The wall behind the speaker is indeed a problem if it is too close (which it often is). In "Sound Reproduction", Toole somewhere cites an author who found that damping the front wall improved imaging and reduced colouration. I've heard many experts agree on this.
However, Bech in his thorough experiment did find that the floor reflection can cause audible colouration as well as a change in imaging. So although some call it a "natural" reflection, I don't think there is any real debate whether it is problematic. The colouration was in the 500-2000 Hz range by the way.
However, Bech in his thorough experiment did find that the floor reflection can cause audible colouration as well as a change in imaging. So although some call it a "natural" reflection, I don't think there is any real debate whether it is problematic. The colouration was in the 500-2000 Hz range by the way.
So I think we all agree the reflection from the wall behind the main speakers is a problem, and floor bounce is too. They create notches where the direct sound and reflected sound are 1/2λ apart, usually in the lower midrange. Some may consider the ground reflection to be a "natural" problem but we all agree it creates a notch if the sound source is a point source. And I think we all agree the ceiling reflection is objectionable too.
If we agree with these things, then we can agree that source directivity, configuration and/or placement is best if it eliminates or at least reduces those particular reflections. That would be somewhat axiomatic, wouldn't you say?
Side and rear reflections are another story. Seems to me most consider the contralateral wall reflection to add spaciousness. What about the back wall? I've heard it said that the back wall can be "live" but it seems to me that it's best if it isn't too live, either by diffusing it or having the listeners distant from it. I personally prefer to diffuse the back wall, using bookshelves or other randomized reflectors. In rooms where this isn't possible, I make sure the listeners are closer to the (front) speaker sources than the (rear) wall reflection. Do any of you do that too?
I think these "helpful" spacious relections sort of fall into the category of "helpful, with caveats." That maybe goes to the point so many have made here that we have to talk about the room if we're going to talk about source directivity. I do think we can safely generalize the back wall and contralateral wall reflections as "helpful" with the suggestion that the listener shouldn't sit too near them.
And what about the ipsilateral wall? I've always considered it to be similar to the wall behind the speakers, not quite as objectionable but potentially problematic, nonetheless. I try to limit reflections from the ipsilateral wall in the mains, although I often take advantage of those reflections in surrounds. What are your thoughts about ipsilateral wall reflections?
If we agree with these things, then we can agree that source directivity, configuration and/or placement is best if it eliminates or at least reduces those particular reflections. That would be somewhat axiomatic, wouldn't you say?
Side and rear reflections are another story. Seems to me most consider the contralateral wall reflection to add spaciousness. What about the back wall? I've heard it said that the back wall can be "live" but it seems to me that it's best if it isn't too live, either by diffusing it or having the listeners distant from it. I personally prefer to diffuse the back wall, using bookshelves or other randomized reflectors. In rooms where this isn't possible, I make sure the listeners are closer to the (front) speaker sources than the (rear) wall reflection. Do any of you do that too?
I think these "helpful" spacious relections sort of fall into the category of "helpful, with caveats." That maybe goes to the point so many have made here that we have to talk about the room if we're going to talk about source directivity. I do think we can safely generalize the back wall and contralateral wall reflections as "helpful" with the suggestion that the listener shouldn't sit too near them.
And what about the ipsilateral wall? I've always considered it to be similar to the wall behind the speakers, not quite as objectionable but potentially problematic, nonetheless. I try to limit reflections from the ipsilateral wall in the mains, although I often take advantage of those reflections in surrounds. What are your thoughts about ipsilateral wall reflections?
Limiting reflections from the mains and using diffusor panels to break up ipsilateral reflections. Seems best balance I've found. Surrounds, ditto 
Do like "flanking" woofers to support filling the void left. Tho we have evolved to take this into account I believe there is a threshold were floor bounce becomes an annoyance. It's like adding an additional reflection to the recorded reflection when the two combine in the wrong way there's a noticeable phase smear in the upper ranges where localization is precise.
Side by side woofer configurations appear to handle this better IMHO. I'm really considering three 6.5" side by side floor coupled some form of MLTL, YES for these very reasons
Do like "flanking" woofers to support filling the void left. Tho we have evolved to take this into account I believe there is a threshold were floor bounce becomes an annoyance. It's like adding an additional reflection to the recorded reflection when the two combine in the wrong way there's a noticeable phase smear in the upper ranges where localization is precise.
Side by side woofer configurations appear to handle this better IMHO. I'm really considering three 6.5" side by side floor coupled some form of MLTL, YES for these very reasons
What specifically were you impressed about? Spaciousness, locatedness, etc.?
Locatedness and stability of the image. Phantom sources were located with great ease and there was a certain naturalness to the image. Distance of the image seemed to depend on the recording; as it should, perhaps. Generally it sounded somewhat further away than most speakers, but this was very pleasing with live music like Clapton Unplugged.
Horizontal directivity was mediocre since it was built with Visaton SC5.9 drivers, but smooth and reasonably wide up to 5 kHz. It did require lots of DSP to sound neutral.
Here's a picture of the pair: http://forum.zelfbouwaudio.nl/viewtopic.php?f=26&t=18432&start=75
Horizontal directivity was mediocre since it was built with Visaton SC5.9 drivers, but smooth and reasonably wide up to 5 kHz. It did require lots of DSP to sound neutral.
Here's a picture of the pair: http://forum.zelfbouwaudio.nl/viewtopic.php?f=26&t=18432&start=75
Last edited:
With the CBT the floor reflection is a benefit. It is coupled to the floor and the reflection contributes what is effectively the other half of the CBT. So it negates the need for a full arc. It effectively eliminates the need for floor treatment that is problematic at best. Especially as any floor treatment typically exhibits a such a high frequency profile as to not be effective as broadband absorption for a traditional speaker. A bare broadband reflective floor surface is optimal. No floor bounce and no specular reflections are arriving from the reflective surface.
The vocal of the CBT36 is perhaps a tiny bit lower then many other speakers, but can easily be lifted by tuning the frequency response.
Imaging of a speaker is depended largely on how the ETC graphs are in the room. It's a speaker and room integration issue more then anything else. Meaning; almost any speaker can produce a great image under the right circumstances.
The vocal of the CBT36 is perhaps a tiny bit lower then many other speakers, but can easily be lifted by tuning the frequency response.
Imaging of a speaker is depended largely on how the ETC graphs are in the room. It's a speaker and room integration issue more then anything else. Meaning; almost any speaker can produce a great image under the right circumstances.
This is true about the mirror image, but I don't think it means that the sound appears to come from the center of the virtual array (the floor level driver). Like most line arrays, I think the perceived source location is more or less the ear level driver (or the driver even with a line perpendicular to the array going through the viewers ear level).
Regards,
David
I agree that the floor and rear wall bounce create audible abberations. The question about source directivity is a practical one. This thread seems to be about the benefit of typical controlled directivity speakers as in "lets give it a waveguide instead of a wide angle dome".
In my experience the usual increased directivity speaker will make no difference in floor bounce problems. These are typically a problem around 200Hz and realistic waveguide sizes aren't going there. Also, a few more dB of directivity index won't put too much of a dent into a floor bounce. The real answer for these is to do the Allison trick of placing LF drivers near the floor and then crossover to higher range drivers well away from the floor. This lets you put bounce frequencies into the stop band of both upper and lower range drivers and essentially do away with the problem.
You can also do some array tricks with 2 or 3 woofers to create a response null in the downfiring floor bounce direction. Also full range line arrays extending to the floor will tend to have no floor bounce issue. These approaches deal with the specifics of the problem rather than making the speaker generally more directional.
David S.
This is a really productive discussion, in my opinion. Thanks for your input, guys.
The thread really has two topics, one a subset of the other. It is about loudspeaker configurations that can provide uniform directivity and the various patterns they create, which ones are most useful and effective. And it is also about the horn/waveguides chosen to build them, their aspect ratios, flare profiles and other details, and how effective they are when used in a loudspeaker system.
The thread really has two topics, one a subset of the other. It is about loudspeaker configurations that can provide uniform directivity and the various patterns they create, which ones are most useful and effective. And it is also about the horn/waveguides chosen to build them, their aspect ratios, flare profiles and other details, and how effective they are when used in a loudspeaker system.
Last edited:
Do you think on a mirror floor in otherwise full space they would image on the axis from the listener through r behind the base of the arc on the ground plane?
I don't know if that is universally true but it would certainly be a good approximation. If the CBT drivers are in a true semi circle and the network doesn't add a lot of excess delay to upper elements then it is probably true. I think the apparent location is primarily tied to the first arival of the HF wave, and the driver it came from.
Somewhat similar is the older JBL diffraction slot CD horns. If you move laterally around them the diffraction slot seems to be the origin. If you move vertically then the entrance at the throat becomes the source. I think Don Keele called it a sort of astigmatism.
David S
You know, the floor (or any other boundary) mirrors a point source too. It's just that with a point source, the mirrored source becomes a second and that's different than multiple sources. Interference is course. With multiple sources, interference becomes dense enough the sound field starts averaging. It's kind of like the various forms of multisub configuations, which are actually all arrays of one form or another. Just something to think about.
Well Wayne, as these types of discussions continue, the underlying reality is almost always revealed as a conflict of objectivity vs subjectivity in regards to what an individual 'enjoys' from a listening experience.
Designing through objectivity is smart engineering 'almost' all of the time. Taking many of the room induced variables out of the equation through constant directivity is a sound approach of course. Using drivers with low harmonic distortion and a smooth frequency response are equally responsible choices.....a design that employs all of the above is sure to please 99% of the people.....it's the 1% that play the wildcard for one reason or another. The factors that define the small percentage of audiophile listeners are quite broad.....some based in real world, quantitate measures and others derived from bias, memory, or other psychologically based rationale.
I think we all need to accept that as we get older and out hearing degrades, we find ourselves constantly searching for more of an audible 'memory' that becomes increasingly difficult to replicate as we age. Once we hit a proverbial wall, we tend to substitute 'new' experiences that never really fully pay off........sad but true IMO....LOL.
As a younger man, my listening focused on the top and bottom of the frequency range, always pleased by HF 'air', 'smoothness', and the overall definition that a well produced top octave could provide. Now that half that octave is gone forever, my brain tends to focus on the midrange.....as there's simple less...um....distractions? and my brain is now looking for intelligibility and less fatigue.
I suspect as an 'older' man in the years of retirement, listening will again shift in priority.....and I'm glad to not yet speculate on what that shift might be. For now, I'll enjoy my middle aged preferences while I still can......LOL.
Designing through objectivity is smart engineering 'almost' all of the time. Taking many of the room induced variables out of the equation through constant directivity is a sound approach of course. Using drivers with low harmonic distortion and a smooth frequency response are equally responsible choices.....a design that employs all of the above is sure to please 99% of the people.....it's the 1% that play the wildcard for one reason or another. The factors that define the small percentage of audiophile listeners are quite broad.....some based in real world, quantitate measures and others derived from bias, memory, or other psychologically based rationale.
I think we all need to accept that as we get older and out hearing degrades, we find ourselves constantly searching for more of an audible 'memory' that becomes increasingly difficult to replicate as we age. Once we hit a proverbial wall, we tend to substitute 'new' experiences that never really fully pay off........sad but true IMO....LOL.
As a younger man, my listening focused on the top and bottom of the frequency range, always pleased by HF 'air', 'smoothness', and the overall definition that a well produced top octave could provide. Now that half that octave is gone forever, my brain tends to focus on the midrange.....as there's simple less...um....distractions? and my brain is now looking for intelligibility and less fatigue.
I suspect as an 'older' man in the years of retirement, listening will again shift in priority.....and I'm glad to not yet speculate on what that shift might be. For now, I'll enjoy my middle aged preferences while I still can......LOL.
Yes, I think that's true to a large degree. Once you get to the point where you're talking about subjective differences, things get murky. And if you need double blind tests to determine which of two things is better - If it's that close, then both are pretty darn good. But I think we still have some objective, measureable facts we can look at, and I think it's worthwhile to talk about what differences are most audible. That's what the whole "baby and the bathwater" comment from the first post was about.
If we know that what's really important is smooth on-axis response and spectral balance of the reverberent field, then that can help dictate our design choices. If we know what directivity patterns work well in small rooms, that can really help guide our designs too. To build the best loudspeakers, we need to know what optimizations are most important to sound quality, e.g. on-axis response, shape of polars, change of directivity with frequency, etc. So I think it's useful to discuss what aspects are most important to optimize for high-quality uniform-directivity loudspeakers used for home hifi and home theater applications.
If we know that what's really important is smooth on-axis response and spectral balance of the reverberent field, then that can help dictate our design choices. If we know what directivity patterns work well in small rooms, that can really help guide our designs too. To build the best loudspeakers, we need to know what optimizations are most important to sound quality, e.g. on-axis response, shape of polars, change of directivity with frequency, etc. So I think it's useful to discuss what aspects are most important to optimize for high-quality uniform-directivity loudspeakers used for home hifi and home theater applications.
I actually always go a little tighter. My HF and MF profiles tend to be just shy of 90°, in the 80° to 90° range.
This is only partially related, but I think it's a somewhat useful and interesting aside. The area expansion of a 90° trihedral corner (three-sided pyramid) is the same as a quadrahedral (four-sided) pyramid with 70° wall angle.
So for DI-matching (this time, talking index, not polars), a 90° axisymmetric horn is not matched with the 90° expansion of a room corner. For DI-matching, one would want a narrower pattern.
In truth though, I think this is almost irrelevant. I agree with Dave Smith, that what we really care about is a near-match in the polars, and not so much the directivity index.
But what I do think is important is that the pattern not be too wide. Ideally, I like 80° to 90° horizontal beamwidth and 40° to 60° vertical. Since beamwidth is defined as the -6dB angle, we're really not talking about a hude attenuation at the edge of the "beam". Even if the stated beamwidth is 70° (+/-35°), out at 45° it's only about 2dB lower than a device that generates a 90° beam. So I'm pretty happy with 70° - 90° horizontal beamwidth, and that puts my design target around 80° or 85°. By happy coincidence, conical horns and waveguides are often smoother with a slightly narrower profile too.
Beamwidth being just under 90° is especially useful for dealing with the ipsilateral reflection, when combined with toe-in. So I like the MF/HF beamwidth to be slightly narrower than 90° for constant directivity cornerhorns or matched-directivity two-way speakers using the (heavy toe-in) crossed-axes placement method. I think we all talk about 90° flares, but that's mostly as a quarter-sphere approximation. In truth, I think it's useful to be a smidge under that.
This is only partially related, but I think it's a somewhat useful and interesting aside. The area expansion of a 90° trihedral corner (three-sided pyramid) is the same as a quadrahedral (four-sided) pyramid with 70° wall angle.
So for DI-matching (this time, talking index, not polars), a 90° axisymmetric horn is not matched with the 90° expansion of a room corner. For DI-matching, one would want a narrower pattern.
In truth though, I think this is almost irrelevant. I agree with Dave Smith, that what we really care about is a near-match in the polars, and not so much the directivity index.
But what I do think is important is that the pattern not be too wide. Ideally, I like 80° to 90° horizontal beamwidth and 40° to 60° vertical. Since beamwidth is defined as the -6dB angle, we're really not talking about a hude attenuation at the edge of the "beam". Even if the stated beamwidth is 70° (+/-35°), out at 45° it's only about 2dB lower than a device that generates a 90° beam. So I'm pretty happy with 70° - 90° horizontal beamwidth, and that puts my design target around 80° or 85°. By happy coincidence, conical horns and waveguides are often smoother with a slightly narrower profile too.
Beamwidth being just under 90° is especially useful for dealing with the ipsilateral reflection, when combined with toe-in. So I like the MF/HF beamwidth to be slightly narrower than 90° for constant directivity cornerhorns or matched-directivity two-way speakers using the (heavy toe-in) crossed-axes placement method. I think we all talk about 90° flares, but that's mostly as a quarter-sphere approximation. In truth, I think it's useful to be a smidge under that.
I was actually thinking pretty much along the lines of your "interesting aside" there - been wondering if a 60x60° or 60x40° -6dB might be worth considering in a lot of home situations. Brings up a question of how best to taper from omni, I suppose, but it seems like that's still a somewhat open question for 90° too.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.