Same. Except not that ambiguous. In my room front wall (diffuse) reflections substantially enhance the sense of space and "naturalness", with no compromise of "imaging", at least with most of my orchestral recordings. With "studio" (rock) recordings there is also a "difference", but since "naturalness" is not at issue it's more a matter of "taste" regarding the overall presentation. I have had both dipoles (panel and point source) and directional monopoles (of comparable "tonal balance") in the room, and *vastly* prefer the sound of the dipoles, for most recordings.
So in one sense I agree with Earl . . . it's *not* "ambiguous" . . . dipoles are clearly better . . .
Draw a typical room layout with dipoles in typical spots and angles. If you know anything about image models the reflections of the speakers off of any surface are the same as the visual reflections as seen via a mirror. The key issue is how strong the radiation is from each reflected image speaker at the radiation angle as you see it in the mirror.
I sketched this out earlier and you can totally kill the back wall bounce with a dipole, as long as you are willing to toe in the system by about 75 degrees. If the frontal response is still good at 45 degrees off axis then that could be pretty good and might also send the main rear lobe off to the sides after a double bounce. Its worth trying.
The frequency range of interest is totally a function of spacing to the back wall, but I would guess that most dipole installations have issues in the middle hundreds. A cardioid mid bass sounds like a good idea but a dipole midbass preserves the pattern in the frequency range you where you don't want it. A dipole may help (reduce) the floor bounce if tall enough. It would, in fact, be equivalent to a line array of the same length. The problem is that the floor bounce angle is pretty shallow so achieving useful directivity is difficult.
David S.
I deaden the back wall, because it makes the room seem bigger. I also have surround speakers on the back wall, for use with (multi-channel) movies. My dipoles are about five feet from the (diffusive) front wall, toed in about 30 degrees, and have no "issues in the middle hundreds". Cello playing friends (acutely sensitive in that range) commonly describe the sound as either "the best" or "the most natural" they've heard.
I'd have to suggest that your "guess" (and/or the theories behind it) needs some adjustment, because reality is . . . different . . .
This isn't true of dipoles in all cases. If you take a large electrostatic (the Sound Labs for an example I am familiar with) they can be scaled to essentially be the height of the room. They are faceted and cover an angle of either 45 deg or 22.5 deg. The create a line source located behind the speaker (follow the dispersion angle of the speaker backward to where the points intersect.
The speakers need to be 4.5 ft + from the rear wall toed in (~22 deg in the case of the speakers with the 45 deg dispersion pattern). Diffusion for both the front wall and rear wall works well. Handle the side walls as you would for a speaker with good controlled dispersion (like the GedLee's for example) There is little vertical dispersion when the panels run from the floor to ceiling, so little to no floor and ceiling early reflections.
The speakers need to be 4.5 ft + from the rear wall toed in (~22 deg in the case of the speakers with the 45 deg dispersion pattern). Diffusion for both the front wall and rear wall works well. Handle the side walls as you would for a speaker with good controlled dispersion (like the GedLee's for example) There is little vertical dispersion when the panels run from the floor to ceiling, so little to no floor and ceiling early reflections.
Acoustics would be so easy if it were all that simple. Unfortunately its not. An arc does not ever act like a line source. The two things are quite different.
You should check with Duke and his observations of the big Sound Labs since he also has high regard for your speakers (he is licensing the OSWG or was a dealer?).
Energy would be the key term here, with the most energetic sound waves being lower in frequency. The higher up you go in the spectrum the less energy. Personally, I don't think that there's that big a difference in sound characteristics between a narrow baffled monopole and any of the dynamic dipoles that I've built. The last commercial speaker that I owned was the Infinity PFr tower that was designed by Laurie Fincham, and while it was most definitely a monopole, it radiated a lot of energy rearward due to its very narrow baffle and didn't stand out as being diametrically opposed in sound to anything that I've built.
I'm not on a mission to convert people to dipoles; rather, I think their weaknesses / foibles are overblown, especially at frequencies above 500 - 1k.
Re: KLH 9's - a true head in a vice speaker... not what I'm talking about.
agreed
And certainly dipoles have some weakness as does every speaker design but they are more than overblown here.
In the right setup, the figure 8 dispersion pattern absolutely decreases the incidence of early reflections from the sidewalls. Also in the case of a properly positioned dipole, the rear output is very near the 10ms delay time which we have discussed is not quite and early reflection any more as well as being attenuated significantly due to the increased path length and semi reflective surface.
Add to this that treatments on the front wall (i.e. behind speaker) can also make a big difference in reducing the impact of cancellation from the rear waves.
If they can be properly positioned, I would much rather have a dipole with good sound coming out the back than a monopole that radiates unbalanced tonal output from all sides. In my experience, the impact of the side and rear output of a monopole can have quite an adverse affect on the power output.
I grant of course that a good CD design and with good box construction can alleviate these inherent problems of the monopole design.
Last edited:
Here are 2 MP3 files. Both are pink noise with a reflection added. The first is a delayed (11ms) reflection, inverted and added at full strength. The second is similar but the reflection is reduced 5 dB, approximately what the extra distance would reduce the reflection level (depends on your distance from the speakers, of course).
THESE ARE NOT ZIP FILES. CHANGE THE ZIP TO MP3 TO PLAY (had to cheat the system to get them to upload).
The reflection is turned on and off every 2 seconds (first 2 seconds without).
Might be nice on Cello.
David S.
...did they noise shape the delay to account for differences in energy vs frequency... reflection angles...dispersion? If not, what does it prove? How does it relate to a real loudspeaker in a real room?
I agree that these ugly cancellation notches encumber most installations. But I see this not just from dipoles but monopoles too.
I think most dipole owners prefer their speakers way out in the middle of the room if they have the space, an attempt to reduce reflections from rear radiation.
But I also find that most monopoles have significant rear reflection and floor bounce problems. Even the directional ones - they just don't have any control at that range. Like you've said, the floor bounce notch is usually in the lower midrange, ~150Hz or so. Often times the rear wall notch is around there too.
I like using multiple sound sources in the 100-200Hz range, stacked vertically. Even just two helps a great deal. I don't like arrays, because I prefer point sources above 200Hz or so. But in the floor bounce region, I think having a couple sound sources stacked vertically works wonders. Blend the two up to 200Hz or so, then rolloff the bottom one to achive point source higher up. That's just how I like it.
Another approach I've always liked it tightly grouping the sound sources in the corner, so none of the boundaries produces a reflection. Radiation from a trihedral corner is a perfect solution in my opinion, because it prevents these troublesome reflections. Of course, the lowest frequencies must be dealt with separately to mitigate room modes. But from the Schroeder frequency up, the constant directivity cornerhorn approach has always been appealing to me.
As far as I can see i proves nothing, and doesn't relate to a real loudspeaker in *my* room at all. In my room there is no simple specular reflection at 11 ms . . . that path encounters a diffusive bookcase. But there are also a large number of diffuse front wall reflections from other angles and at other delays, and all the "second bounce" back wave reflections off the ceiling and side walls . . . so the reflected sound is *temporally" diffuse . . . it does not produce any significant comb filtering, and the "first bounce" (one or two reflection) delay(s) extend smoothly from 10 or 11 ms to perhaps 20 ms. Both before and interwoven with those reflections are the side wall and ceiling reflections of the front wave, and soon after come the third and fourth bounce reflections from around the room.
In reality nothing like simply adding two signals with an 11 ms delay actually occurs, so the "example" tells us nothing about loudspeakers in rooms.
Last edited:
there is no problem with floor reflection when a flooder is used
either there is no problem with front wall reflection when Beveridge placement is used
best,
graaf
They?
Remember that the absorption coefficient of drywall is less tha 0.1 at 250Hz and above. A true dipole will be sending as much energy backwards as forwards. The only reduction of level of the delayed rear reflection is from toe-in angle and the extra path distance. I think the 5 dB reduction is quite realistic. If you have a specific case, let me know the details and I'll create a simulation.
The point is that the rear firing reflection is clearly audible and not very pleasant. Sure, that wall reflection can be treated but why pick a dispersion pattern where unwanted energy has to be treated when you could start with a better polar pattern? (And if people think a bookcase provides diffusion for low frequencies they are fooling themselves.)
Could some dipole advocate give me a valid technical arguement for their speaker's superiority?
David
That starting assumption is incorrect. Our hearing have evolved to expect reflections, hence the more uniform the room is illuminated, the more natural our perception of the sound.
I have omni, dipoles, and CD speakers and clearly the ability of these transducers to dissapear is exactly in that order.
Hence the walls should never be absorbed. Diffused perhaps.
Listen to the noise samples. Those are the reflections. Studies show that our hearing is very poor at separating out such reflections that come on or near the median plane. If they came from the sides they might be ignored.
Numerous studies, numerous researchers, common conclusion: lateral reflections = spaciousness, frontal reflections = coloration.
David S.
Those tables may serve to visualize distances and
frequencies.
But to have a balanced excitation of the room below
Schroeder frequency at the listening seat or area, there
is the whole room with its boundaries and
speaker/listener positions to be taken into account.
I don't want to discuss about 'how a dipole sounds' with
respect to LF, but point to the fact that the excitation
of room modes can be influenced by toeing the dipole ...
a dipole subwoofer could be turned rather freely, and
the excitation of modes in the room changes, which is
also true for a cardioid due to its directivity.
The monopole can only be scrolled in the room, to change
the modal pattern excited.
There are many occasions, especially in small to mid sized
rooms - having low modal density at LF - where you can
find rather balanced solutions using a dipole, where a
monopole could only excite the room in a balanced way
from positions not suitable for stereo listening.
Of course if you have dedicated subwoofers,
things change a lot.
With just stereo speakers it is imporant IMO which
dispersion pattern the speaker exhibits when entering
the transitional range above Schroeder frequency and the
midrange region.
To change that pattern in a discontinuous manner
somewhere at higher frequencies is not a good idea.
A consistent pattern is what is needed.
Siegfried Linkwitz as well as John Kreskovsky extended
the dipole pattern to the upper frequency limit with
their recent designs.
If you start with a dipole pattern, you should stick with it
and maybe introduce an increasing DI at higher frequencies.
But the same can be said with respect to other patterns.
A cardioid pattern should be implemented from the
transitional range of the room upwards to be consistent.
The consistency of the radiation pattern is what
contributes to "homogeneity" of a speaker in a given room.
Having a consistent pattern also gives clear advice for
room treatment and speaker placement.
Surely with a dipole the "back radiation" has to be
accounted for, some proposals have already been made
in here, like directing it to the sides, diffusing, delaying
and broadband absorption.
But the "back radiation" of a small "monopole box" speaker
places a problem of equal dimension, even worsened by
the radiation pattern being inconsistent.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.