say is that the safest way to go would be seperate enclosed spaces. That way driver interactions will be purely electrical and the speaker will end up behaving closer to the T/S model.
By the way, I found another good line array link:
http://www.meyersound.com/products/concertseries/m3d/line_array_theory.htm
Bill
Dave,
I've heard that curved baffles work very well--with one big caveat: the sweet spot is focused into a tiny area defined by the curve of the baffle--sorta like a pair of virtual headphones hanging in space.
At that special little point, time coherence is excellent since all drivers are equidistant from your ears. Since the array is focused, you're also probably beating the 3dB decay rate of a perfect line source. The apparent efficiency (spl) of the array will be magnified, basically beaming its energy straight into your earholes.
Step forward, behind, or to the side of this zone, and all bets are off.
Bill
I've heard that curved baffles work very well--with one big caveat: the sweet spot is focused into a tiny area defined by the curve of the baffle--sorta like a pair of virtual headphones hanging in space.
At that special little point, time coherence is excellent since all drivers are equidistant from your ears. Since the array is focused, you're also probably beating the 3dB decay rate of a perfect line source. The apparent efficiency (spl) of the array will be magnified, basically beaming its energy straight into your earholes.
Step forward, behind, or to the side of this zone, and all bets are off.
Bill
In the 7 (really 6) speaker bessel array, two sets of speakers (the closest and furthest pair from the center) are out of phase. What if these pairs were made to share an enclosure? I suppose their response would be similar to an infinite baffle. Is the infinite baffle a good alignment for speakers of Qts of 0.47? I could probably raise the Q to 0.5-7 by playing with the amplifier output impedence.
Thanks, Won
Thanks, Won
Bill F. said:
I have had this paper "The Line-Source Loudspeaker and its Applications" by Paul H Taylor from 1963 and have finally completed conversion from paper to pdf.
Some mathematical analysis, with some very illustrative pictures of comb filtering, and some practical applications.
Line-Source paper by P Taylor
dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.