Thanks for this Juha. I think the Finn I corresponded with wasn't Backman but Salmi. Actually I did correspond with Backman but it wasn't about dipole speakers.
Option 1 is more like the Gradient Revolution than the Helsinki, but it predates Revolution by more than a decade.
Gradient Revolution loudspeaker John Atkinson March 1997 | Stereophile.com
There aren't any good pics of Revolution to show how the dipole is placed. I agree with some of Salmi's recommendations for placement but not all.
Last edited:
Personally I think there is much to be said for the OB approximations of dipoles, particularly in average sized living rooms, but this thread is about SOTA dipoles?
Wanna give us an example of a SOTA dipole?
Or if you can't, how about a precise definition of what you mean?
When you do that, we'll see if these can be good speakers
No it's not a trick question. Yes I have read every post in this thread from the beginning.Err, is that a trick question?
Strictly speaking, Option 1 and the Gradients are large open baffles and ESL63 is a large unbaffled diaphragm. All three have Fig-8 directivity over a large frequency range.
In my book, these are 'dipoles' cos they have a dipole's fig-8 directivity over a large frequency range.
Your last post said OBs are only approximations to a dipole and referred to SOTA dipoles. This suggests your definition of a SOTA dipole doesn't include the likes of the Gradient, Option 1 or ESL63 so presumably, your definition is stricter than mine and there are speakers which are closer to your SOTA definition. (Yes there are several stricter definitions of 'acoustic dipole' but I'd like to know IN PRACTICAL TERMS which one you subscribe to.)
If you tell us which speakers these are and their characteristics, eg their frequency response, directivity with frequency, do they need bass boost, positioning requirements bla bla ... or even your strict definition ... we will have a better idea of what you consider 'SOTA dipole'
I am of course most interested in whether they can be good speakers.
Charlie, the OP, was/is interested in whether a dipole can cover 20Hz-20kHz. I brought up Option 1 as an example of what has been achieved in the past. ie not theoretical pontificating but for real in real rooms.
Last edited:
I am not quite getting it anymore, but I wanted to say that large open baffle exists for a long time, but only "recently" various designers converged to as little baffle as possible, to extend the CD "figure of 8" pattern over as large as a frequency as possible.
So open baffle is not synonimous to dipole. Logical, but not for everyone.
Gradient Revolution has many versions, even active . The dipole bass unit has two 12" drivers and it can be rotated. The upper part contains a "cookie" cardidoid midrange and tweeter (SEAS coaxial custom unit) Very popular speaker here in Finland, and still in production!
"DIPOLE BASS
The bass part of Gradient Revolution loudspeaker consists of two 300mm custom made long throw woofers. They are assembled on an open baffle giving the low frequencies a "figure eight" or dipole radiation pattern similar to most panel loudspeakers. These two woofers operate from below 30Hz up to 200Hz.
An ordinary box speaker radiates in all directions at low frequencies. This has the unfortunate effect of generating standing waves in all dimensions (HxWxD). In the typical listening room, standing waves distort the sound at frequencies below 200Hz. Gradient Revolution does not excite standing waves between the floor and the ceiling, it can only generate them in the direction of the radiation pattern.
The woofer section of the Revolution can be rotated and directed in three different ways, to minimize undesirable standing waves in the listening area. This ability to "tune" out standing waves in the bass is unique. The result is a smooth articulate bass response which has correct musical pitch and dynamics."
"DIPOLE BASS
The bass part of Gradient Revolution loudspeaker consists of two 300mm custom made long throw woofers. They are assembled on an open baffle giving the low frequencies a "figure eight" or dipole radiation pattern similar to most panel loudspeakers. These two woofers operate from below 30Hz up to 200Hz.
An ordinary box speaker radiates in all directions at low frequencies. This has the unfortunate effect of generating standing waves in all dimensions (HxWxD). In the typical listening room, standing waves distort the sound at frequencies below 200Hz. Gradient Revolution does not excite standing waves between the floor and the ceiling, it can only generate them in the direction of the radiation pattern.
The woofer section of the Revolution can be rotated and directed in three different ways, to minimize undesirable standing waves in the listening area. This ability to "tune" out standing waves in the bass is unique. The result is a smooth articulate bass response which has correct musical pitch and dynamics."
OK. In that case I have some practical experience.
Doing a dipole with back to back units as in Option 1 achieves far more equal radiation front & rear and constant directivity over a large frequency range than single units in a baffle. We do this up to 5kHz in Option 1.
You need a certain consistency in your drive units which might be difficult unless you make them in house in large numbers to benefit from the quality only available from the best engineering and mass production.
Of course the directivity isn't strict cosine Fig-8 at 5kHz .. but neither is ESL63.
Salmi's single unit will depart from your ideal well below 1kHz and so witll Linkwitz's RIP if it is also single unit.
Erik is right about constant directivity and this was one of our goals. OB with single units aren't CD enough.
Somewhere in this thread or the one that spawned it I suggested that, but it was dismissed due to the distance between the sources
kgrlee, can you please tell about Option 1 and 2 construction details? Looking at photos I thought that mid-tweeter box is regular monopole 2-way, but obviousy it had opposing drivers in diff. polarity? Nice to see you in this thread, sir!
I haven't heard Gradient Helsinki or Revolution, sad to say, only 1.3 and some 2-ways. It would be fun to measure these extensively with different positioning, also near a wall. Many owners will never sell them, but some sell them right away. Obviously dipole bass in a small room is a real challenge... Here some measurements of Revolution.
My take on this challenge is here, bass is sealed, LR2 at 180Hz like in Gradient 1.3. More info by clicking my signature!
I haven't heard Gradient Helsinki or Revolution, sad to say, only 1.3 and some 2-ways. It would be fun to measure these extensively with different positioning, also near a wall. Many owners will never sell them, but some sell them right away. Obviously dipole bass in a small room is a real challenge... Here some measurements of Revolution.
My take on this challenge is here, bass is sealed, LR2 at 180Hz like in Gradient 1.3. More info by clicking my signature!
Attachments
Last edited:
kgrlee said:
I'm pleased to confirm your original suggestion works & works well in real life
I'm sorry I forgot your contribution. Mea maxima culpa. As penance, I've re-read the whole thread
I'd forgotten what an onerous task this was ... wading through seas of liquid BS myth to find the tiny gems of truth.
I couldn't find any posts which dismissed your suggestion. I used back to back units instead of a totem pole cos I was concerned with vertical directivity ... though this meant the distance between 'sources' was greater. Directivity starts taking on 2nd order spherical harmonics earlier instead of staying pure 1st order cosine Fig-8 But it is still symmetrical.
Last edited:
This is what I was referring to
shame it seems to have petered out a bit
I had a bit of a reread earlier
shame it seems to have petered out a bitI just went back and looked at the Option 1 speaker that you mentioned HERE, kgrlee. Hmmm, that is very interesting indeed.
At first I really could not understand why someone was yet again making the mistake of putting a dipole right up against a wall... but what I seem to see now is a dipole "at" the wall and rotated 90 degrees so that the dipole null is pointing at the (front) wall and the "on axis" position is parallel to the wall. Is that correct?
Also, in the image, if the left side of the speaker is placed against the wall, the wall would become part of the baffle. This would increase the baffle size at low frequencies for free. For the higher frequency drivers, you would want them on a narrower section of baffle, and out and away from (the baffle not extending to) the wall. Is that the case with those speakers? I can't quite tell from the pics.
A narrow baffle, oriented perpendicular to the wall but not next to it, would not illuminate the wall all that much in the direction parallel to the plane of the baffle, because that is where the null is located. As you move to, say, 45 degrees there is sound power being radiated and it will illuminate the wall, except farther away from the loudspeaker.
By positioning the speakers far apart from each other, you might be able to sit relatively close to the wall and still enjoy dipole sound, with reflections from other walls filling in like the front wall reflection usually does with the traditional way of setting up a dipole speaker.
Well, that is my guess of how this kind of system would work in a room...
At first I really could not understand why someone was yet again making the mistake of putting a dipole right up against a wall... but what I seem to see now is a dipole "at" the wall and rotated 90 degrees so that the dipole null is pointing at the (front) wall and the "on axis" position is parallel to the wall. Is that correct?
Also, in the image, if the left side of the speaker is placed against the wall, the wall would become part of the baffle. This would increase the baffle size at low frequencies for free. For the higher frequency drivers, you would want them on a narrower section of baffle, and out and away from (the baffle not extending to) the wall. Is that the case with those speakers? I can't quite tell from the pics.
A narrow baffle, oriented perpendicular to the wall but not next to it, would not illuminate the wall all that much in the direction parallel to the plane of the baffle, because that is where the null is located. As you move to, say, 45 degrees there is sound power being radiated and it will illuminate the wall, except farther away from the loudspeaker.
By positioning the speakers far apart from each other, you might be able to sit relatively close to the wall and still enjoy dipole sound, with reflections from other walls filling in like the front wall reflection usually does with the traditional way of setting up a dipole speaker.
Well, that is my guess of how this kind of system would work in a room...
Well done Charlie! You have it almost 100% correct.
Mid & Treble are in the little rotatable 'box' at the top, set a bit away from the main 'baffle'.
My only quibble is with
The shaded area is the listening area. It's quite scary to stand in the middle with your eyes shut and walk forward towards the 'back' wall. The stereo soundfield is plausible all the way until you open your eyes and find you are at the wall.
One aspect of this Room Integration might be useful to more conventional dipoles.
If you place Option 1 at odd fractions of the room width from the side walls, you find the response in the listening area smooths out considerably. That's 1/3 as in the 2nd pic or 1/5 in smaller rooms. Left & right can be different. The effect is quite precise. You almost hear the speakers 'click' into place when you move them towards these optimum positions.
Juha, I suspect you have seen this in your experiments with placement without realising it. It's probably why REG & JA measure Revolution to have such a flat in-room response.
What is happening is that the 'resonances' due to the dipole and its reflections in the side walls are exactly interlaced with the resonances of the room in that direction. Room resonances in the other 2 directions are 'not excited' cos they are in the dipole null.
There is no such serendipituous relation with omni speakers in rooms.
I observed this before realising what was happening and pre10ing 2 doo da obfuscating maths in support
Mid & Treble are in the little rotatable 'box' at the top, set a bit away from the main 'baffle'.
My only quibble is with
This may be the first time a speaker system has been designed considering the room it will operate in right from the start. This speaker NEEDS the room to work properly and the front & side walls are part of the sound.
The shaded area is the listening area. It's quite scary to stand in the middle with your eyes shut and walk forward towards the 'back' wall. The stereo soundfield is plausible all the way until you open your eyes and find you are at the wall.
One aspect of this Room Integration might be useful to more conventional dipoles.
If you place Option 1 at odd fractions of the room width from the side walls, you find the response in the listening area smooths out considerably. That's 1/3 as in the 2nd pic or 1/5 in smaller rooms. Left & right can be different. The effect is quite precise. You almost hear the speakers 'click' into place when you move them towards these optimum positions.
Juha, I suspect you have seen this in your experiments with placement without realising it. It's probably why REG & JA measure Revolution to have such a flat in-room response.
What is happening is that the 'resonances' due to the dipole and its reflections in the side walls are exactly interlaced with the resonances of the room in that direction. Room resonances in the other 2 directions are 'not excited' cos they are in the dipole null.
There is no such serendipituous relation with omni speakers in rooms.
I observed this before realising what was happening and pre10ing 2 doo da obfuscating maths in support
No. Both bass & mid units fire 'parallel' to the wall behind the speakers. They point at the side walls.
You are never 'on-axis' to the speakers unless you move directly between them at the wall.
Thank you guys for explaining this. I see one caveat, because of lack of direct sound, stereo imaging at "spot" suffers or is even lost! Personally I might use parallel to wall dipole only for bass, crossed around 200Hz, like Salmi of Gradient.
I decided to use monopole bass because I thought that they are 1) cheaper 2) smaller 3) require less power and eq and 4) the Gradient 1.2 and 1.3 had them
My monopole downfiring bass units are 3m apart and room width is 8,5m. length 4,5m, height 3m(tilted). With such separation bass gets diffused and modes attenuated too, to some degree. My room has wide openings too, and has modes practically only below 100Hz. Spot response for L and R bass are very different <100Hz, and so is when L+R are playing usinsono. I have done many measurements with different settings and mic placement.
One nice test is to play sine wave of a modal ferquency and start walking/creeping around in the room. Pretty mild modes when two speakers playing. With dipole bass it is like four drivers playing! Remembering Todd Welti et al. multiple subwoofer studies.
I decided to use monopole bass because I thought that they are 1) cheaper 2) smaller 3) require less power and eq and 4) the Gradient 1.2 and 1.3 had them
My monopole downfiring bass units are 3m apart and room width is 8,5m. length 4,5m, height 3m(tilted). With such separation bass gets diffused and modes attenuated too, to some degree. My room has wide openings too, and has modes practically only below 100Hz. Spot response for L and R bass are very different <100Hz, and so is when L+R are playing usinsono. I have done many measurements with different settings and mic placement.
One nice test is to play sine wave of a modal ferquency and start walking/creeping around in the room. Pretty mild modes when two speakers playing. With dipole bass it is like four drivers playing! Remembering Todd Welti et al. multiple subwoofer studies.
Last edited:
Hi kgrlee,
Indeed great to have your contribution here! >30 years ago open baffle speakers must have been quite a thing.
Just want to add some links and small comments regarding some SOTA dipole speakers, as I think you missed these.
You are talking about the ESL63, but this wide panel sort of speaker, as well as the Wharfedale SFB3 Open Baffle are still being done, but not what is considered SOTA in this thread. Some speakers I consider SOTA.
4 way LX521 by Linkwitz – commercial speaker
LX521 Description
4 way Nao Note – commercial speaker
NaO Note II RS Details
Various 3 ways by Daudio.nl – commercial speaker
Daudio - dipole – design – audio
3 way Totem of tone (+ sealed subwoofer) – well documented DIY project.
The Totem of Tone, an active 3-way dipole and active subwoofer
As you see, all share a minimal baffle to achieve constant directivity pattern over a wide frequency range. Achieving a more equal front and rear radiation is not done equally in these designs. Back to back units is done for the LX521 tweeter, the AMTs in the daudio speakers and the planar drivers in the “totem of tone”. Gerrit Boers, author of the totem of tone, showed some measurements with single ad back-to-back units in this thread, and their outcomes were far from the excellent results of the B&G NEO10 planar and his own tweeter (results in his thread) – reason he used these units.
Therefore, there is recent effort in achieving equal radiation and some excellent results, which were probably not possible 30 years ago.
Two pictures of my own attempt. It looks a lot like the totem of tone but with cheaper units and electronics.
best regards, Erik
Indeed great to have your contribution here! >30 years ago open baffle speakers must have been quite a thing.
Just want to add some links and small comments regarding some SOTA dipole speakers, as I think you missed these.
You are talking about the ESL63, but this wide panel sort of speaker, as well as the Wharfedale SFB3 Open Baffle are still being done, but not what is considered SOTA in this thread. Some speakers I consider SOTA.
4 way LX521 by Linkwitz – commercial speaker
LX521 Description
4 way Nao Note – commercial speaker
NaO Note II RS Details
Various 3 ways by Daudio.nl – commercial speaker
Daudio - dipole – design – audio
3 way Totem of tone (+ sealed subwoofer) – well documented DIY project.
The Totem of Tone, an active 3-way dipole and active subwoofer
As you see, all share a minimal baffle to achieve constant directivity pattern over a wide frequency range. Achieving a more equal front and rear radiation is not done equally in these designs. Back to back units is done for the LX521 tweeter, the AMTs in the daudio speakers and the planar drivers in the “totem of tone”. Gerrit Boers, author of the totem of tone, showed some measurements with single ad back-to-back units in this thread, and their outcomes were far from the excellent results of the B&G NEO10 planar and his own tweeter (results in his thread) – reason he used these units.
Therefore, there is recent effort in achieving equal radiation and some excellent results, which were probably not possible 30 years ago.
Two pictures of my own attempt. It looks a lot like the totem of tone but with cheaper units and electronics.
best regards, Erik
Attachments
Last edited: