Is baffle shape really important? What about these ideas?

JRS · Jun 3, 2023

sarumbear said:
I disagree. It’s about understanding the theory. Audio is a complicated discipline and few can understand it.

Of course one can admire the theory while respecting the limitations. Waveguides are well understood across a large range of dimensions, And so I was perplexed but not surprised when it was suggested that there is a gap between theory and practice. In other words, you can simulate the hell out of something--even with FEA and supercomputers, you still gotta build the damn thing, measure and refine.

I am in no position to argue the case either way--it strikes me that if one can model a hypersonic aircraft wing, one should be able to pin down an acoustic lens. The assessment provided suggested it just isn't so, leaving me to lament that all those PDE's I studied in school are subject to significant limitation--presumably due to overly zealous simplifying assumptions. The same problem crops up in most of physics; even the most gorgeous model of all--the so called standard model is mute on the matter of dark stuff which appears to be most of the stuff in the universe. Doesn't that strike you as a wee bit paradoxical?

I love good theory as much as the next guy. And am relieved to learn that the software appears to be very capable indeed. And I think that it would be a phenomenal service to the DIY community to be able to carve up some plastic for a modest fee in service of controlling dispersion. At present that, and access to Klippel standard msmts, is the only thing holding back even more successful design in the broader DIY community. I love most of the speakers I have built--but then again I assume from the git go that the nearest boundary is a few feet so that the impact of radiation mismatches is minimized.

What I would like to understand more completely is the role of asymmetry in listening rooms as a benefit. Often is seems an an a priori assumption that a proper listening environment is symmetric, but not too symmetric (as in avoiding the perfect cube as an example). I suspect that bilateral symmetry needs to be approximated just to avoid left vs right arguments between our ears, but wonder whether strict obedience is always in our best interests. Some of the weirder spaces I've inhabited have had a wonderful acoustic--the perfect parallelopiped LR not so much.

gab · Jun 3, 2023

JRS said:
What I would like to understand more completely is the role of asymmetry in listening rooms as a benefit. Often is seems an an a priori assumption that a proper listening environment is symmetric, but not too symmetric (as in avoiding the perfect cube as an example). I suspect that bilateral symmetry needs to be approximated just to avoid left vs right arguments between our ears, but wonder whether strict obedience is always in our best interests. Some of the weirder spaces I've inhabited have had a wonderful acoustic--the perfect parallelopiped LR not so much.

How big should a great listening room be?

This recent study concluded that, "the length-width ratio should be within 1.15–1.45 to get a reasonably good frequency spacing of the room modes" Preferred dimension ratios of small rectangular rooms As far as size goes, he points out that if the longest dimension of your room is under 8.6m...

www.audiosciencereview.com

sarumbear · Jun 3, 2023

As an exercise I asked Bard the following. Images are from Bard. Google is trying to differentiate their AI tool by offering images.

Q. Show me different levels of diffraction on different speaker enclosure shapes

A. Sure, here are some examples of different levels of diffraction on different speaker enclosure shapes:

Sphere: A spherical enclosure has the lowest level of diffraction, as the sound waves are able to radiate in all directions without being obstructed by the edges of the enclosure. This results in a smooth and even frequency response.
Cube: A cubic enclosure has a higher level of diffraction than a spherical enclosure, as the sound waves are diffracted by the edges of the enclosure. This results in a more uneven frequency response, with some frequencies being louder than others.
Rectangle: A rectangular enclosure has an even higher level of diffraction than a cubic enclosure, as the sound waves are diffracted by the edges of the enclosure in a more pronounced way. This results in a very uneven frequency response, with some frequencies being much louder than others.

It is important to note that the level of diffraction also depends on the size of the enclosure, with larger enclosures having a lower level of diffraction than smaller enclosures. This is because the wavelength of sound waves is inversely proportional to their frequency, so lower frequencies have longer wavelengths. As a result, lower frequencies are less likely to be diffracted by the edges of an enclosure.

In general, it is best to use a speaker enclosure with a smooth and even shape, such as a sphere, to minimize diffraction and achieve the best possible sound quality.

Q. Show me the math

A. Sure, here is the math behind diffraction of sound waves on different speaker enclosure shapes:

The amount of diffraction is determined by the wavelength of the sound wave and the size of the obstacle that it is diffracting around. The smaller the obstacle relative to the wavelength, the less diffraction will occur.

For example, a 100 Hz sound wave has a wavelength of about 11 feet. If a speaker enclosure is 2 feet wide, then the diffraction will be relatively small. However, a 10 kHz sound wave has a wavelength of about 1.1 inches. If a speaker enclosure is 2 feet wide, then the diffraction will be much larger.

The following formula can be used to calculate the amount of diffraction:

D = (w/λ)^2

Where:

D is the amount of diffraction
w is the width of the obstacle
λ is the wavelength of the sound wave

For example, if a speaker enclosure is 2 feet wide and the sound wave has a wavelength of 100 Hz, then the amount of diffraction would be:

D = (2/11)^2 = 0.036

This means that the diffraction would be very small and would have a negligible impact on the sound quality.

However, if the sound wave had a wavelength of 10 kHz, then the amount of diffraction would be:

D = (2/1.1)^2 = 3.6

This means that the diffraction would be much larger and would have a significant impact on the sound quality.

In general, it is best to use a speaker enclosure with a smooth and even shape, such as a sphere, to minimize diffraction and achieve the best possible sound quality. This is because a sphere has no sharp edges, which means that the sound waves will not be diffracted as much.

Duke · Jun 3, 2023

JRS said:
What I would like to understand more completely is the role of asymmetry in listening rooms as a benefit.

My understanding is that de-correlation in the reflection field is, or can be, desirable, and that lateral acoustic asymmetry can contribute to this, but that we don't want strong asymmetry in the first lateral reflections if they arrive too soon after the direct sound. My understanding is that left/right de-correlation in the reflection field contributes to a sense of spaciousness.

@Floyd Toole's current room has a wall on one side (with drapes if I understand correctly) and a large opening on the other side. The distance from the listening area to the left-side wall with the drapes is fairly far, or so it appears in photos I've seen.

Earl Geddes deliberately introduces acoustic asymmetry into his dedicated listening room; one side wall has a tapestry on it, and the other is made of large irregular rocks for diffusion. Earl's constant-directivity speakers are toed-in 45 degrees such that their first significant sidewall reflections are contralateral (off the opposite side wall), and therefore they arrive significantly later than ipsilateral (same-side-wall) reflections would.

So in Earl's room the asymmetrical first lateral reflections occur much later than the first lateral reflections normally do, and I think that is also the case in Floyd Toole's room, the distance to the side with the wall being longer than normal. I suspect that having a fairly long time delay before the first significant lateral reflections arrive avoids the image shifting that could otherwise occur with earlier-onset asymmetrical lateral reflections.

fluid · Jun 4, 2023

GM3 said:
Just thinking, would no baffle be the best baffle?

Maybe, here is a contrast of no baffle cylindrical vs sphere with a 25mm flat radiating element of constant acceleration.
There has been some progress since Olsen, but given the time and resources available he was amazing.

Deleted member 21114 · Jun 4, 2023

sarumbear said:
As an exercise I asked Bard the following...
...
D = (2/11)^2 = 0.036
...
D = (2/1.1)^2 = 3.6

tuga · Jun 4, 2023

About baffle design, edge diffraction, secondary sound sources, ... ... and their influence on the sound quality of loudspeakers

Basic article | Baffle geometry

About baffle design, edge diffraction, secondary sound sources and their influence on the reproduction quality of loudspeakers.

heissmann-acoustics.de

sarumbear · Jun 4, 2023

Marquee Smith said:

Care to describe the reason for your mark?

tuga · Jun 4, 2023

Duke said:
My understanding is that de-correlation in the reflection field is, or can be, desirable, and that lateral acoustic asymmetry can contribute to this, but that we don't want strong asymmetry in the first lateral reflections if they arrive too soon after the direct sound. My understanding is that left/right de-correlation in the reflection field contributes to a sense of spaciousness.

@Floyd Toole's current room has a wall on one side (with drapes if I understand correctly) and a large opening on the other side. The distance from the listening area to the left-side wall with the drapes is fairly far, or so it appears in photos I've seen.

Earl Geddes deliberately introduces acoustic asymmetry into his dedicated listening room; one side wall has a tapestry on it, and the other is made of large irregular rocks for diffusion. Earl's constant-directivity speakers are toed-in 45 degrees such that their first significant sidewall reflections are contralateral (off the opposite side wall), and therefore they arrive significantly later than ipsilateral (same-side-wall) reflections would.

So in Earl's room the asymmetrical first lateral reflections occur much later than the first lateral reflections normally do, and I think that is also the case in Floyd Toole's room, the distance to the side with the wall being longer than normal. I suspect that having a fairly long time delay before the first significant lateral reflections arrive avoids the image shifting that could otherwise occur with earlier-onset asymmetrical lateral reflections.

The ceiling slopes across the axis of the speakers, which not being coaxial mid-tweeter should result in some degree of imbalance. I wonder if that is why the speakers are upside down?

Duke · Jun 4, 2023

tuga said:
The ceiling slopes across the axis of the speakers, which not being coaxial mid-tweeter should result in some degree of imbalance. I wonder if that is why the speakers are upside down?

It looks to me like @Floyd Toole's big Revels are atop a cabinet (presumably so they wouldn't occupy floor space), and my guess is that positioning them upside-down was the most practical way to get the tweeter height correct.

sarumbear · Jun 4, 2023

Duke said:
It looks to me like @Floyd Toole's big Revels are atop a cabinet (presumably so they wouldn't occupy floor space), and my guess is that positioning them upside-down was the most practical way to get the tweeter height correct.

Salon2 is too tall for listening while sitting down. I tilted them down to aim ear height while sitting on my sofa.

Deleted member 21114 · Jun 4, 2023

sarumbear said:
Care to describe the reason for your mark?

Using language models to do maths is an awful idea. You don't learn to add by looking at a million (unreliable) examples of addition - you learn the rules of addition. This is demonstrated in your post.

sarumbear · Jun 4, 2023

Marquee Smith said:
Using language models to do maths is an awful idea. You don't learn to add by looking at a million (unreliable) examples of addition - you learn the rules of addition. This is demonstrated in your post.

I was not doing math. No calculations were done, is there? I asked the formula involved and Bard showed me. Do you disagree on anything on the answers?

What awful idea was demonstrated? I like to know your point of view.

Timcognito · Jun 4, 2023

sarumbear said:
In general, it is best to use a speaker enclosure with a smooth and even shape, such as a sphere, to minimize diffraction and achieve the best possible sound quality. This is because a sphere has no sharp edges, which means that the sound waves will not be diffracted as much.

Question: is a flat infinite wall better or worse than a sphere?

sarumbear · Jun 4, 2023

Timcognito said:
Question: is a flat infinite wall better or worse than a sphere?

Irrelevant question in the context of diffraction caused by enclosure edges.

Timcognito · Jun 4, 2023

sarumbear said:
Irrelevant question in the context of diffraction caused by enclosure edges.

Yes, so an infinite baffle is no baffle? I asking about the no edge case. Seems relevant, at least curious to