advertisement


Thinking of ESLs as more similar to omni than box?

Coda II

getting there slowly
Hearing some Quads recently got my wondering whether I was thinking about them the wrong way. As a dipole, sound goes out the back as well as the front and generally I see this written about in terms of having them far enough out so that the reflections don’t lead to cancellation, or if this isn’t possible to get absorption in behind them.
But there was a comment in a thread recently along the lines of getting them to work with the room; that the reflections from the back should be viewed as a positive thing possibly. And this is where the omni bit comes in, where a speaker is intended to include the room by bouncing sound in all directions.
If this is the case then is diffusion or at least a broken up surface behind Quads a better way to go?
 
Remember the sound radiating out the back of a Dipole is out of phase with the front, whereas generally in an Omidirectional speaker, it's in phase all-round.

Yes, this is true but how the sound leaves the box is secondary to how it arrives at the listener.

For what it's worth, I've heard Quads working perfectly well in relatively small rooms with no fancy room treatment. I'm not convinced they much harder to accommodate and position than a conventional full-range loudspeaker.
 
Yes, this is true but how the sound leaves the box is secondary to how it arrives at the listener.

For what it's worth, I've heard Quads working perfectly well in relatively small rooms with no fancy room treatment. I'm not convinced they much harder to accommodate and position than a conventional full-range loudspeaker.

I've used 57s, 63s and now Magnepans in a small room (14 X 11) and it works fine.
In fact this forces listening in a semi-neafield position which combined with the smaller room lessons power requirements and you don't have to push the 'speaker so hard.

Smaller rooms also seem to lift the bass end in a way that can compliment panels, which start falling away pretty fast below 40 - 60Hz depending on the design. The LRS I'm using at the moment is rated to 50Hz but in room I can clearly hear to about 35Hz.
On the flip side a dipole excites two of the three primary room modes less than something with omni bass, which is just about everything else, so bass tends to be quite clean especially compared to the smaller ported boxes that often get matched to small rooms.
 
Using 63s on Stand & Deliver stands in a 4.2m x 3.8m x 3.25m room I experienced significant bass cancellation between 45Hz-80Hz, even with the speakers 1m out from the front wall and toed-in so that the rear of each speaker was aiming toward my GIK corner traps.

49035434557_9130caffbd_o.jpg


49723576053_06a2a36a00_o.jpg
 
Omnis and dipoles are my favourite types of speaker because, for the music I mostly listen to, they give the most realistic illusion of having the musicians performing in front of me.

I think they do this by making use of the room in a positive way, but that does mean that great care needs to be taken in positioning them. Sadly, the position that gives the best sense of realism might not be the best for domestic harmony or deep even bass and this is where judicious use of DSP or a sub can help.

I have found, at least in my room, a very subtle but for me important difference between omnis and dipoles. They both give a stunning, realistic illusion of a performer in front of one but the image of a performer with the dipole is relatively 2D, lacking in depth whilst the omni is more 3D. This is something I’d never noticed over many years of listening to various Quads and Martin Logan’s but it did become apparent after using MBLs for a while and then going back to Quads.

So, if you’re happy with your Quads don’t try an omni!
 
I've used 57s, 63s and now Magnepans in a small room (14 X 11) and it works fine.
In fact this forces listening in a semi-neafield position which combined with the smaller room lessons power requirements and you don't have to push the 'speaker so hard.

Smaller rooms also seem to lift the bass end in a way that can compliment panels, which start falling away pretty fast below 40 - 60Hz depending on the design. The LRS I'm using at the moment is rated to 50Hz but in room I can clearly hear to about 35Hz.
On the flip side a dipole excites two of the three primary room modes less than something with omni bass, which is just about everything else, so bass tends to be quite clean especially compared to the smaller ported boxes that often get matched to small rooms.

Does the semi-nearfield position mean that the relationship of direct to reflected out of phase sound is tipped sufficiently in favour of direct for the issue to be minimised and/or that there is a psychoacoustic effect going on which separates the two sound sources (can’t quite remember how this works and when the time gap is sufficiently small the brain treats it as one sound source not two, or with an echo two sources not one).
Think it must have been your 57s I heard when I brought a Quad 405-2 around for work some years back.
 
When I had ESL57s I tried them at just about every possible distance from the wall behind them, ended up preferring them a couple of metres or more away (I had a big room), with them on low coffee tables as stands, and me sitting a couple of metres away rigorously on axis vertically and horizontally. As others have said, the rear radiation is out of phase with the front and if I’m not mistaken goes up higher in frequency than a conventional box speaker so is problematic. Might’ve been interesting to try having lots of soft furniture behind them, but I’d spent all my money on hi-fi so blankets on the wall was the best I could do. Still magic on the right material.
 
When I had ESL57s I tried them at just about every possible distance from the wall behind them, ended up preferring them a couple of metres or more away (I had a big room), with them on low coffee tables as stands, and me sitting a couple of metres away rigorously on axis vertically and horizontally. As others have said, the rear radiation is out of phase with the front and if I’m not mistaken goes up higher in frequency than a conventional box speaker so is problematic. Might’ve been interesting to try having lots of soft furniture behind them, but I’d spent all my money on hi-fi so blankets on the wall was the best I could do. Still magic on the right material.
It sounds like there is a bit of a parallel with the Tannoy discussion around crossing them in front of your head and getting a really solid image but with a very definite sweet spot and having them crossed behind for a more diffuse but also more accommodating sweet spot.
 
It sounds like there is a bit of a parallel with the Tannoy discussion around crossing them in front of your head and getting a really solid image but with a very definite sweet spot and having them crossed behind for a more diffuse but also more accommodating sweet spot.
Here’s what I understand the argument to be about speaker crossing (assuming an equilateral arrangement).

If the speakers cross in front of you, then as you move your head to one side you actually move more off axis of the speaker on that side, so it’s output reduces a bit, so a central image will be pulled less in that direction than otherwise.

Otoh, you have two ears, so for your ears to be spot-on on-axis you want the speakers to cross slightly behind you, so if you buy that argument you want to sit slightly forward of the “tip”.

There is at least one good thread on gearspace with contributions from exceptionally experienced studio design engineers like Northward Audio, who says:

“You have to sit a bit forward, among other things due to spacing between your ears, but also because moving out of the tip degrades stereo a lot more than moving a bit in.

Since engineers will change position while working, the ideal spot is a bit forward at the average working position. That way the changes are very small if the engineer leans back on his chair for a bit, or leans forward to adjust an EQ etc.

Same with lateral movement.”

Full thread here

But ESL57s are a law unto themselves, the horizontal and vertical dispersion is so poor you have a sweet spot that is tiny, and you have to take a fair bit of care finding it, and staying rooted in it. I used to use a mono recording of George Malcolm playing the harpsichord. When you nail it, you know.
 
My living room has the wall behind the ETs covered with wedge acoustic foam panels to absorb some or most of the rear sound.
 
My living room has the wall behind the ETs covered with wedge acoustic foam panels to absorb some or most of the rear sound.
That’s certainly one approach, but don’t you find you lose that sense of realism? I’ve not used Eminent speakers but assuming the principles of using dipoles apply I’ve found the best approach is to make use of and tune the rear radiation by placing the speakers away from the wall behind them. A certain amount of damping can pay dividends but if too much is used the magic of dipoles is negated. As an aside, I think Peter Walker suggested that the speakers should be a third of the way down the room on the diagonal of the room. Of course he may have had a larger room them most of us!
 
That’s certainly one approach, but don’t you find you lose that sense of realism? I’ve not used Eminent speakers but assuming the principles of using dipoles apply I’ve found the best approach is to make use of and tune the rear radiation by placing the speakers away from the wall behind them. A certain amount of damping can pay dividends but if too much is used the magic of dipoles is negated. As an aside, I think Peter Walker suggested that the speakers should be a third of the way down the room on the diagonal of the room. Of course he may have had a larger room them most of us!
I think that rear wall reflection just confuses the direct sound. There is enough room reflection from direct sound as it is.

And consider that Quads have sound absorbers built-in behind the panels. ETs don't - so they have lots of rear sound.
 
I think that rear wall reflection just confuses the direct sound. There is enough room reflection from direct sound as it is.

And consider that Quads have sound absorbers built-in behind the panels. ETs don't - so they have lots of rear sound.
The 57s have felt behind the panels; later models don’t, and neither do Martin Logan’s although iirc the rear of the Quad panels has an extra layer of transparent material. Quite why I don’t know, they certainly seem to give very similar output to the rear as to the front. I don’t think the rear radiation confuses the sound as long as you have the speakers far enough from the wall behind them. The secret imo is to have the rear sound sufficiently delayed and subsequently reduced in volume that the sound does not smear the direct sound from the front. Instead it gives that natural sound that is the hallmark of a well setup dipole and has a positive effect on sound quality. It takes time to find the best speaker and listening position but is worth it.

I wonder if people who try to “remove” or over reduce the rear radiation might be better with more conventional speakers? I found that the “opposite” of dipoles, cardioid speakers, are very good for analysing the sound but don’t, to my ears, sound so much like music. Perhaps a lot depends on the type of music we listen to; I’d be happy with cardioid or box speakers for my made in the studio rock and pop recordings, but for classical which we often associate with acoustic instruments in a concert hall I find that dipoles, and particularly omnis, give me the closest approach to the original sound. All down to personal taste and perception though.
 
The 57s have felt behind the panels; later models don’t, and neither do Martin Logan’s although iirc the rear of the Quad panels has an extra layer of transparent material. Quite why I don’t know, they certainly seem to give very similar output to the rear as to the front. I don’t think the rear radiation confuses the sound as long as you have the speakers far enough from the wall behind them. The secret imo is to have the rear sound sufficiently delayed and subsequently reduced in volume that the sound does not smear the direct sound from the front. Instead it gives that natural sound that is the hallmark of a well setup dipole and has a positive effect on sound quality. It takes time to find the best speaker and listening position but is worth it.

I wonder if people who try to “remove” or over reduce the rear radiation might be better with more conventional speakers? I found that the “opposite” of dipoles, cardioid speakers, are very good for analysing the sound but don’t, to my ears, sound so much like music. Perhaps a lot depends on the type of music we listen to; I’d be happy with cardioid or box speakers for my made in the studio rock and pop recordings, but for classical which we often associate with acoustic instruments in a concert hall I find that dipoles, and particularly omnis, give me the closest approach to the original sound. All down to personal taste and perception though.
Perhaps audiophiles like different kinds of illusion.

Musicians playing in your living room vs. a window into the original event.

If it's the latter, then the acoustic contribution of your listening space should be minimized, as it will interfere with the acoustic of the original venue that was actually recorded.
 
... The secret imo is to have the rear sound sufficiently delayed and subsequently reduced in volume that the sound does not smear the direct sound from the front. Instead it gives that natural sound that is the hallmark of a well setup dipole and has a positive effect on sound quality. It takes time to find the best speaker and listening position but is worth it….
This is one of the bits I am trying to reconcile and I’m not entirely sure I grasp exactly what you are saying compared to eg. @AndyU above.
If you go for the approach of getting the dipole to work with the room, to attain a sound that comes across as ‘natural’ in that space, then does that side~step the very narrowly confined sweet-spot issue? Although you still have a ‘best listening position’ is that now a broader area?
 

This is one of the bits I am trying to reconcile and I’m not entirely sure I grasp exactly what you are saying compared to eg. @AndyU above.
If you go for the approach of getting the dipole to work with the room, to attain a sound that comes across as ‘natural’ in that space, then does that side~step the very narrowly confined sweet-spot issue? Although you still have a ‘best listening position’ is that now a broader area?

Different things.

Dipoles radiate front and back - by definition. The lateral sound distribution depends on the dipole design.

Original Quads were "beamy" locking you into a tight "sweet spot." In practice, they were not all that limiting.

Subsequent Quads were nearly ideal, as they approximated a point source. They are the only speaker leveraging the phased array radar technology coming online at that time. If Peter Walker developed these ideas independently, it would be additional testimony to his genius.

Martin-Logans use a curved panel to spread the sound around.

Magnepans and ETs use narrow panels to approximate a line source, which has very good lateral dispersion.

Bottom line, modern panel speakers don't have a "small sweet spot" problem.

But they do throw a lot of sound back, and you have to decide what to do with that.
 
Omnis and dipoles are my favorite types of speaker because, for the music I mostly listen to, they give the most realistic illusion of having the musicians performing in front of me.

I would let Isobariks sneak into that category and one thing they are very good at is this particular trick. You need the right ancillary kit to do it but a lot of speakers can't do it no matter what you drive them with.

For me a Hi-Fi has to let you forget that you are listening to a machine and feel that there are real people there somewhere. That doesn't have to mean they sound like they are in the room, they could be playing through a PA, but the essence of the instruments and voices has to be intact enough that any colouration can be ignored. From what I've heard, most loudspeakers are hopeless at this.
 
Martin-Logans use a curved panel to spread the sound around.

From my earlier investigations of M-L spec's, the angle of dispersion is 30 odd degrees, which is way narrower than Quads. Okay if you sit a fair way from the speakers, OR like near-field listening. However, I feel that panels of any sort should be as large as you can get and go into a sufficiently commodious room
 


advertisement


Back
Top