Paul McGowan - getting speaker close to wall changes tonal balance?

[abdo123]

So, do vocals and other small instruments sound as high as the speaker? I can imagine it'd work maybe in a room 15 or 20 feet wide, but forgive me for struggling to see it work in the room pictures as it looks like mono to me as far as 'width of perceived image' to me and having heard in the past, large tall ribbon types of speaker which alter the 'height' of everything, it's a bit odd to me... Forgive me here, I'm confused...

All I can say from current experience is that free space mounting can improve perceived stereo images and many ported speakers these days seem to like it. Lean toned sealed boxes do seem better able to be used close to the wall behind them as long as bass below 80 - 100Hz doesn't fall off a cliff as baby speakers do.

I'm not the biggest expert here on this either, these are all a niche within a niche in acoustics.

As for the distance between the speaker and the side walls, it doesn't really matter as as Paul described open baffle speakers don't radiate side-ways.

As for the interaction between the ceiling and the floor, well i have to get a little bit more technical here.

In a normal home listening room the sound reflected from the walls, floor and ceiling creates reflected sound images (or just reflections) in the same way an optical images would appear if the walls, floors and ceilings were made out of mirrors.

We hear both the reflected images and the direct sound of the speaker as 'sound', it takes about ~5-30ms (frequency dependent) before our brain registers the reflected images as reverberation. as the listening position moves away from the speakers the ratio of reflected images to direct sound increases in the overall 'perceived sound', the reflected images are a distortion objectively.

Now consider what happens when we place a line array loudspeaker in a room where the line array spans from floor to ceiling (see pic below). The first order reflections TRIPLE the effective length of the array. Including the second order reflections we see the height of the array increased FIVE FOLD over the actual speaker. A seven foot long array is reflected into a 35 foot array by consideration of just the first two reflections. A seven foot array of 24 speakers is transformed into a 35 foot array with 120 sound sources based on two reflections. In reality the higher order reflections are significant and the array is effectively longer with even more sound sources. In a perfectly reflective room the lines would extend to infinity giving the room an infinite reverberation time. The finite reverberation time of real rooms indicates that the reflections actually fade to inaudibility as they extend toward infinity in all directions. The combination of direct and reflected sound sources actually forms a very long array with the output progressively more tapered toward each end.

This behavior emulates a diffuse field more than monopole speakers, since the reflections are so close to the origin of sound, while a regular speaker is well a regular speaker.

The widely spaced images of the point-source playback system produce a colored frequency response that is unique to each room and varies with speaker position in the room making consistent performance almost impossible. The line-array with its standard speaker placement (on the z-axis) and well managed reflections is, in my opinion, to be a superior solution to the overall application of loudspeaker playback systems in home environments.

 

[youngho]

In a normal home listening room the sound reflected from the walls, floor and ceiling creates reflected sound images (or just reflections) in the same way an optical images would appear if the walls, floors and ceilings were made out of mirrors.

What I never understood about this model is that non-omnipolar speakers radiate differentially across the frequency spectrum in different directions, so shouldn't the optical image analogy would be more like funhouse mirrors showing distorted reflections of the original? Also, in theory, shouldn't the fusion time for the precedence effect be affected by the severity of the distortion of the reflection?

 

[abdo123]

What I never understood about this model is that non-omnipolar speakers radiate differentially across the frequency spectrum in different directions, so shouldn't the optical image analogy would be more like funhouse mirrors showing distorted reflections of the original? Also, in theory, shouldn't the fusion time for the precedence effect be affected by the severity of the distortion of the reflection?

everything you said is absolutely correct. was there anything i said that conflicted with that?

 

[youngho]

everything you said is absolutely correct. was there anything i said that conflicted with that?

Yes, I would rewrite your sentence "In a normal home listening room the sound reflected from the walls, floor and ceiling creates reflected sound images (or just reflections) in the same similar way an as optical images would appear if the walls, floors and ceilings were made out of funhouse mirrors." The sound doesn't reflect off boundaries in the same way as light off mirrors, especially since boundary reflections become increasingly non-specular at lower freqencies.

 

[abdo123]

Yes, I would rewrite your sentence "In a normal home listening room the sound reflected from the walls, floor and ceiling creates reflected sound images (or just reflections) in the same similar way an optical images would appear if the walls, floors and ceilings were made out of funhouse mirrors." The sound doesn't reflect off boundaries in the same way as light off mirrors, especially since boundary reflections become increasingly non-specular at lower freqencies.

Wait hold up. Why fun house mirrors? the reflections (because there are 6 refelective boundaries) will extend to infinity without any special or different mirrors.

 

[youngho]

Wait hold up. Why fun house mirrors? the reflections (because there are 6 refelective boundaries) will extend to infinity without any special or different mirrors.

It seems to me that with conventional speakers, the sound that bounces off the wall behind the speakers is very different in terms of the frequency spectra compared with the sound that comes directly out of the speaker, just as a typical box speaker sounds very different when you're standing in front of it versus behind it, since higher frequencies radiate increasingly forward while lower ones radiate in an increasingly omnidirectional manner. Similarly, with non-constant-directivity speakers, the reflected sound off the side wall is likely to differ from the direct sound because of off-axis variation. Most non-solely-coaxial speakers will have floor and ceiling reflections with different spectral content, due to interference between drivers, compared with the direct sound. Thus, it seems to me that first (and other order) reflections don't just interact with the direct sound through comb filtering but are themselves already distortions of the direct sound.

In my funhouse mirror analogy, it seems like the front wall reflection is like a distorted image (technically should be your backside) where the lower part of your body (the lower frequency range) is increasingly wider and the upper part (the upper frequency range) increasingly narrow. The sidewall reflections would show you at an angle but likely a little wider lower down and unevenly so further up as directivity changes unsmoothly. Etc.

 

[Newman]

I'm not the biggest expert here on this either, these are all a niche within a niche in acoustics.

As for the distance between the speaker and the side walls, it doesn't really matter as as Paul described open baffle speakers don't radiate side-ways.

As for the interaction between the ceiling and the floor, well i have to get a little bit more technical here.

In a normal home listening room the sound reflected from the walls, floor and ceiling creates reflected sound images (or just reflections) in the same way an optical images would appear if the walls, floors and ceilings were made out of mirrors.

We hear both the reflected images and the direct sound of the speaker as 'sound', it takes about ~5-30ms (frequency dependent) before our brain registers the reflected images as reverberation. as the listening position moves away from the speakers the ratio of reflected images to direct sound increases in the overall 'perceived sound', the reflected images are a distortion objectively.

Now consider what happens when we place a line array loudspeaker in a room where the line array spans from floor to ceiling (see pic below). The first order reflections TRIPLE the effective length of the array. Including the second order reflections we see the height of the array increased FIVE FOLD over the actual speaker. A seven foot long array is reflected into a 35 foot array by consideration of just the first two reflections. A seven foot array of 24 speakers is transformed into a 35 foot array with 120 sound sources based on two reflections. In reality the higher order reflections are significant and the array is effectively longer with even more sound sources. In a perfectly reflective room the lines would extend to infinity giving the room an infinite reverberation time. The finite reverberation time of real rooms indicates that the reflections actually fade to inaudibility as they extend toward infinity in all directions. The combination of direct and reflected sound sources actually forms a very long array with the output progressively more tapered toward each end.

This behavior emulates a diffuse field more than monopole speakers, since the reflections are so close to the origin of sound, while a regular speaker is well a regular speaker.

The widely spaced images of the point-source playback system produce a colored frequency response that is unique to each room and varies with speaker position in the room making consistent performance almost impossible. The line-array with its standard speaker placement (on the z-axis) and well managed reflections is, in my opinion, to be a superior solution to the overall application of loudspeaker playback systems in home environments.

You remind me of Murphy (even your diagrams!), except he goes further and puts the line in the corner and says it is even better optimised that way. (Which eliminates the dipole option BTW)

 

[JRS]

Like me. Vegetarian is NOT the same as low calorie.

Plants invented sugar and oils. Animals invented fat. Hence no fat plants, but plenty of fat plant-eating animals.

 

[youngho]

Plants invented sugar and oils. Animals invented fat. Hence no fat plants, but plenty of fat plant-eating animals.

Nuts and avocado have fat

 

[SIY]

Nuts and avocado have fat

I make Szechuan peanut noodles every Friday for a weekly special. Fat from peanuts AND sesame. I think there's a few percent in the wheat. I shudder to think about the per serving caloric content...

Our other vegan dishes use coconut oil, soy oil, corn oil, and canola as well.

 

[mhardy6647]

Plants invented sugar and oils. Animals invented fat. Hence no fat plants, but plenty of fat plant-eating animals.

Don't get the biochemists 'round here riled up.
Plants do, indeed, have all of the classes of biomolecules that... umm... nominally higher organisms have.
Including fat.



Mind you, there are two basic kinds of fats (lipids): saponifiable (triacylglycerides, which is what most folks think of as "fats") and nonsaponifiable (e.g., cholesterol)

 

[JRS]

Nuts and avocado have fat

Indeed they do, but I'm making a distinction based on plant fats as being generally healthy to consume, given the majority are either monosaturated or omega fatty acids. Another distinction is that they are used for storing raw materials needed for growth vs energy. It really is kind of an mushy distinction as the difference is whether they are semisolid or gel at room temperature. The structural differences are often not great and may result from differences in length and/or the number of double bonds. And then you have sterols which are also common to plants and animals. So yes lipids are ubiquitous in life and form the fundamental barrier between the inside and outside of cells--truly a stroke of genius for creation to make self assembling amphipathic lipids. So yes, all of life depends on fatty acids.

PS: Do biochemists really hang out here?

 

[youngho]

Indeed they do, but I'm making a distinction based on plant fats as being generally healthy to consume, given the majority are either monosaturated or omega fatty acids.

So basically nuts, seeds, avocados, and many fish on one side and non-fish animals (including shellfish and crustaceans) in the other?

 

[SIY]

PS: Do biochemists really hang out here?

A few. And some of us are biochem-adjacent.

 

[tuga]

These videos are part of a series worth watching:

 

[Newman]

@abdo123 I have been very excited about the ‘perfect line source’ in recent years, as I have been aware of the ‘five stories high virtual speaker with no floor or ceiling’ advantage. However, a couple of connected issues have made me pause to develop one into my home.

Firstly, if it is not full range above 200 Hz, then the crossover between two lines (HF and MF) side-by-side is going to be problematic. Suddenly the lobing, that is normally vertical and less of an issue psycho-acoustically, is happening horizontally where it is a significant issue. So either it needs to be full range (whoops, beaming) or coaxial drivers (whoops, HF units too far apart).

Secondly, open baffles/dipoles are not as clean to the side walls as you imply. At 60, even 75 degrees off axis, a perfect line source has plenty of output and hence plenty of interaction with side walls. It is only in the 80-90 degree area that cancellation is strong. So, unless you carefully angle the dipole so its sides point straight to the first reflection point, the advantage is not that great, and the forward sound might not be pointed to the seats optimally. Comparatively, if that reflection is not wanted for a normal speaker, just put a high quality absorptive pad over it. Plus, a normal speaker with tweeter waveguide probably has less side wall interaction than a dipole with a flat panel (unless it is a beamy-treble panel, which is bad news for other reasons).

And finally….is this all just about 2-channel again? Your words, “a superior solution to the overall application of loudspeaker playback systems in home environments”, can’t ring true to me unless we can apply it to multi-channel and even object oriented audio for music, which is already here. If it’s 2-channel, it’s not “a superior solution”, it’s inferior. So I try to imagine floor-ceiling multi-channel line sources, hang the expense, and the issues just multiply. Centre speaker? Massively impractical or compromised. Height channels? Impossible. Then the question becomes how do we integrate two or several perfect lines with numerous normal speakers, or can we even do it at all, and just how compromised will the integration be?

In summary, we start with this wonderful and ideal drawing of reflected images, and we end up with disappearing advantages and proliferating compromises. Anyhow, that is where my thinking on this has landed, so far.

cheers

 

[JRS]

A few. And some of us are biochem-adjacent.

That would be me. Medicine and physiology here, which intersect with biochem on occasion.

I'm always pleased to see fellow lifers.

 

[youngho]

And finally….is this all just about 2-channel again? Your words, “a superior solution to the overall application of loudspeaker playback systems in home environments”, can’t ring true to me unless we can apply it to multi-channel and even object oriented audio for music, which is already here. If it’s 2-channel, it’s not “a superior solution”, it’s inferior. So I try to imagine floor-ceiling multi-channel line sources, hang the expense, and the issues just multiply. Centre speaker? Massively impractical or compromised. Height channels? Impossible. Then the question becomes how do we integrate two or several perfect lines with numerous normal speakers, or can we even do it at all, and just how compromised will the integration be?

@Newman This could possibly be addressed with my inchoate idea of a modular CBT line array units that could be stacked and wall mounted, potentially going all the way from floor to ceiling, which I referenced here: https://www.audiosciencereview.com/...nsducer-cbt-speakers.12060/page-3#post-671043. For an 11.1.8 Atmos setup, the center channel and regular (90-110 degree) surround channels would go halfway up; the L/R front, wide, surround 1 (117-127 degree), and rear channels would go from floor to ceiling, but the bottom half would correspond with the traditional surround channels, and the top half would be inverted and correspond with the height channels, which would be close enough in terms of compromise to the recommended angles. When you factor in the fall-off-with-distance, then the angles become less critical. Also, as you now know, the CBT effect does not collapse with boundary proximity.

 

[mhardy6647]

PS: Do biochemists really hang out here?

umm... yeah.

 

[SIY]

I'm always pleased to see fellow lifers.

I've been paroled. These days, I'm a cook at a college town café.