I agree. Looks like a 3.5 way with the .5 moved up top so as to reduce the vertical footprint.
I also like the baffle; the radii start tangent to the face, whereas many other enclosures with curved walls start at a non-zero angle.
I don't think you get my point. Obviously you want symmetry, however how would you adjust the speaker spacing (and therefore angular separation) and toe in to optimize imaging at the desired listening location ? You can't if they are in a fixed location flat on a wall.
What if you decide to move your listening position further back ? That would require the separation to be increased and toe in changed for optimal imaging.
Most speakers have some diffraction issues, why add a whole lot more ? A "just in front of wall" speaker doesn't solve any of the diffraction effects of a free standing speaker, in fact because of the shallow depth there is less opportunity to have a large radius rounding of the cabinet, and then there is the inevitable reflection from the surrounding wall that will be well under 1ms and so will smear the perceived direct field from the speaker. (It's within the hearing integration window)
So, no TV or furniture on the same wall as the speakers then ? That doesn't sound very WAF friendly.
If there is a TV in front of the same wall as the speakers, between them, it doesn't matter that it is not in a line to block the direct path - there will still be diffraction off the TV. The wider the dispersion the speakers have, the worse the problem becomes. The only solution to this (other than a flush mounted TV in the wall) is to have the front of the speakers slightly ahead of the front most object - eg the TV and/or AV cabinet.Don't be so sure of that. I'm living in the UK and interior walls are just as poor as elsewhere. Gypsum board or plasterboard with little in the way of bracing. Hollow and resonant. Even a solid panelled timber wall will still resonate and add colouration.4. Only a problem in US homes
This is one reason why imaging improves with a certain minimum distance from speakers to the front wall and sidewalls (somewhere around 0.7 metres) - the resonances of the walls are attenuated in level and delayed enough in time to be ignored as a delayed arrival. When the speakers are right up against either front or side wall the resonances of the walls become part of the first arrival.
Depending on your room it may not be practical. For many seating positions and normal speaker locations direct to reflected ratio is only just high enough (see earlier in this thread) - another metre of listening distance is enough for you to be out of the direct field, especially if the wall mount speakers lack directivity.6. Nobody will stop you from moving closer to the speaker
Last edited:
Any design where the drivers are asymmetrically offset to one side horizontally (tweeter and mid in that example) shows that the designer doesn't really understand whats bad about diffraction IMHO. Yes it might make the on axis frequency response look a bit flatter, but the diffraction is still there just as strong - just smeared out a bit more in time. The time domain response is still a mess, and now you also have an asymmetric polar pattern which is not good for imaging.
Good imaging doesn't require drivers jammed extremely close together anyway, certainly not at the expense of having them horizontally offset. If the drivers are all vertically aligned, symmetrically placed, and using a well designed crossover you can get excellent imaging with surprisingly large separations.
No it is mainly dictated by the room, but depending on the speaker the triangle can/should start at 2m or even less. This is also driven by the ratio of direct/reflected sound.
I absolutely agree with this. It is pure geometry that suggests this. Let me do a little excursion before I continue here:
There seem to be 2 major speaker properties:
1. Imaging ("sound quality" in Toole) and
2. "spatial quality" (caused by lateral reflections from mainly 60° after >=6ms...10ms).
In this forum and discussion you can get the impression that for good imaging you need narrow dispersion loudspeakers. This is not so and Toole also says the following (page 139 bottm, 140 top):
"In summary, it is clear that the establishement of a subjective preference for the sound from a loudspeaker incorporates aspects of both sound quality and spatial quality, and there are situations when one may debate which is more imprtant (6.283: this happens here to some degree). The results discussed here all point to the same direction: that wide dispersion loudspeakers, used in rooms that allow for early lateral reflections, are preferred by listeners especially, but not exclusively, for recreational listening (6.283: and this is what we are talking about here, recreational listening. We are building speakers for the living room, not work place).
There appear to be no notable sacrifices in the "imaging" qualities od stereo reproduction. Indeed, there are several comments about excellent image stability and sensations of depth in the soundstage."
Personally, I am shooting for both, good imaging and spatial quality. I also agree with this rephrased Geddes statement.
The only problem is that with speakers with such a narrow dispersion of +-45° you simply cannot generate lateral reflections coming from 60° in your typical small living room ! It does not work ! The reflections can come from all kinds of directrions and many of them pass the listener behind the back/couch/seat and coming from behind they are less effective. This is also confirmed by Toole.
So "doesn't this simply put a requirement on the amount of directivity?" yes !
I haven't had the chance to create a real drawing but I believe in order to fulfill both "requirements" the dispersion should at least be +-60° if not more.
Now people can again come along with the Ghostbusters trick: "Cross the rays !"...of narrow dispersion speakers. That widens the sweet spot and allows reflections from 60°. But now sound from the left channel comes from the right wall and vice versa. Are you sure that in this case the imaging is of the same high quality compared to the "none crossed rays" situation ? I have serious doubts.
Last edited:
I look forward to your reasoned response, because quite simply I'm baffled by your current response. I actually had to look up Straw man to make sure it meant what I thought it did. I can't see any instances where I used a straw man argument however I can see a couple where you did.
I'm not even sure if this is in response to the original question or my query about the driver layout of the speaker you presented.
Your original question was "why aren't wall speakers used more often" to which I listed several potential technically valid reasons. You might disagree on some points, sure, but I fail to see how this is short on substance.
My point about colouration from nearby interior walls due to their resonant panel nature is refuted by "Only a problem in US homes", which is both a Straw man and an answer with very little substance when you could have responded with a counter argument about the properties of the materials I listed or their prevalence or lack of in certain countries.
Nowhere did I hold up US homes as the reference for living rooms, it should be apparent from my profile that I'm in the UK, and in fact I've never lived in or even been to the US. When listing those materials I'm drawing on personal experience from living in both the UK and New Zealand where acoustically poor interior wall cladding like that is common. Timber panelled walls are no solution either, as timber is highly resonant. (And not used in speaker cabinets for that reason, as well as warpage)
Even if we imagine for a moment US homes had interior walls with acoustically poor materials and everywhere else didn't (a preposterous stance) then that's still a large market of people affected who might do well to not use in or on wall speakers, and could be a valid reason for them not being common.
As for the photo of the speaker you provided, I simply queried the odd layout of the drivers and their purpose, you made the claim that it would have been done for the benefit of imaging. Perhaps that was the designers intention, but if so it was misguided.
You'll have a very steep road convincing me that bunching together a group of drivers as close together as possible in an asymmetric pattern is going to improve imaging over a straight vertical line of drivers.
We can't tell from looking whether its a 3 way with two woofers, or a 3.5 way as suggested by 454Casull, but lets examine both possibilities.
If they are truly only a pair of woofers in a 3 way system there is no benefit to putting one woofer at the top and to the side of the tweeter (instead of below the other woofer) due to the likely large wavelengths at the crossover frequency. However putting a large woofer right to the side of the tweeter will be a large source of diffraction for the tweeter - an asymmetric source, since the woofer is only on one side. This will be detrimental for imaging and polar response smoothness and symmetry.
If it is 3.5 way with one of the woofers producing midrange, you have much the same scenario if the lower woofer is the midrange one (diffraction from the bass only woofer beside the tweeter) and if its the other way around then you have the worse problem of having horizontally spaced drivers crossing over with each other giving horizontal lobing. There may be some other reason for the choice of driver location, but "thinking about imaging" isn't it.
If you have any technical points to respond to I'm happy to hear them and discuss them, but claiming my posts are lacking in substance or contain straw man arguments seems like a bit of a cheap shot, not in your usual character.
Last edited:
This Toole quote has been discussed previously in this thread. Toole was discussing mainstream cone-dome speakers, and I believe 'wide dispersion' he refers to speakers that do not beam excessively.
You might have missed this, so I'm posting it again. It's a little more about Toole's opinion.
This is mainly to widen the sweet spot, not for intentionally generating stronger side wall reflections (in my opinion, because the contralateral reflection is still quite early). Because the speaker is directional and you're listening off-axis in the middle, phantom sources will move less strongly to the nearer speaker as you move off-center. For example, if you're moving to the left, you're moving more off-axis from the left speaker and more on-axis from the right speaker.
In a crossfired setup the left channel comes from the left speaker, etc. It's just that the strongest reflection of that channel comes from the other side.
Having tried both triangle and crossfired approaches, they do image differently. Subjectively, the crossfired setup puts one further back in the hall, whereas the apex-of-the-triangle approach is more pin-pointy. I prefer the former, but can see why one would prefer the latter too.
Let's approach the on-wall speaker concept form a different perspective. Instead of comparing advantages/disadvantages, good/bad, let's list differences of on-wall compared to box speaker:
1. Adjacent-boundary effects from front wall
2. Output at lower frequencies (no baffle step)
3. Coupling to certain modes
4. Mounting on wall (resonances)
5. Toe-in of speaker
Anything else?
1. Adjacent-boundary effects from front wall
2. Output at lower frequencies (no baffle step)
3. Coupling to certain modes
4. Mounting on wall (resonances)
5. Toe-in of speaker
Anything else?
WAF !
Agreed
but I think Keyser started this thread with "Psychoacoustics first, practical implications later"
Yes, distinction needs to be drawn between on wall (shallow baffle hung on or pushed hard up against the wall) and actual flush mounted in wall speakers. Some factors are the same for both and some are not, and there can be a tendency for people to mix the two interchangeably in conversation.
If one has a loudspeaker with sufficient controlled directivity to allow it to operate next to an identical unit and not produce an audible seam, then that speaker can usually be placed on a physical boundary at the same plane without incurring close reflections. A physical boundary and acoustic mirror source being similar.
With a boundary compatible system the size of an SM-60, one only sees a rise below about 300Hz due to the boundary loading at the lower frequencies. This is possible because the horn wall and room wall are only about 3 inches apart and are parallel. In commercial sound, this boundary is usually the ceiling however it works the same on the floor or on the walls or with adjacent cabinets.
An advantage in narrow rooms is the on wall placement produces a much wider speaker position than normal and in a narrow room where the side walls produce strong reflections, this position eliminates them so far as the close wall.
Like all loudspeakers with a radiation lobe, it is usually best to aim the speakers at the farthest seats, in the case of hifi, that is the kitty corner aiming or as they say in hifi toe in.
Fwiw, a flat baffle driven by a point source also acts like a horn, but it has a 180 degree pattern. Examine Don Keele's pattern loss F formula and consider an angle of 180 degrees.
Best,
Tom
With a boundary compatible system the size of an SM-60, one only sees a rise below about 300Hz due to the boundary loading at the lower frequencies. This is possible because the horn wall and room wall are only about 3 inches apart and are parallel. In commercial sound, this boundary is usually the ceiling however it works the same on the floor or on the walls or with adjacent cabinets.
An advantage in narrow rooms is the on wall placement produces a much wider speaker position than normal and in a narrow room where the side walls produce strong reflections, this position eliminates them so far as the close wall.
Like all loudspeakers with a radiation lobe, it is usually best to aim the speakers at the farthest seats, in the case of hifi, that is the kitty corner aiming or as they say in hifi toe in.
Fwiw, a flat baffle driven by a point source also acts like a horn, but it has a 180 degree pattern. Examine Don Keele's pattern loss F formula and consider an angle of 180 degrees.
Best,
Tom
What is "sufficient" expressed in numbers? And, you're strictly talking about mounting the horn on the side walls not the front wall, right?
Last edited:
Before they take down the web site take a look at the Snell range:
Snell Acoustics | In-Wall Speakers
I designed these as very high end home theater products. I was able to deal with toe in as well as low diffraction mounting, and also used real cabinets that didn't rely on the wall cavity to be a proper enclosure.
In-walls done right can give smooth response and remove the back wall bounce. So can studio monitors that are properly flushed in.
David S.
Snell Acoustics | In-Wall Speakers
I designed these as very high end home theater products. I was able to deal with toe in as well as low diffraction mounting, and also used real cabinets that didn't rely on the wall cavity to be a proper enclosure.
In-walls done right can give smooth response and remove the back wall bounce. So can studio monitors that are properly flushed in.
David S.
So far as a number, I have not examined where or if there is a sharp transition, I suspect it is a shades of grey issue. What can be said is that if one had an axial Q of 10, then one has contained about 90% of the speakers energy into the desired beam, with a Q of 20, one has about 99% of the energy within the pattern and 1% outside.
The greater the Q, the less the response of the off axis energy matters as the reverberant field will be driven by the direct spectrum.
In a room, the greater the Q, the larger the direct field is, the larger the distance where words can be understood.
So far as mounting, as long as the outer horn wall angle tips in or is parallel to the room wall, then one can examine the polar plot looking at 90 degrees off axis to estimate the level of side energy.
So far as what placement in front of a wall, a similar situation exists in that one can examine the polar plot and get a feel for how strong the rear radiation is relative to the forward direction.
Normally the rear radiation would be lower than the 90 degree off axis level. If the rearward energy is only -3 dB relative to the front level, then that will make a strong (50%) delayed reflection, if the rearward level was -20dB, then I would guess the reflected energy (about 1%) would be approaching or actually inaudible.
While what I am suggesting is an observation made after developing directional speakers for commercial use where directivity is critical.
I would suggest those interested in this area, to do the outside experiment I have described and think of that as the blank sheet of paper because outdoors the room effects are obviously gone, and directivity doesn’t matter, leaving only the other things in the chain.
Consider that in addition to the room, the loudspeaker ALSO is a strong part of what governs the realism of the stereo/mono phantom image.
Best,
Tom Danley
6. Nobody will stop you from moving closer to the speaker
Close to a speaker brings up another issue... driver integration. Maybe not a big deal for a coaxial but with other speakers i've notice that it takes 1.5, even 2m before it even measures correctly WRT driver summing.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.