Mabat posted frequency and excursion plots of a regular ST260 horn including a generic driver model and extended with a conical section here:
Anyway, this really seems to be a way how to extend the usable range of a driver to lower frequencies, without changing the directivity.
A regular ST260 including a generic driver model (left) and extended with a conical section (right):
Perhaps only the resonances are better damped in the real world (?).
A regular ST260 including a generic driver model (left) and extended with a conical section (right):
Perhaps only the resonances are better damped in the real world (?).
The conical extension increases sensitivity at 1000Hz by ~+6dB, and reduces peak excursion at 110dB/2m from ~0.25mm to under 0.05mm, while having little effect on the polar pattern.
Art
We can't see the effects of moving the fundamental lower. We would need for the FR to not be normalized to see it.
We can assume that the LF roll off knee moves left and the boost is lower. The Fs of driver would of the driver would be exposed eventually as well.. Sound about right?
I'm starting to think the only alternative that would appeal, to me, outside of this ~200hz on one Acoustical Axis concept...
Is an MTM....
I've been experimenting with my little desktop monitors and the REW sine generator, to see how well the phantom center is.
(pic for reference)
2 redundant sources, being on the horizontal axis should create a harder challenge vs vertical. Sweeping up and down the spectrum, and paying attention to how strong of a phantom center is created. No time alignment, 30" away from the speakers, that are 40inches apart. As I run up n down, the image moves left and right at times, due to room modes and reflections I suppose but generally its not horrible given the amount of effort going into getting it right. It makes me feel that one; if this was two redundant sources on the vertical axis, centered above n below my head, it should, sell itself, better, and that two; maybe the MTM would allow me to use higher crossovers on a close proximity monitor without experiencing the wandering of vocals down to the lower axis, as I can perceive in large 2 ways that are Not crossed over, like mine.
I am not dissatisfied with what I have, to me, it is superb. My pattern has been to second guess everything to see if I can convince myself to do something different. Last I checked I was and inch ore 2 away from 1/4wl coupling between my tweeter and woofers, and it is a benefit. I could only sense the vocals wandering to the lower axis with Barry White.
Then I think about this huge *** horn I got, and do I want it... The goal was Max Direct Sound, a large horn/waveguide and wide baffles, is the only way to get there(aside from active/passive cardioid). At ~34" wide the largest guided wave outside of baffle step is 423hz. With a MTM I could cross the horn there and keep vocals centralized, get the 18"s out of slot, impress the board with my proper THD measurements that I can't even hear anyway, configured as is.....
I'll probably do nothing lol! But I still wonder how much of a gimmick is this vibration cancelling thing with my dual opposed woofers.... I mean I have to use a dolly to move this 26hx32x32 box with dual 18's as is.... You'd think the mass alone would keep it from moving back n forth due to cone movement.
Then I go to XDir and completely confuse myself.... Why did it just occur to me that Comb filtering cares not about vertical or horizontal axis....
For example if you have two 1' dynamic tweeters spread apart at 60inches horizontally as in a stereo setup..., this is what the horizontal response looks like at 1000Hz? No 💩 ?
Is an MTM....
I've been experimenting with my little desktop monitors and the REW sine generator, to see how well the phantom center is.
(pic for reference)
2 redundant sources, being on the horizontal axis should create a harder challenge vs vertical. Sweeping up and down the spectrum, and paying attention to how strong of a phantom center is created. No time alignment, 30" away from the speakers, that are 40inches apart. As I run up n down, the image moves left and right at times, due to room modes and reflections I suppose but generally its not horrible given the amount of effort going into getting it right. It makes me feel that one; if this was two redundant sources on the vertical axis, centered above n below my head, it should, sell itself, better, and that two; maybe the MTM would allow me to use higher crossovers on a close proximity monitor without experiencing the wandering of vocals down to the lower axis, as I can perceive in large 2 ways that are Not crossed over, like mine.
I am not dissatisfied with what I have, to me, it is superb. My pattern has been to second guess everything to see if I can convince myself to do something different. Last I checked I was and inch ore 2 away from 1/4wl coupling between my tweeter and woofers, and it is a benefit. I could only sense the vocals wandering to the lower axis with Barry White.
Then I think about this huge *** horn I got, and do I want it... The goal was Max Direct Sound, a large horn/waveguide and wide baffles, is the only way to get there(aside from active/passive cardioid). At ~34" wide the largest guided wave outside of baffle step is 423hz. With a MTM I could cross the horn there and keep vocals centralized, get the 18"s out of slot, impress the board with my proper THD measurements that I can't even hear anyway, configured as is.....
I'll probably do nothing lol! But I still wonder how much of a gimmick is this vibration cancelling thing with my dual opposed woofers.... I mean I have to use a dolly to move this 26hx32x32 box with dual 18's as is.... You'd think the mass alone would keep it from moving back n forth due to cone movement.
Then I go to XDir and completely confuse myself.... Why did it just occur to me that Comb filtering cares not about vertical or horizontal axis....
For example if you have two 1' dynamic tweeters spread apart at 60inches horizontally as in a stereo setup..., this is what the horizontal response looks like at 1000Hz? No 💩 ?
Last edited:
No wonder a high degree directivity sounds so good. I was only considering the comb filtering from reflections, but a stereo pair of loudspeakers, as a redundant set of sources, causes comb filtering even in the absence of boundary reflections. The higher the directivity, the less they interact with each other, the less comb filtering.
@AllenB, Sitting nearfield, as you know I do, is better for accuracy but non the less, XDir is showing comb filtering as result of just the sources. I guess this is just apart of normal stereo listening because by the time the wave reaches the listener the beamwidth should be wide enough to have some degree of comb filtering over the whole spectrum. That would suggest that the main increase in accuracy is a result of the mitigation of boundary reflections... Things we've covered already in this thread but I never really thought about comb filtering as result of stereo itself.
You sit in the room's nearfield. That's not the same as the overlap's nearfield.
You'd have to elaborate. I sit 1meter on average. I don't know how to define that, but I bet in the higher DI of my horns I am sitting in the overlaps nearfield.
And more "claustrophobic" and headphone-like the sound is. But that doesn't bother many.
To learn where the effect blooms into it's ultimate form, consider drawing your case in the way I've suggested.
I've been in the longest session of beatification, but the design was built with the expectations of 1 meter listening. Ka is under 2 which is what Dr Geddes suggest as a max for nearfield monitoring. Listening window is wide enough for almost 2 people at that distance, thus plenty for 1 which is all I was concerned with.
"On a regular basis?"? Did you think 80db was louder at 1m than it is at 3m? ☺️
Hmmm... this?
....OK I think I figured it out... you can see it in the last polar above.
Last edited:
@GM can you hold my hand for second please.
You gave me some math to calculate desirable round overs before. I have it in my notes. I just would like to understand it better.
"round over; ~13543.8"_sec (~34400 cm_sec)/25.133 = 538.885 Hz
~13543.8/pi/538.885 = 8"
8"/2 = 4" round over"
I'd have to dig to find what the round over was for even for but you'd probably look at this and know.
You gave me some math to calculate desirable round overs before. I have it in my notes. I just would like to understand it better.
"round over; ~13543.8"_sec (~34400 cm_sec)/25.133 = 538.885 Hz
~13543.8/pi/538.885 = 8"
8"/2 = 4" round over"
I'd have to dig to find what the round over was for even for but you'd probably look at this and know.
Hmm, been awhile and my notes are MIA, but from dim memory need the baffle width to calculate its pipe end correction (technically incorrect IIRC, but don't recall ever finding a panel one), so just going to assume that the panel + end correction = 25.133" = 538.885 Hz = 8" dia divided by 2 edges = 4" round over each.
And, to skip the math, as big as fits 🙂 if there is 8" space between driver and baffle edge, the 8" radius is biggest that fits and most effective, down until the wavelength is long enough not to get affected much by the box. In other words, the bigger the baffle, the lower in frequency it affects, and the bigger roundover you need.
What I was thinking is that with such formula one could predict the max or acceptable rate of change in the tangent line for a wave guide profile.
The expansion rate should be able to be higher with large distances with less diffraction versus boundary interaction with higher frequency.