I'm not sure I would agree with that. See posts #904 and 912 in this thread.
We are much better off with a speaker that promotes lateral reflections and reduces back wall and floor reflections. This will increase spaciousness and reduce response effects. The dipole goes in the wrong direction.
Read Ken Kantor's article on the Magic Loudspeaker he designed for AR for background on this. http://www.kenkantor.com/publications/magic_speaker/magic_speaker.pdf
I would think that a cardioid pattern would be a good thing. You can also adjust the room instead of the speaker with heavy damping behind a dipole as Dan suggested, or cross firing CD horns as Earl recommends. In the end you are trying to reduce the early reflections, especially those in line with the system where the ear has a hard time seperating them out, and increase later lateral reflections for spaciousness.
Of course, 10 or more speakers in an anechoic chamber might be a good way to go.
David
I think that's wrong . . . dipoles combined with a diffusive front wall (and, perhaps, a dead back wall) go entirely in the *right* direction . . . front wall diffusion combined with travel path, and still-significant output at 60 degrees, gives plenty of (sufficiently delayed and thoroughly diffused) lateral reflection. And, properly done, they do it reasonably uniformly over their frequency range (no 2pi/4pi transition as is inherent in almost all "boxes").
All the flaws, none of the advantages. Difficult to implement, with negative return . . .
I have to mostly agree with Dave, but a few further points (I almost made these before but decided not to because of the flame wars that it will create). First the assumption "That's one reason dipoles sound good in real rooms" is incorrect in my opinion. They don't sound that good to me and my biggest customers are ex-dipole owners. So lets dispense with this assumption as being a given and understand that it is an opinion that is not offered with any real foundation.
Dipoles do send a lot of sound in the wrong direction and that is to the wall behind the speakers - this is wasted if it is absorbed and it is disruptive if it is not. For good spaciousness with low coloration and image degradation, the reflections should come from the sides and back of the listener with little to none from the front. So why have the back wave at all? A cardiod is a better choice, but as stated not as easy as a dipole to build and broadband not directive enough. To me, dipoles also do not have enough directivity at higher frequencies.
Dipoles do send a lot of sound in the wrong direction and that is to the wall behind the speakers - this is wasted if it is absorbed and it is disruptive if it is not. For good spaciousness with low coloration and image degradation, the reflections should come from the sides and back of the listener with little to none from the front. So why have the back wave at all? A cardiod is a better choice, but as stated not as easy as a dipole to build and broadband not directive enough. To me, dipoles also do not have enough directivity at higher frequencies.
At approximately what frequency do you consider "higher frequencies" to begin, where a dipole does not meet your definition? Are they adequate below this range in part or in whole?
Dave
You seem to be approaching spaciousness as an immersive phenomena, i.e., sound all around you. Dipole reproduction, properly set up, creates depth and not immersion. The image begins at the plane of the speaker and ends at some distance behind it... at least that's what I strive for.
Two different beasts.
My writtings make this clear in detail, but for now I look for less than 90 degrees of coverage about 500-700 Hz. Below this we do not have the same sensitivity to reflections so the need for high directivity is not as importnat.
If you look at my data for the Orion you will see two things. First its no where near constant directivity above about 4 kHz, so this doesn't count and below that it is much wider than 90 degrees being more like 120-150. Compare this to a Summa which is pretty close to being CD above 400-500 Hz and is much narrower at precisely 90 degrees.
The general usage of the term spaciousness in audio has to do with lateral reflections giving a sense of envelopment, hence putting you into a space. I'm not sure what factors contribute to depth.
I accept that there is variation of opinion on how much you might want the room to contribute, as generally determined by speaker directivity. If you like the impression that headphones give, then a deader room and higher directivity speakers may suit you. If you want more space and don't mind giving up some precision then a wider dispersion system may suit you. Two channels will always be a compromise, so people can pick a preference, just as they might when choosing near or far seats at a concert.
Beyond this I think the research clearly shows that increasing later lateral reflections is a good thing for spaciousness. Other research shows that back wall bounces are detrimental. Dipoles are not good in either regard.
David S.
I've never heard a ribbon or dynamic based dipole that had limited disperion : large planars and stats are another subject. As to where any null lies and its effects on dispersion, that's determined by placement, toe-in, driver type and size, crossover, etc..
The whole front wall bounce argument is a bit of a red herring, as has been pointed out a number of times. How far back is front?
How reflective? What materials / objects does the back wave encounter in its journey.Toole plopped some Quad 63's in place of a pair of monopoles in a test optimized for monopoles with no consideration of the Quad's radiation patterns. In this sense, he was treating them like appliances; something that a Quad owner would never do. And from this we arrive at a conclusion that wide dispersion monopoles are superior to Quads -- and then take it a step further to all dipoles?
Last edited:
What about only for midbass? I'm trying to follow you here... and with the following configuration...
1. Not below the schroeder frequency as wavelengths are too long to kill the backwave (etc.)
2. The pattern changes with frequency so keep dipoles narrowband. Stick with a waveguide up top?
3. between fs and the CD, they might be matched in directivity at the crossover like we might do with a large monopole, and the back wave could be damped if we could get enough space around them.
Dr Geddes once entertained the notion of a midbass dipole, and I see that waveguides are designed to limit/control the lateral reflections.
A dipole might control floor and ceiling reflections? I can't see much benefit to them otherwise.
Last edited:
And then there is what I recommend, which is highly directional speakers in rooms which are highly reverberant, particularly in the rear. This gives you both spaciousness and precise imaging. There is no need to compromise between the two, it just takes some judicious design.
I don't see any fish. There shouldn't be anything coming from behind the loudspeakers, this is not ambiguos.The whole front wall bounce argument is a bit of a red herring, as has been pointed out a number of times. How far back is front?
Toole plopped some Quad 63's in place of a pair of monopoles in a test optimized for monopoles with no consideration of the Quad's radiation patterns. In this sense, he was treating them like appliances; something that a Quad owner would never do. And from this we arrive at a conclusion that wide dispersion monopoles are superior to Quads -- and then take it a step further to all dipoles?
I will agree that Toole botched that test. He also points to that test as proof that high directivity is not ideal. I told him that I was not convinced that this test was that conclusive.
I did and have a design on paper. The increase in cost is simply not worth pursuing (to wit the Orions). If I were to go down that road I would do a two driver cardiod as the costs are about the same but I like the cardiod idea better.
The waveguides do limit the lateral Very Early Reflections, but no source limits the later ones. Thats purely room design. So design the speakers for the early reflections and the room for the later ones. Wait - I said that already.
I can see no benefit in disregarding the dipole or the
monopole (narrowing with frequency) approach
that globally.
It depends on placement. If you e.g. cannot avoid near
side walls, a dipole has its benefits in mitigating
(early) reflections from the same speaker's side.
The front wall (where the listener is looking at) reflections,
are worse, if the dipole is placed with the radiation axis
orthogonal to the front wall ... a placement which i
would avoid, it also is not good usually for balanced mode
excitation at LF.
To have dipoles toed in, so that the larger portion of
the back radiation hits the side wall first and "visits"
the listener via 2nd, 3rd ... reflections is advantageous
in delaying and diffusing that reflections, especially
if there is no diffusive front wall at hand.
Remember that not all listeners are willing or allowed
to treat their rooms in an offensive manner ...
At larger listening distances this may -depending on
the room's proportions - lead to crossing the axis
noticeably in front of the listener, which is no good idea
with dipoles having high directivity in the highs and are
restricted to aim at the listener directly for some brilliance.
Some "CD like " virtues in the highs are very desirable
with a dipole IMO, though not achieved by many large
panel ESL.
On the other hand panels and line arrays have less
problems with floor bounce.
The typical (non CD) 2 Way is suffering from near and
undamped side walls. It has to be placed rather freely
and also toed in to perform in many rooms.
Not so very different what you and up with in some
rooms ...
monopole (narrowing with frequency) approach
that globally.
It depends on placement. If you e.g. cannot avoid near
side walls, a dipole has its benefits in mitigating
(early) reflections from the same speaker's side.
The front wall (where the listener is looking at) reflections,
are worse, if the dipole is placed with the radiation axis
orthogonal to the front wall ... a placement which i
would avoid, it also is not good usually for balanced mode
excitation at LF.
To have dipoles toed in, so that the larger portion of
the back radiation hits the side wall first and "visits"
the listener via 2nd, 3rd ... reflections is advantageous
in delaying and diffusing that reflections, especially
if there is no diffusive front wall at hand.
Remember that not all listeners are willing or allowed
to treat their rooms in an offensive manner ...
At larger listening distances this may -depending on
the room's proportions - lead to crossing the axis
noticeably in front of the listener, which is no good idea
with dipoles having high directivity in the highs and are
restricted to aim at the listener directly for some brilliance.
Some "CD like " virtues in the highs are very desirable
with a dipole IMO, though not achieved by many large
panel ESL.
On the other hand panels and line arrays have less
problems with floor bounce.
The typical (non CD) 2 Way is suffering from near and
undamped side walls. It has to be placed rather freely
and also toed in to perform in many rooms.
Not so very different what you and up with in some
rooms ...
Last edited:
In my experience that depends on the speaker, and to some extent the room. So I find it ambiguous.
We were talking dipoles here. I've seen the polars and d.i. curves of the ESL 63 (and talked to Peter Walker about how he emperically derived the drive to the outter rings) and they are as close to an ideal dipole as is out there. They have a figure 8 pattern over a wide range and narrow less at the top than most planer speakers. The back wall reflection will be fed to a higher level than with other radiation patterns. If you go with broader rear energy via a ribbon, then you're making the back wall reflection issue worse, not better.
The whole front wall bounce argument is a bit of a red herring, as has been pointed out a number of times. How far back is front?
Most every dipole listening setup I've seen has had the speakers in front of drywall type wall. Spacing typically puts cancellation notches in the middle hundreds. You can treat the wall, but then why do you want a dipole if you intend to kill the back wave? For any back wall you propose a speaker with less rear radiation will still be better off.
Toole plopped some Quad 63's in place of a pair of monopoles in a test optimized for monopoles with no consideration of the Quad's radiation patterns. In this sense, he was treating them like appliances; something that a Quad owner would never do. And from this we arrive at a conclusion that wide dispersion monopoles are superior to Quads -- and then take it a step further to all dipoles?
I'm sure he did better than just "plop them in a test optimized for monopoles". Everything I've read shows his tests are carefully constructed. To clarify what he found in blind listening tests, listeners found the dipoles at the bottom of the group with regards to creating a realistic image (spaciousness) in single speaker tests. When tested in stereo (he typically tested in mono because he found test results more consitent that way) the ESL 63s caught up somewhat with the group.
I had KLH 9s years ago and loved some of their special qualities. Imaging was interesting in a way, much like wearing headphones from across the room. I guess they pulled you into the recording rather than bringing the musicians into the room. Your preference is a matter of taste.
Beyond the Toole test, which is on point for dipoles, are many other studies that make a good case that back wall and floor bounces are bad but delayed sidewalls bounces are good. If you want to pursue those ideals, then dipoles are a bad starting point.
David S.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.