omnidirectional loudspeakers = best design available

[anmpr1]

If I stand in front of you and talk, my voice is not anything like omnidirectional, and if I play a violin or guitar or clarinet while standing there, its music is also not omnidirectional, not even close.

As we discussed earlier, it depends on how 'close' is 'close' The closer one is to an instrument the more directional it will be. But to the listener it will likely be a lot different. What I wrote was:

A musical instrument (acoustic) is more omni than directional. So you will always get a lot of ambient sound from a musical instrument--a sound 'quality' that will just not be present from a directional point source.

If I was writing now, I'd not use the 'more omni' wording, but use the word 'sufficiently'. In any case, the point is still valid. In a typical acoustical setting the reflected sound will play a large part in determining the quality of the instrument. And no one in the audience is going to be on axis to more than one or two instruments in any case. So reflected sound (which is closer to omni than strictly on axis) is going to dominate for most listeners.

Some instruments will radiate omnidirectionally at typical listening distances in spite of their directionality close to the instrument (see the Scientific American article cited above about the kettle drum--which is essentially omni several wavelengths out from the instrument).

Some instruments are almost entirely based on reflected sound, such as the horn, which throat points behind (or to the side of) the listener. A large double bass is going to sound differently off axis, but away from the instrument, in the audience, the sound radiating from it will be more omnidirectional than direct.

The link to different microphone placements in one of the recent threads here at ASR demonstrates the acoustic differences among a variety of instruments at various microphone placements.

Please note: I'm not making the argument that omnidirectional loudspeakers will recreate the microphone's image more accurately than a direct radiator. I'm not saying that at all. However, in certain acoustic spaces an omni might be the best choice, depending upon what the listener is attempting to achieve.

 

[pozz]

Yes, but time/phase shifted ambience and reverb etc. do not contain information about the off-axis response of the loudspeaker on which they will be reproduced or the acoustical properties of the reproduction room (although of course they may indirectly contain such information about the objects recorded or the acoustical properties of the recording space).

To add to this. If we're talking about accuracy, stereo inherently cannot reproduce soundfields correctly. It only records information at specific points in space. The more accurate technique, spatial sampling, which like all sampling techniques requires a minimum of two sample points per frequency, is only possible with a large number of microphones, and there are inherent limitations given that the spacing between any two have to smaller than the wavelength of the highest frequency you want to capture.

Regardless, there is still no way to arbitrarily reproduce directivity, even if it's recorded.

 

[anmpr1]

Regardless, there is still no way to arbitrarily reproduce directivity, even if it's recorded.

I believe what you wrote has to be pretty much the final word.

As I mentioned, in a smaller space I certainly prefer a forward firing loudspeaker. However, if I had twenty to forty thousand cubic feet of living room, my sonic requirements would be different. In that case, 180 degree horizontal dispersion would be what I'd want, possibly without much sound directed behind the speakers, to the floor, or ceiling. Something like the Beolab 5. In a large room where listening is likely going to be off axis, and with diverse seating arrangements, when I'm stretched out reading a book, what we usually call imaging and specific instrument placement in the mix would not be as important to me, as it is in a smaller listening space, where I generally sit between the two loudspeakers, accordingly.

I think everyone agrees that the microphone captures the direct/reflecting waves at its placement. Adding more at the loudspeaker is going to be a kludge, but might be a necessary kludge depending upon the listening space, and what one is attempting to wind up with, from an overall sonic perspective.

 

[andreasmaaan]

The sound of a narrow dispersion CD loudspeaker and a true omni are going to be very different, so how can you argue both satisfy some ill-defined fidelity requirement?

Sorry, I didn't notice this question until later.

Firstly, I don't think the requirement is "ill-defined". It's actually very clear: both on- and off-axis, the output should have the same frequency response as the input. That's all.

Secondly, I agree that a narrow-dispersion CD and an omni will sound very different, but I don't begin from the assumption that fidelity and how something sounds are the same thing.

This is, as mentioned, because IMO:

• Fidelity is the extent to which output = input.
• The recording does not contain information about the DI of the reproduction system or the acoustics of the reproduction space (i.e. this information is not present in the input).
• Yet how something sounds is highly dependent upon the DI of the reproduction system and the acoustics of the reproduction space.

Finally, a narrow-dispersion speaker with an upward-tilted DI will tend to sound very different from a wide-dispersion speaker with an upward-tilted DI. Does this mean that one or the other is higher-fidelity? The two questions (of fidelity on one hand, and of sound on the other) are best not conflated IMO.

Also please see my post #221, where I make it clear that I don't believe fidelity (that is, extent to which output = input) is necessarily the only valid goal in loudspeaker design.

 

[pozz]

I think everyone agrees that the microphone captures the direct/reflecting waves at its placement. Adding more at the loudspeaker is going to be a kludge, but might be a necessary kludge depending upon the listening space, and what one is attempting to wind up with, from an overall sonic perspective.

Though it's difficult to admit, at least from the perspective of spatial accuracy, stereo is a kludge to begin with. That said, when comparing stereo with multichannel techniques rising in complexity, including Auro 3D, ambisonics and WFS, two-channel stereo has a lot of wins (wide sweet spot, acceptable coloration, clear phantom image), not the least of which is that it's practical and cheap for just about anyone to make and listen to stereophonic music.

The most interesting thinking about omnis and dipoles in general was probably by Linkwitz. He put together auditory scene analysis and phantom imaging and came up with the term "phantom scene" to describe what these speakers do to a room. You're still strapped to your chair, and it's not like you can walk around the sound sources and phantom events, although the difference in presentation, which @MattHooper has written about before, is intense and undeniable. Really, I wouldn't make any arguments about accuracy or preference or anything of the sort; rather, the most interesting aspect of omnis is that they represent an attempt to manipulate perception. Kludgy, of course, but effective. A clear step towards understanding what it takes to influence or produce phantom effects in general.

 

[Blumlein 88]

Maybe you guys should look into J_Js perceptual soundfield reconstruction instead of going round in circles.
https://www.stereophile.com/content/wheres-real-magazine-we-see-it-february-2001

https://www.stereophile.com/news/11035/index.html

 

[oivavoi]

Maybe you guys should look into J_Js perceptual soundfield reconstruction instead of going round in circles.
https://www.stereophile.com/content/wheres-real-magazine-we-see-it-february-2001

https://www.stereophile.com/news/11035/index.html

Yap, I read about it a while back. It sounds really interesting. Would have loved for it to take off commercially!

As far as I can understand it (and I might be mistaken) the philosophy was very similar to auro 3d though: There are a set of microphones which correspond exactly in their locations to a given set of speakers. In this way you short-circuit the whole circle of confusion. With Auro 3D recordings like 2L does, they place the microphones in the same locations that the the auro 3d manual dictates that one has to place the speakers at home. This then translates and transfers the whole soundfield to the home. I've listened to a calibrated auro 3d setup, and it is without any doubt the best reproduced audio experience of my life. It was so far beyond anything else that I just had to laugh. I believe dr. Toole had the same experience the first time he heard Auro 3D, and there are few people more conversant in the language of multichannel than dr. Toole.

It seems reasonable and logical to me that as far as audio philosophies go, auro 3d really is the endgame.

But the thing is most people can't have 9, 11 or 13 speakers in their living rooms. Some people have dedicated music or movie rooms, but many don't. For many people it also becomes too expensive with such a large number of speakers. What then? This is where the discussion about omni vs conventional, wide vs narrow directivity etc comes into play. If one goes multichannel it doesn't matter that much anymore.

 

[aarons915]

Might as well kill 3 birds with one stone.

If I stand in front of you and talk, my voice is not anything like omnidirectional, and if I play a violin or guitar or clarinet while standing there, its music is also not omnidirectional, not even close. So, if an omnidirectional speaker is placed where I am standing and plays a recording of my speech and instrumental playing, and redirects all the sound omnidirectionally, that loudspeaker is performing a gross reproductive inaccuracy. An act of high unrealism.

The whole idea that omnidirectional playback is correct, or ideal, is a myth.

Here are the actual directivity vs frequency plots of voice and various instruments, © Toole:

View attachment 104251

The fact that they are not all the same means that no one loudspeaker can be perfect at reproducing all their directionalities. However, in broad and general terms, the speaker type that does best at reproducing the directionality of some of the most critical original sources such as voice, string and wind instruments, is ..... your typical front-firing loudspeaker that trends to omnidirectional in the bass!

(P.S. An omnidirectional speaker is a flat line on the zero axis at the bottom of the above chart. Pretty bad eh?)

So, if we compare the human voice directivity curve, above, to some forward-firing loudspeakers © Toole:

View attachment 104252

Again, imagine an omnidirectional loudspeaker on the above graph as a flat line on the zero axis. We can easily see that various conventional loudspeakers on the graph are far superior to an omnidirectional at reproducing something approaching the directivity of a human voice in a natural way. Furthermore, if we imagine adding the various other instruments from the top graph onto the bottom chart, we see the conventional loudspeakers even better at matching their directivities, and the omnidirectional even worse again. In fact, the omnidirectional is completely uncompetitive at the one feature that has been promoted as its natural advantage.

Thus we see that all the above arguments that idealise the directivity of omnidirectional loudspeakers, are based on a false premise, which changes all your conclusions.

Cheers

Agreed but probably with a different conclusion about the "ideal" speaker. Those directivity charts do basically prove that we don't want a pure omnidirectional speaker, a speaker like that would have flat early reflections and would likely sound bright in a room. So it does seem that a typical cone and dome is the closest to this directivity and I would take it further to say that coaxial speakers are even better since they maintain this even directivity in all directions. I've compared coaxials to very good speakers like Revels (blind) and felt the coaxials had a more clear and natural sound to them even though the Revels had the more neutral measurements. So I don't believe a speaker needs to be truly omnidirectional 360 deg around it but I do believe there is an ideal directivity and that we can't neglect the vertical plane.

 

[pozz]

Yap, I read about it a while back. It sounds really interesting. Would have loved for it to take off commercially!

As far as I can understand it (and I might be mistaken) the philosophy was very similar to auro 3d though: There are a set of microphones which correspond exactly in their locations to a given set of speakers. In this way you short-circuit the whole circle of confusion. With Auro 3D recordings like 2L does, they place the microphones in the same locations that the the auro 3d manual dictates that one has to place the speakers at home. This then translates and transfers the whole soundfield to the home. I've listened to a calibrated auro 3d setup, and it is without any doubt the best reproduced audio experience of my life. It was so far beyond anything else that I just had to laugh. I believe dr. Toole had the same experience the first time he heard Auro 3D, and there are few people more conversant in the language of multichannel than dr. Toole.

It seems reasonable and logical to me that as far as audio philosophies go, auro 3d really is the endgame.

But the thing is most people can't have 9, 11 or 13 speakers in their living rooms. Some people have dedicated music or movie rooms, but many don't. For many people it also becomes too expensive with such a large number of speakers. What then? This is where the discussion about omni vs conventional, wide vs narrow directivity etc comes into play. If one goes multichannel it doesn't matter that much anymore.

I think the main downside is the narrow sweet spot. The setup also becomes less tolerant to the weirdness of some living rooms, unless you are not using the full powers of multichannel (like content that is mostly LCR plus presence channels). I don't think omnis would work, either.

Here's a comparison by Theile and Wittek: https://www.linkwitzlab.com/Links/AURO3D_Theile-Wittek_Dec_2011.pdf

 

[Blumlein 88]

Yap, I read about it a while back. It sounds really interesting. Would have loved for it to take off commercially!

As far as I can understand it (and I might be mistaken) the philosophy was very similar to auro 3d though: There are a set of microphones which correspond exactly in their locations to a given set of speakers. In this way you short-circuit the whole circle of confusion. With Auro 3D recordings like 2L does, they place the microphones in the same locations that the the auro 3d manual dictates that one has to place the speakers at home. This then translates and transfers the whole soundfield to the home. I've listened to a calibrated auro 3d setup, and it is without any doubt the best reproduced audio experience of my life. It was so far beyond anything else that I just had to laugh. I believe dr. Toole had the same experience the first time he heard Auro 3D, and there are few people more conversant in the language of multichannel than dr. Toole.

It seems reasonable and logical to me that as far as audio philosophies go, auro 3d really is the endgame.

But the thing is most people can't have 9, 11 or 13 speakers in their living rooms. Some people have dedicated music or movie rooms, but many don't. For many people it also becomes too expensive with such a large number of speakers. What then? This is where the discussion about omni vs conventional, wide vs narrow directivity etc comes into play. If one goes multichannel it doesn't matter that much anymore.

Interesting paper on putting together a 9 channel Auro microphone rig.
https://arxiv.org/pdf/2010.05877.pdf

Auro to my understanding is more a matched channel system, while Atmos is more an object oriented system where placing objects means you have no relation between the number of channels recorded and the number used upon playback.

J_J's system was for 5 microphones in a pattern slightly larger than head sized and equally spaced around the circle. With an upward and downward shotgun microphone for height. These were mixed into the 5 channels to provide some sense of height and space. There was a simple multiplicative factor for mixing that in. So you could use 5 equally space speakers for playback.

 

[Duke]

But the main argument for why omnis can make sense is actually... about how the indirect sound becomes more similar to the direct sound. That's the heart of it. The point is still to listen to the direct sound - as with conventional speakers - but that the indirect soundfield becomes more spectrally correct. This makes the loduspeakers disappear more easily, and there are indications that it may enhance our spatial perception (re "the second look" hypothesis Indescribed before).

Agreed.

I think that one of the benefits of minimizing the discrepancy between the direct and reverberant sound is an effective reduction in the noise floor.

At some point in its decay, the spectral balance of a reflection has been degraded so much (primarily through loss of the short wavelengths) that the ear/brain system can no longer correctly classify it as a reflection. At that point, it effectively stops being "signal" and becomes "noise". Imo starting out with off-axis energy which has the same spectral balance as the direct sound will enable reflections to be correctly classified as "signal" further down into the reverberation tails.

Not that this is the only thing that matters of course, but imo it is one of them.

...a speaker like that would have flat early reflections and would likely sound bright in a room.

Yes, but imo the way you'd "voice" an omni is with the same gently downward-sloping response at all horizontal angles, including of course "on axis".

I've compared coaxials to very good speakers like Revels (blind) and felt the coaxials had a more clear and natural sound to them even though the Revels had the more neutral measurements.

Very interesting. That makes sense to me, though I don't recall reading about anyone else making the direct comparison. Which coaxials, if you don't mind? I'm thinking it must have been a very competent design to beat Revels in a blind test.

So I don't believe a speaker needs to be truly omnidirectional 360 deg around it but I do believe there is an ideal directivity and that we can't neglect the vertical plane.

I agree.

 

[krabapple]

Conceptually, the Walsh invention would seem hold the most promise, as it consists of a single driver for mids and highs. Ohm never really got it to work as advertised. Morrison is a variation of the Hegeman design. Three way (I think) with passive crossover, conventional mid and HF drivers firing up into a spherical lens.

The MBL site is not exactly forthcoming with technical information, but the large speakers appear to be multidriver with crossovers. Interestingly they claim their wave launch is 180 degrees, but the drivers appear 360. MBL literature mimics the Amar Bose idea of direct/reflective concert hall thing.

Their large speakers are shown in huge 'living rooms' and I imagine that in such large open spaces the speakers do present an impressive sonic artifact. I'm guessing that many ASR readers have smaller listening rooms, and are using the small monitor-type speakers reviewed here, in a nearfield listening position, or possibly larger floor standing speakers in a more normal sized living room.

But consider, if you have twenty or thirty thousand cubic feet of space to fill, small boxes, or even larger floor standing conventional loudspeakers just aren't going to fill the bill. So what are you supposed to do? Obviously in such an environment you are not going to employ a 'critical listening' monitor in an attempt to 'recreate' whatever the recording engineer heard at his console location. But that is not what the buyer of these is probably looking for, in such a large space, is it?

View attachment 104125

If anyone's curious to see how those MBLs up there are manufactured, the TV show "How It's Made" (to which I am addicted) did a segment on it.

It's split in two parts on youtube:

 

[krabapple]

Just following up on this (I think you edited it in afterwards):
There are actually large companies pushing omni or omni-ish designs. But they do it in the consumer entry-level products. Apple does it with the homepod (quasi-omni), Samsung with several of their wireless speakers, Harman Kardon as well with one of their speakers. Most of the commercial audio research these days goes into these kind of products, so I think there must have been some testing leading them to go in that direction. Samsung actually has one of the largest audio research facilities at the moment. They stole quite a lot of smart folks form Harman. But their speakers are not comparable to any of the high-end true omnis of course.

As to why there haven't been any commercial attempts at omnis in the high-end, I think that WAF factors have a lot to do with it. Omnis need air, and should be placed one meter from walls. That's a tough call in most homes. Similarly, horns generally benefit from being large. But almost all commercial offerings with compression drivers etc have fairly small waveguides. I can't think of any really big horn offerings from the larger companies. But it's not because it's acoustically superior (it's not), but because that's what it's possible to sell.

Here's the current system of Bjørn Kolbrek, probably the biggest authority of horn loudspeakers:
View attachment 104196
The reason those beasts are not sold by klipsch or harman is not because current klipsch waveguides are better (far from it), but because Klipsch assumes that they would not be able to sell many such speakers.

Audiophile home decor never fails to amuse.

 

[Newman]

Toole says that the dominant factor where directivity matters is spatial, not timbral. And that when you move to multichannel, even upmixed stereo multichannel, this factor fades into insignificance.

Insignificance.

So I wouldn’t be recommending the kind of importance to omnidirectional speakers, particularly “matching reverberant field spectrum to direct spectrum”, that some here have adopted.

cheers

 

[Duke]

Toole says that the dominant factor where directivity matters is spatial, not timbral. And that when you move to multichannel, even upmixed stereo multichannel, this factor fades into insignificance.

Insignificance.

Can you provide the citation? I'd rather read Toole's exact words. I'm not disagreeing with you, just want to read what he actually said. (It's human nature to "see what we are looking for", and I do it too.)

Thanks.

 

[aarons915]

Yes, but imo the way you'd "voice" an omni is with the same gently downward-sloping response at all horizontal angles, including of course "on axis".

Very interesting. That makes sense to me, though I don't recall reading about anyone else making the direct comparison. Which coaxials, if you don't mind? I'm thinking it must have been a very competent design to beat Revels in a blind test.

Yes I think that's the only way voice an omni but then I wonder if they would sound too laid back since the on-axis wouldn't be neutral?

I compared the LS50 to the M105. I did note in the review that the M105 were the closest to a point-source that I've heard that isn't a coaxial but there was still a slight difference in how clear the LS50 sounded. I've noticed a much larger difference in driver integration between KEFs and other line source speakers, especially ribbons and 3 ways.

I wrote a more detailed writeup here: (35) KEF LS50 vs Revel M105 | AVS Forum

 

[Duke]

Yes I think that's the only way voice an omni but then I wonder if they would sound too laid back since the on-axis wouldn't be neutral?

Ime the perceived spectral balance is a weighted average of the direct and reverberant sound, so I'm pretty sure a happy medium could be found.

I compared the LS50 to the M105. I did note in the review that the M105 were the closest to a point-source that I've heard that isn't a coaxial but there was still a slight difference in how clear the LS50 sounded. I've noticed a much larger difference in driver integration between KEFs and other line source speakers, especially ribbons and 3 ways.

I wrote a more detailed writeup here: (35) KEF LS50 vs Revel M105 | AVS Forum

FASCINATING!!

Imo such "anomalies" - speakers that sound better than they "should" - may well be the most interesting, because they are indications that there is still something useful left to be learned!

 

[Newman]

The main thing I learned was that some people blindlfold their girlfriends, only to disappoint.

 

[Blumlein 88]

Apparently one of the more popular Auro 3D microphone arrangements is this Schoeps ORTF 3D rig. 8 Supercardioids.

 

[MattHooper]

Agreed.

I think that one of the benefits of minimizing the discrepancy between the direct and reverberant sound is an effective reduction in the noise floor.

That's an interesting idea.

On one hand, I have found that for most speakers I've owned the more room reflections the more "live" the sound, but the more hashy it becomes.
Subjectively a sort of scrim of whitening hash seems to accompany everything. This could of course me the particular sound of my room under certain conditions. (However, as I've said, I have good flexibility in my room in terms of making it sound more live or more dead).

On the other hand, one of the distinctive characteristics I found with my MBLs, subjectively, was a "super low noise floor" sensation. They sounded very free of hash, and the low level sonic information, e.g. even just the finger padding of someone playing a classical guitar lightly, seemed more easily heard and more natural than I have heard anywhere else. That's part of what sounded so convincing with my MBLs, it didn't *sound* hyped where upper frequency transients have exaggerated clarity. Rather it was more like someone just playing a guitar in front of me and I could simply choose to listen down in to as much subtle detail as I cared to.

Dang, I'm starting to miss my MBLs now!