• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Speakers that measure well and work in small room

andreasmaaan

Major Contributor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,757
#61
I know that the trend in speaker design has been toward narrow baffles, moving the frequency point where sound becomes more directional from the speaker higher. I’m not sure what benefit this brings. I guess it brings more of the curve into a range where there is less attenuation of off axis sound?
This is a complex one. My view is that wide baffles are superior for two main reasons:
  1. they bring the point at which directivity transitions from frontal to omni down to (or closer to) a region where the room tends to dominate (both in terms of measurement and perception)
  2. they widen the frequency band between the points where the woofer(s) tend to become omnidirectional and where the enclosure does, ensuring that the transition from highly directional to omni is less abrupt than for an enclosure that is roughly the same width as the woofer(s) it houses
The other issue is edge diffraction which I don’t understand. The descriptions I’ve seen seem to describe almost as a reflection from the sound traveling across front of baffle, when it hits the edge, some bounces back causing comb filtering interference.

The modern studio monitors tend to have curved baffles, which I think is supposed to minimize the problems with diffraction.
That’s correct. Also FWIW, the main argument against wide baffles is that these diffracted waves arrive later than they do with a narrower baffle, and from a point more distant from the source. But yes, curving the baffle edges is the most effective way to minimise this diffraction. It’s also possible that you like this diffraction effect sonically - I recall for example Vance Dickason writing that he found highly diffractive baffles to sound more "lively". I’m not aware of any serious research into this, however.

I totally get what you’re saying: that my hypothesis proposed is only relevant if there actually is some kind of “family sound” for different approaches to speaker designs. Essentially that is my tentative hypothesis. That is my subjective experience.
That's interesting and yes you could be right. I think the one aspect I can't imagine this could be the case with would be your discussion of resonant enclosure walls. This just couldn't be consistent across speakers as there are so many random factors at play in determining the frequencies and and decay of such resonances.

But I do think you might be onto something when it comes to baffle dimensions and edge diffraction.

What I need to do for at least sanity check is find a home speaker that is designed in accordance with the principles espoused by Toole, and see what I think.

Do folks think any of the older Revel models would qualify? I would like to find something $1000 or less, used.
The Revel M106 qualifies and can probably be had for under $1k used. However, apart from the non-curved edges, I doubt it would sound much different from your studio monitors, as it's a similar size/width and has a similar driver complement to a small studio monitor, i.e. 6" cone woofer + 1" dome tweeter in a small constant directivity waveguide.
 

MattHooper

Senior Member
Joined
Jan 27, 2019
Messages
402
Likes
424
#62
andreasmaaan,

Your post about wider baffles echoes similar sentiments I've seen expressed elsewhere.

And it would seem to at least partially explain why, after decades of enjoying narrow baffle speakers I've "discovered" wider baffle speakers. They seem to sound a bit different. I've mentioned elsewhere that I really enjoyed the two way, wide-baffle Devore Fidelity 0/96 and0/93 speakers which do indeed seem to have a fuller, livelier sound than most other speakers I've encountered. Hard to tease out all the variables, though.
 

andreasmaaan

Major Contributor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,757
#63
andreasmaaan,

Your post about wider baffles echoes similar sentiments I've seen expressed elsewhere.

And it would seem to at least partially explain why, after decades of enjoying narrow baffle speakers I've "discovered" wider baffle speakers. They seem to sound a bit different. I've mentioned elsewhere that I really enjoyed the two way, wide-baffle Devore Fidelity 0/96 and0/93 speakers which do indeed seem to have a fuller, livelier sound than most other speakers I've encountered. Hard to tease out all the variables, though.
Yeh. I don't have any rigorous evidence that the effects of wider baffles are subjectively preferred by listeners, but the wider baffle's effects on polar response and diffraction are objectively true, and it's at least plausible that this could be subjectively important IMHO.
 
Joined
Apr 22, 2016
Messages
57
Likes
67
Location
Princeton, Texas
#64
Yeh. I don't have any rigorous evidence that the effects of wider baffles are subjectively preferred by listeners, but the wider baffle's effects on polar response and diffraction are objectively true, and it's at least plausible that this could be subjectively important IMHO.
In addition to baffle step/polar response implications, imo there are imaging implications associated with baffle width.

My understanding is that reflections/diffractions which arrive very early, within the first .68 milliseconds, are especially detrimential to imaging. This corresponds to a path length of about 9 inches, which in turn is roughly the distance around the head from one ear to the other. A reflection within this very early interval is like a false imaging cue indicating the sound is coming from somewhere to the side, and the closer to .68 milliseconds, the farther off to the side it seems to be. In practice I think the true localization cues usually dominate, but the image specificity is degraded the false localization cues in these very early reflections. After about .68 milliseconds the precedence effect kicks in and the ear/brain system ignores directional cues from reflections.

So with a narrow-baffle speaker, the false cues from edge diffraction inherently arrive early within the initial .68 millisecond window, so they indicate a relatively small incorrect angle and therefore are not very detrimental. As the baffle gets wider, the edge diffraction pushes closer and closer to .68 milliseconds, and so the false imaging cues indicate a larger incorrect angle and thus become correspondingly more detrimental. If the baffle is wide enough to give us more than .68 milliseconds (9 inches) before the sound from the edge of the driver reaches the edge of the baffle, imaging should improve.

One implication of this is, narrow-baffle speakers usually image better than wide-baffle speakers (assuming similar edge diffraction), but probably not better than VERY wide-baffle speakers... such as flush-mounted monitors in a recording studio. Another implication is that, if we can avoid or minimize edge diffraction to begin with, then baffle width won't really matter as far as imaging goes, though it will still matter for other reasons.

Diffraction is detrimental for other reasons as well, as it can be a source of harshness at high sound pressure levels. This is because it is an aberration that arrives later than the original signal, and therefore will not be masked by the original signal. The ear/brain system tends to mask an aberration that occurs at the same time as the original signal if its spectral content is included in the (considerably louder) original signal, but not if the aberration arrives at a different time.

The Snell Type A was arguably one of the most intelligent designs ever from the standpoint of minimizing diffraction, and imo its imaging and non-fatiguing sound quality were superb. https://www.audioasylum.com/message...-type-a-restoration-insulation-against-baffle
 
Joined
Dec 15, 2018
Messages
55
Likes
57
#65
That's interesting and yes you could be right. I think the one aspect I can't imagine this could be the case with would be your discussion of resonant enclosure walls. This just couldn't be consistent across speakers as there are so many random factors at play in determining the frequencies and and decay of such resonances.
.
Thanks for the perspective on baffle width very interesting.

The issue I’m noticing with modern studio monitors is they are made of very dense materials, with rounded shapes. This trend started sometime in the last 15 years.

What I find disconcerting is that I don’t hear much cabinet resonance at all. I was attributing this to the inert materials and reduction of standing waves and rectangular panels.

But you’re right, the diffraction effects are reduced to.

This brings me to my limits of understanding of speaker design.

My vague concept was that the drivers have fundamental resonance, which you try to move out of the with the crossover frequency.

But there’s also a calculation for the volume of the cabinet. I’ve never been clear whether the physical structure and any resonance characteristics are part of the equation.

My perception has always been that more “traditional” wooden box designs are generating sound from the whole cabinet, in the way a violin would.

Speakers from the 60s were sometimes sealed empty plywood boxes. I have a pair of EV “The Layton” home speakers, and they sound crazy, the box is definitely part of the sound.

One issue I’m not grasping is how the cabinet enables such a louder and more “full” from the driver.

If you have a very dense, modern cabinet shape, presumably stuffed with absorption, how does whatever impact the cabinet air volume contribute to the sound?

The studio monitors I’ve been dissing are ported, so some sound must come out of the port. Does it actually change the movement of the driver.

As an extreme example I was trying to fix some JBL Duet computer style speakers recently. As far as such speakers go, these aren’t bad.

They have a bulbous plastic housing with a proportionate big port.

I pulled the driver out and it’s this tiny full range driver that when played out of the enclosure is just a tinny, fizzy. The difference the enclosure makes borders on magical.

Anyway, always more to learn!
 
Joined
Dec 15, 2018
Messages
55
Likes
57
#66
In addition to baffle step/polar response implications, imo there are imaging implications associated with baffle width.

My understanding is that reflections/diffractions which arrive very early, within the first .68 milliseconds, are especially detrimential to imaging. This corresponds to a path length of about 9 inches, which in turn is roughly the distance around the head from one ear to the other. A reflection within this very early interval is like a false imaging cue indicating the sound is coming from somewhere to the side, and the closer to .68 milliseconds, the farther off to the side it seems to be. In practice I think the true localization cues usually dominate, but the image specificity is degraded the false localization cues in these very early reflections. After about .68 milliseconds the precedence effect kicks in and the ear/brain system ignores directional cues from reflections.

So with a narrow-baffle speaker, the false cues from edge diffraction inherently arrive early within the initial .68 millisecond window, so they indicate a relatively small incorrect angle and therefore are not very detrimental. As the baffle gets wider, the edge diffraction pushes closer and closer to .68 milliseconds, and so the false imaging cues indicate a larger incorrect angle and thus become correspondingly more detrimental. If the baffle is wide enough to give us more than .68 milliseconds (9 inches) before the sound from the edge of the driver reaches the edge of the baffle, imaging should improve.

One implication of this is, narrow-baffle speakers usually image better than wide-baffle speakers (assuming similar edge diffraction), but probably not better than VERY wide-baffle speakers... such as flush-mounted monitors in a recording studio. Another implication is that, if we can avoid or minimize edge diffraction to begin with, then baffle width won't really matter as far as imaging goes, though it will still matter for other reasons.

Diffraction is detrimental for other reasons as well, as it can be a source of harshness at high sound pressure levels. This is because it is an aberration that arrives later than the original signal, and therefore will not be masked by the original signal. The ear/brain system tends to mask an aberration that occurs at the same time as the original signal if its spectral content is included in the (considerably louder) original signal, but not if the aberration arrives at a different time.

The Snell Type A was arguably one of the most intelligent designs ever from the standpoint of minimizing diffraction, and imo its imaging and non-fatiguing sound quality were superb. https://www.audioasylum.com/message...-type-a-restoration-insulation-against-baffle
I’m sure many folks must have had the idea of putting absorption at the edge of the speaker. Since I’ve never seen this, I presume this offers no solution?

Some of the AR designs from the 80s did have a bunch of foam right on the baffle, surrounding the tweeter and mid range...
 
Joined
Apr 22, 2016
Messages
57
Likes
67
Location
Princeton, Texas
#67
I’m sure many folks must have had the idea of putting absorption at the edge of the speaker. Since I’ve never seen this, I presume this offers no solution?
I think a thick foam "halo" surrounding the front baffle would reduce edge diffraction considerably. Unfortunately it would probably reduce sales even more. I recall seeing a manufacturer demonstrating something like this at CES about fifteen years ago, and to the best of my knowledge it never ended up on a production model.

Absorption on the front baffle helps with the shorter wavelengths, but doesn't absorb the longer wavelengths very well unless it is fairly thick.

My understanding is that sound waves travelling across a theoretically sound-absorbing surface, like across a felt-covered front baffle, are attenuated very little. However where the sound waves travel INTO the absorbent material, like into the edge of a thick felt ring around a tweeter, they are absorbed where they hit the inner edge of that material. So the thickness of the felt (or foam or whatever) matters for that reason.

Imo a big round-over is probably the best solution. Using drivers which effectively "beam" enough to largely "miss" the edges of the enclosure also usefully reduces edge diffraction.

Some of the AR designs from the 80s did have a bunch of foam right on the baffle, surrounding the tweeter and mid range...
Ken Kantor's "MGC-1" design, for AR, made very aggressive use of absorbent foam for pattern control. There is a prosound company that does so today but I can't remember their name offhand.
 

andreasmaaan

Major Contributor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,757
#68
In addition to baffle step/polar response implications, imo there are imaging implications associated with baffle width.

My understanding is that reflections/diffractions which arrive very early, within the first .68 milliseconds, are especially detrimential to imaging. This corresponds to a path length of about 9 inches, which in turn is roughly the distance around the head from one ear to the other. A reflection within this very early interval is like a false imaging cue indicating the sound is coming from somewhere to the side, and the closer to .68 milliseconds, the farther off to the side it seems to be. In practice I think the true localization cues usually dominate, but the image specificity is degraded the false localization cues in these very early reflections. After about .68 milliseconds the precedence effect kicks in and the ear/brain system ignores directional cues from reflections.

So with a narrow-baffle speaker, the false cues from edge diffraction inherently arrive early within the initial .68 millisecond window, so they indicate a relatively small incorrect angle and therefore are not very detrimental. As the baffle gets wider, the edge diffraction pushes closer and closer to .68 milliseconds, and so the false imaging cues indicate a larger incorrect angle and thus become correspondingly more detrimental. If the baffle is wide enough to give us more than .68 milliseconds (9 inches) before the sound from the edge of the driver reaches the edge of the baffle, imaging should improve.

One implication of this is, narrow-baffle speakers usually image better than wide-baffle speakers (assuming similar edge diffraction), but probably not better than VERY wide-baffle speakers... such as flush-mounted monitors in a recording studio. Another implication is that, if we can avoid or minimize edge diffraction to begin with, then baffle width won't really matter as far as imaging goes, though it will still matter for other reasons.

Diffraction is detrimental for other reasons as well, as it can be a source of harshness at high sound pressure levels. This is because it is an aberration that arrives later than the original signal, and therefore will not be masked by the original signal. The ear/brain system tends to mask an aberration that occurs at the same time as the original signal if its spectral content is included in the (considerably louder) original signal, but not if the aberration arrives at a different time.

The Snell Type A was arguably one of the most intelligent designs ever from the standpoint of minimizing diffraction, and imo its imaging and non-fatiguing sound quality were superb. https://www.audioasylum.com/message...-type-a-restoration-insulation-against-baffle
I actually did mention this effect in both my previous posts ;) But what is news to me is the 9”/0.68ms threshold. Do you have a source for this pls?
 

andreasmaaan

Major Contributor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,757
#69
I think a thick foam "halo" surrounding the front baffle would reduce edge diffraction considerably. Unfortunately it would probably reduce sales even more. I recall seeing a manufacturer demonstrating something like this at CES about fifteen years ago, and to the best of my knowledge it never ended up on a production model.
All the Dunlavy models used foam absorbers around the tweeter, and I recall seeing a few other models over the years but I can’t recall which now. I agree with you that rounding over is probably a better, albeit more expensive, solution.
 

andreasmaaan

Major Contributor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,757
#70
My vague concept was that the drivers have fundamental resonance, which you try to move out of the with the crossover frequency.
True of tweeters and mids, although of course not in the case of woofers.

But there’s also a calculation for the volume of the cabinet. I’ve never been clear whether the physical structure and any resonance characteristics are part of the equation.
It depends on the type of cabinet. For closed boxes, as box volume decreases the resonance frequency increases, while for ported boxes the ratio of box volume to port length and surface area determines the resonance frequency. Ported boxes are far more sensitive to changes in box volume.

My perception has always been that more “traditional” wooden box designs are generating sound from the whole cabinet, in the way a violin would.
True, although it’s been generally agreed since the very early days that this is an undesirable effect, so it’s generally been considered a flaw rather than a feature.

The studio monitors I’ve been dissing are ported, so some sound must come out of the port. Does it actually change the movement of the driver.
Definitely. At the port resonance frequency, the driver theoretically doesn’t move at all, with all output at this frequency coming from the port. The more rigid and dense the cabinet, the closer to theoretical perfection this works.

As an extreme example I was trying to fix some JBL Duet computer style speakers recently. As far as such speakers go, these aren’t bad.

They have a bulbous plastic housing with a proportionate big port.

I pulled the driver out and it’s this tiny full range driver that when played out of the enclosure is just a tinny, fizzy. The difference the enclosure makes borders on magical.
Yep, cabinets are fundamentally important :) Also keep in mind that when you run a driver in free air there will be cancellation of the front wave by the out-of-phase back wave, which will greatly diminish low frequency performance.
 
Joined
Dec 15, 2018
Messages
55
Likes
57
#71
True of tweeters and mids, although of course not in the case of woofers.



It depends on the type of cabinet. For closed boxes, as box volume decreases the resonance frequency increases, while for ported boxes the ratio of box volume to port length and surface area determines the resonance frequency. Ported boxes are far more sensitive to changes in box volume.



True, although it’s been generally agreed since the very early days that this is an undesirable effect, so it’s generally been considered a flaw rather than a feature.



Definitely. At the port resonance frequency, the driver theoretically doesn’t move at all, with all output at this frequency coming from the port. The more rigid and dense the cabinet, the closer to theoretical perfection this works.



Yep, cabinets are fundamentally important :) Also keep in mind that when you run a driver in free air there will be cancellation of the front wave by the out-of-phase back wave, which will greatly diminish low frequency performance.
Right, but where does their contribution to the sound emanate from? It’s clear the drivers actively create sound waves by vibrating. If you put a speaker in a cabinet, does the cabinet just change the actual motion of the driver cone?
 

andreasmaaan

Major Contributor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,757
#72
Right, but where does their contribution to the sound emanate from? It’s clear the drivers actively create sound waves by vibrating. If you put a speaker in a cabinet, does the cabinet just change the actual motion of the driver cone?

In the case of a closed box, the air inside the box exerts pressure on the woofer which impedes its motion, raising the resonance frequency and creating a 12dB/octave high pass filter. The smaller the volume, the higher the resonance/cutoff.

In the case of a ported speaker, the port is a Helmholtz resonator: a region of air having lower acoustic impedance than the air from which the sound wave is propagated (in this case, the air behind the woofer inside the box). When sound is reflected through such a low-impedance region, its phase is reversed. Since the phase of the back wave of the woofer is already out of phase with the wave coming from the front of the woofer, when the back wave’s phase is reversed by the port it ends up exiting the port in-phase with the front wave (except below the frequency of the ported box's resonance frequency, where the port is ineffective). The result is that the outputs from the port and the front of the woofer add constructively from the resonance frequency up, which a sharper (24dB/octave) roll-off below the resonance frequency due to cancellation between the port's output and the output from the front of the woofer (due to their being out of phase below this frequency).

In reality, box materials are not infinitely rigid or reflective, and boxes are not infinitely tightly sealed, so there are losses and unintended resonances that contribute additionally to the output. By this I mean that:
  • panels vibrate much like a speaker, especially at particular frequencies defined by cabinet geometry and materials
  • sound passes through panels
  • sound escapes through gaps in the panels or e.g. through the woofer’s cone or dustcap etc
This can of course be audible if it’s significant enough.

In most high quality speakers since the 80s, losses and resonances have been quite well suppressed. I believe Celestion is widely credited with making the first serious attempt to address this issue in the late 70s, but I’ve forgotten which model it was now. Even long before this though, it was understood that losses and resonances were theoretically undesirable, and some effort was made by various manufacturers to brace cabinets, use thicker ply or other materials, etc (to the best of my knowledge - I wasn’t actually around back then ;))
 

Soniclife

Major Contributor
Patreon Donor
Joined
Apr 13, 2017
Messages
1,615
Likes
1,190
Location
UK
#73
My understanding is that reflections/diffractions which arrive very early, within the first .68 milliseconds, are especially detrimential to imaging. This corresponds to a path length of about 9 inches, which in turn is roughly the distance around the head from one ear to the other.
That's very similar to the width of a lot of the towards the top of the range slim floor standers, if rounded corners don't eliminate the issue that seems an odd position to end up in.
 

andreasmaaan

Major Contributor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,757
#74
That's very similar to the width of a lot of the towards the top of the range slim floor standers, if rounded corners don't eliminate the issue that seems an odd position to end up in.
The distance Duke is talking about is from the acoustic centre of the driver to the baffle edge, not from edge to edge (if I understood his post correctly).
 

Soniclife

Major Contributor
Patreon Donor
Joined
Apr 13, 2017
Messages
1,615
Likes
1,190
Location
UK
#75
In most high quality speakers since the 80s, losses and resonances have been quite well suppressed. I believe Celestion is widely credited with making the first serious attempt to address this issue in the late 70s, but I’ve forgotten which model it was now. Even long before this though, it was understood that losses and resonances were theoretically undesirable, and some effort was made by various manufacturers to brace cabinets, use thicker ply or other materials, etc (to the best of my knowledge - I wasn’t actually around back then ;))
There is also the BBC approach, of thin walled and damped, as used still by Harbeth and others. I've not seen convincing test results that show their approach works, or fails, I've liked many of the speakers that use it though.
 

andreasmaaan

Major Contributor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,757
#77
There is also the BBC approach, of thin walled and damped, as used still by Harbeth and others. I've not seen convincing test results that show their approach works, or fails, I've liked many of the speakers that use it though.
Yeh that’s a good point, and something that was certainly happening prior to the 80s IIRC.
 
Joined
Apr 22, 2016
Messages
57
Likes
67
Location
Princeton, Texas
#79
I actually did mention this effect in both my previous posts ;) But what is news to me is the 9”/0.68ms threshold. Do you have a source for this pls?
I might have read it in an article in Audio magazine many years ago. I don't remember the name of the article but it dealt with recording studio acoustics, and I think the authors were Pisha and Bilello or something like that.

Or it may have been in something I read at the local university library after reading the Pisha and Bilello article... I wasn't keeping track of sources of information back then, I was just trying to learn stuff that I could use. It might have been in a paper by James M. Kates.
 
Joined
Apr 23, 2019
Messages
33
Likes
3
#80
Yeah, I second the Genelec One 8331. Nothing better at that price point or higher that I have heard. They will blow you away with their DSP room correction and precision.
 
Top Bottom