I wouldn't venture to give a straight answer here, there is too much to consider.
However if you can make certain changes within a short distance, the driver may be able to feel its way into the gaps and help toward the shortfall. A small portion of a wavelength away, the wave begins to gain independence from the driver.
I'm going through something similar fitting a cone driver to a quarter space radial waveguide. I'm trying to develop the intended wavefront shape, and most of the solution so far is bordering on the upper band limit.
When something isn't right geometrically sometimes there is no absolute fix, only varying degrees of wayward energy and, the audibility of this.
There is a mainstream waveguide of an odd shape that has issues which it spreads around near the driver, and which some report is OK to listen to.
Last edited:
And you call that experimenting ? Further, you come along with conclusions, so it might seem some scientific approach but it is not.
What is foam ? Does it have which acoustic properties ? Form, geometry, density, displacement behind the driver ??? No, just foam strips
With dipoles, one must just just use narrow baffle until around 6kHz that gets too difficult. This means a 4-way speaker is necessary. Simple as that!
Dipolplus - Alles über offene Schallwände
Dipolplus - Alles über offene Schallwände
I'll just throw this question out there. I'm building a directional speaker using a woofer run up into the 2-3khz range crossed over to a waveguided tweeter. Also will be using thick synthetic fur on a large baffle.
So with a directional speaker like this does one usually listen exactly on axis, or try to optimize for listening off axis at say 20 degrees. So then the high frequencies are a bit off axis and add some room "ambience"?
Yea I agree diffraction is an issue and not easily dealt with. If one uses a large baffle and somewhat smaller woofer then the woofer can have off axis diffraction issues as low as 1-3 khz, which is more of an issue than up high. Also the directivity approach basically forces you to run the woofer into break up.
But all things considered if you have a small lively room this is the best approach.
"mad max" dipole.
So with a directional speaker like this does one usually listen exactly on axis, or try to optimize for listening off axis at say 20 degrees. So then the high frequencies are a bit off axis and add some room "ambience"?
Diffraction is a problem that has to be dealt with if you want good results, that much is obvious. Slapping a driver with no directivity control like a standard dome tweeter on a flat baffle shoebox with sharp edges and then tweaking the crossover to try to compensate for the horrible response anomalies the diffraction will cause is not a solution. Nor is offsetting the driver to try to make the on axis response look flatter. (There is still just as much diffraction as before - just "moved around" and smeared a bit differently in time. Sorry asymmetric driver positioning fans...
There's two main ways to deal with diffraction, or at least high frequency diffraction, which is most damaging kind as it affects imaging, subjective smoothness of the sound, and the smoothness of the polar response of the speaker... Driver directivity, or cabinet/baffle shape. (Curves, rounding of edges etc)
Personally I find using drivers with built in directivity MUCH easier than trying to build some very elaborate curved monstrosity of a cabinet. Typical router bit rounding of the edges of a shoebox cabinet are only marginally effective, and only at high treble frequencies doing little if anything at midrange or low treble frequencies. You have to build something massively curved and pretty unusual looking to have any real effect on diffraction down to high midrange frequencies through cabinet shape alone - something like a KEF blade.
Yea I agree diffraction is an issue and not easily dealt with. If one uses a large baffle and somewhat smaller woofer then the woofer can have off axis diffraction issues as low as 1-3 khz, which is more of an issue than up high. Also the directivity approach basically forces you to run the woofer into break up.
But all things considered if you have a small lively room this is the best approach.
"mad max" dipole.
What on earth is your problem ?And you call that experimenting ? Further, you come along with conclusions, so it might seem some scientific approach but it is not.
What is foam ? Does it have which acoustic properties ? Form, geometry, density, displacement behind the driver ??? No, just foam strips
The mention of foam damping strips is completely irrelevant to the discussion of open baffle vs closed box in my post. I mentioned it only in passing to give some context for how I ended up measuring the front panel of a normal box speaker as an open baffle.
If you genuinely want to know more about foam damping strips then I suggest you read these links, which are now 7 years old:
https://www.diyaudio.com/forums/multi-way/192215-phase-plug.html#post2640287 (post #42 onwards)
https://www.diyaudio.com/forums/full-range/185012-whizzer-intelligibility.html#post2660277 (post #29 onwards)
Even if they do cancel perfectly there will still be diffraction. That was the point that I was trying to make.
I design for 20 degrees off axis and listen with speakers well toed-in. It's not for the reason that you state since the directivity is constant and the on-axis response is the same as 20 degrees. It yields a much wider "sweet-spot" that way as I have often discussed.
And yes going with a large woofer like I do, 15", does push the limits of cone control, but some woofers can do this, like the B&C 15s. But my crossover is much lower than you suggest, more like 900 Hz.
What on earth is your problem ?
The mention of foam damping strips is completely irrelevant to the discussion of open baffle vs closed box in my post.
Agreed! Not a very polite response - far too typical around here IMO.
I don't know about others, but I designed mine to measure flat on axis (time gated / anechoic) from about 250Hz up, but I listen off axis with the speakers toed to intersect behind the listener maybe a couple of feet. Not sure what that would be as an angle, I haven't tried to work it out. Maybe 10 degrees ?
I've also tried toeing the speakers in slightly in front of the listener and that also works well - the sweet spot gets a lot larger, the phantom centre is stronger than toeing out, but it makes the room sound slightly too "dead" and compresses the image width somewhat. I'm undecided on whether I prefer toed out or toed in. I think if the speakers were spaced further apart the toe in approach would be better, but I'm very limited on where I can position the speakers due to obstacles on all sides so they're much too close to each other to create a classic equilateral stereo triangle. (Angular offset of each speaker from the centre line is probably only 20 degrees)
Because the ribbon tweeter is CD in the horizontal plane there's not much change in frequency response from on axis out to nearly 45 degrees, then it drops fairly quickly and uniformly. So at 10 degrees off axis the response is almost identical to on axis, so no reason to "optimise for listening off axis".
With an 8" midwoofer baffle diffraction becomes an issue below about 2Khz. This can be seen in the frequency response graphs I posted back in post #213 - if you compare the on axis and 30 degree off axis responses (light blue line in both graphs - ignore the red grill cloth on lines) there's very little change in the response from 2-3Khz (crossover frequency 3Khz) while there is a significant bump at about 1.2Khz off axis. I haven't investigated to confirm the cause of that bump but it's probably diffraction off the edge of the 39cm wide baffles, however that is manifesting at a relatively wide 30 degree angle, and no amount of normal rounding of the baffle would have any effect at that frequency anyway, so there is not much I can do about it with a shoebox cabinet design.
But from about 2Khz up there is very little in the way of diffraction related frequency response changes. Even I was surprised when I took some recent measurements. (I still don't have a comprehensive set of measurements of the finished design as I just don't have the time and space to set up the necessary measurement facilities especially in amongst house renovating)
Sure, but that just means you need to pick a woofer with well controlled breakup, which means no metal cones etc... it rules out a lot of drivers, but there are still some that are suitable.
You want a woofer that takes advantage of controlled cone breakup to extend the frequency response and reduce beaming but without having any nasty resonances.
This is one reason why I'm using a full range driver as a "mid woofer". The driver was designed from the beginning to optimise the high frequency response and not just be a mid woofer, and was designed to be listened to without any low pass filter to hide any resonances. It stands to reason that if it's clean enough at high frequencies to be listened to on it's own full range, it will easily be clean enough at high frequencies to cross over to a tweeter, and then the only issue is directivity matching with the tweeter at the crossover frequency.
I've performed some damping modifications to the cones of mine to further clean up and flatten out the response of an already good driver in the 2-5Khz region (and then evaluated how they sound as a full range driver) and the remaining two bumps at 2Khz and 4Khz are equalised in the low pass section of the crossover to achieve an almost perfect LR 4th order acoustic roll off for the low pass section.
The driver by itself is capable of going all the way to 15Khz so in that sense it's a "waste" to use it crossed over at 3Khz, on the other hand it gives me a driver with a great deal of good quality overlap with the tweeter, freeing me to choose a crossover frequency based on criteria other than trying to avoid cone breakup resonances.
My room is a typical small British living room - speakers positioned on the long wall which is about 4.8 metres long, and 4 metres across, with the sofa right back at the other wall - there literally is no other configuration that works in this room.
Despite the speakers being relatively directional and only sitting about 2.5 metres from them I still find the room very lively and a bit more lively than I would prefer. (I tend to prefer a drier sound with less room influence) There are no acoustic "treatments" in the room other than two fabric sofas and a carpet. There are no curtains installed at the moment, only blinds.
Last edited:
Thanks for the replies. Yes with a constant directivity the spectrum stays the same off axis but just drops in level a bit, I suppose this is one the big advantages. Listening axis might not be as important.
This is the woofer I'm using. From what I've read the cone is designed to have a neutral damped response up into breakup. I couldn't hear the "cone" at all when I listened to it the other day. Very neutral sounding, but maybe a bit less detail than a metal cone. This size matches directivity with the tweeter best.
Dayton Audio RS150P-4A 6" Reference Paper Woofer 4 Ohm
A 15" cone is monstrous in comparison, perhaps peoples definition of a "small lively room" varies. The Dayton 10" RS paper would fill a larger room nicely.
This is the woofer I'm using. From what I've read the cone is designed to have a neutral damped response up into breakup. I couldn't hear the "cone" at all when I listened to it the other day. Very neutral sounding, but maybe a bit less detail than a metal cone. This size matches directivity with the tweeter best.
Dayton Audio RS150P-4A 6" Reference Paper Woofer 4 Ohm
A 15" cone is monstrous in comparison, perhaps peoples definition of a "small lively room" varies. The Dayton 10" RS paper would fill a larger room nicely.
Most definitely, but I use such a big driver for directivity, not "filling a room". But it does "fill the room" nicely even down to 20-30 Hz which your 6" woofer is going to choke on. There is also the efficiency and the extremely low distortion, not to mention the MaxSPL that is achievable. No small woofer is going to get you those things.
I am designing for a DI in the 9-12 dB range. A 6" woofer would be lucky to get 4 dB.
I was asking
Yes, impolitely.
I like to measure at various angles and compute power from this. One axis is likely to have a smoother looking DI than the others. Call it luck if you will, designing at this level is like trying to hold too many items at once.
Right and isn't that "borrowing from Peter to pay Paul"? All this heart-rendering crying about the evil of diffraction when there's no consideration that total sound may be fine even if there are little undetectable dips in the FR in particular places in the room. As if there were any room where that weren't true even with no diffraction at all.
B.
I re-read your Peter/Paul post, I had also considered responding then..
Maybe I see this in a lower mid design compromise (but errors are not easy to fix), and maybe I see it at bass frequencies for the room itself, and it just needs to be fixed.
But I can't see it in the top half of the spectrum.. or maybe I should have said I can hear it there..
Maybe I see this in a lower mid design compromise (but errors are not easy to fix), and maybe I see it at bass frequencies for the room itself, and it just needs to be fixed.
But I can't see it in the top half of the spectrum.. or maybe I should have said I can hear it there..
Wow, if that's asking I'd hate to see if you were actually upset at me..I was asking
I find that felt on the baffle works quite well to lessen diffraction, but the aesthetics is... questionable. In my case I managed to reduce +/-2 dB variations to +/-1 dB. Sorry, no off-axis measurements.
My room and speaker setup is almost identical, just a bit smaller room (4,62*3,45 m) and speakers with wide, but controlled directivity (DI is around 6 dB from 500 Hz to 10 kHz). And thick blindout curtains No carpet though, I don't like carpets I started with no treatments except two broadband bass traps in front corners and one 60*125 cm 10 cm thick absorber directly above the sofa and then added small amounts at a time to first reflection points.
The end result is 90*150 cm patch of 75 mm thick foam on left wall, nothing on the right wall (a big window and an armchair there), 60*60 cm 50 mm foam directly behind the speakers and four 30*60 cm 50 mm foam patches on the ceiling. But in addition to absorption I also added six 60*60 cm diffusers to the ceiling and two in the center of the front wall. In the end I have 8:10 (not counting bass traps, since they're away from first or even second reflection points) of diffusion to absorption ratio. And IMO diffusion is the key for small room acoustics. It doesn't deaden the room, but makes it sound bigger and takes care of the flutter echo. I think that diffusers on the ceiling is the most important element, since this is the biggest, flattest surface of the room that is not broken up with furniture.
Each time I added another acoustic element I got an improvement in imaging. The center is really nailed down, but I also get a lot of envelopment. In some pieces of music not only do the reverb of the instruments appear from the sides and back of the room, but some instruments themselves are mixed so that they appear from the side/back, and stay there firmly throughout their frequency range. One great piece that demonstrates this is: YouTube
My room and speaker setup is almost identical, just a bit smaller room (4,62*3,45 m) and speakers with wide, but controlled directivity (DI is around 6 dB from 500 Hz to 10 kHz). And thick blindout curtains
No carpet though, I don't like carpets I started with no treatments except two broadband bass traps in front corners and one 60*125 cm 10 cm thick absorber directly above the sofa and then added small amounts at a time to first reflection points.The end result is 90*150 cm patch of 75 mm thick foam on left wall, nothing on the right wall (a big window and an armchair there), 60*60 cm 50 mm foam directly behind the speakers and four 30*60 cm 50 mm foam patches on the ceiling. But in addition to absorption I also added six 60*60 cm diffusers to the ceiling and two in the center of the front wall. In the end I have 8:10 (not counting bass traps, since they're away from first or even second reflection points) of diffusion to absorption ratio. And IMO diffusion is the key for small room acoustics. It doesn't deaden the room, but makes it sound bigger and takes care of the flutter echo. I think that diffusers on the ceiling is the most important element, since this is the biggest, flattest surface of the room that is not broken up with furniture.
Each time I added another acoustic element I got an improvement in imaging. The center is really nailed down, but I also get a lot of envelopment. In some pieces of music not only do the reverb of the instruments appear from the sides and back of the room, but some instruments themselves are mixed so that they appear from the side/back, and stay there firmly throughout their frequency range. One great piece that demonstrates this is: YouTube
Very nice to see such nice data. Thanks.
I think it is fair to say that the random improvement of 1 dB in colouration is undetectable by ear even if it is apparent to your eye on plots. While the felt clearly helps a bit, the benefits seem as big outside the diffraction band as in.
Sure nice to see your attention to room acoustics although it would be good to see measurements or A/B blind testing to support your lyrical assessments.
Pity the acoustics sub-forum is so well-hidden elsewhere at DIYaudio when it should be in Loudspeakers. The visitor stats show nobody knows it is there.
B.
I don't know if I would agree to that as a general statement. There are situations where 1 dB would be audible. But the real point is that diffraction effects on the steady-state response are not really the issue. They have a strong effect on the direct sound and hence on imaging - probably far in excess of the steady-state effect.
Sorry, I misspoke earlier - our living room is only 4.8m (including bay window) by 3.5 metres - so almost the same as yours. We do have a "high" 2.8 metre ceiling though, (1930's house) so the room volume is considerably more than a 2.4 metre ceiling would provide, and I suspect the high ceiling helps reduce the influence of the ceiling on imaging, as I don't really hear any effects from the ceiling. (Although to be fair the ribbon tweeter is relatively directional in the vertical plane - on purpose)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.