jlo said:
post scritpum and corrigendum
I'm sorry - I got suggested by Rybaudio "attached image"
clearly what You have described is a different case
but I wonder how You got this result: "about 2.1ft and the one from back wall is about 3.4ft"?
after all it is a function of the listener's distance from the walls and that is also the question of the listener's distance from stereo base and of the angle of stereo triangle
What assumptions have You made?
nota bene Blumlein recommended for a stereo triangle an isosceles at 45 degrees
also 16 feet listener's distance from loudspeaker is also quite far, something rather not typical in small rooms
ok, why not
then let's make it 2.5 feet
my point is that it is not a practical problem in a typical living room
6 feet would be surely a practical problem
but it appears that nobody needs those "6 feet"
best regards,
graaf
with a speaker radiating near 4 walls (floor, ceiling, rearwall, sidewall), you will get the direct sound plus 4 first reflections. All those reflections will have a different travel time to the listener : depending on the speakers/listener geometry, generally you'll get direct sound, than floor, than sidewall, than ceiling or rear wall. I gave approximative numbers for the typical 60° triangle. But it means that you'll get the first reflections between 0.5 and 10 ms after direct sound.
I did a software to try to simulate listening with/without first reflections (and other things), just download "the final cut" on my website and play with listener and speaker position.
graaf said:
I am not joking, and I am also not trying to make any statements regarding typical listening rooms. I simply posted an easy to calculate case that illustrates how a reflection can have a path length difference of less than the objects distance from the source. I think I misunderstood you; I thought you were asking something more basic than you were. The floor reflection is the only early reflection that is approximated by the previous drawing I attached.
If you are asking about the first horizontal reflections in typical configurations in rectangular rooms I'm inclined to agree with you that the reflections have a path length difference close to their distance from the walls. I've attached another image of a case of this.
Attachments
a part of: "Design considerations for an idealized domestic surround sound listening space"
You seem to imply I don't know what it is I'm reading, or perhaps something else. The title of Moultons paper is indeed Design considerations for an idealized domestic surround sound listening space .
You ask:
why have You omitted these two words: "surround sound "
I didn't think they were important, Moulton certainly doesn't. And to support this, we should look at what he says.
In the second paragraph, first page of that article Moulton writes:
Criteria for a high-quality domestic listening environment
Multiple uses
It is not reasonable, given the investment involved and the similarity of architectural and hardware requirements, to devise separate and unique spaces for listening to original music, to reproduced music or for viewing film and/or video. I assume, therefore, a single, multi-use space. I have even envisaged the integration of live acoustic music performance as well, but that implementation remains beyond the scope of this paper. Some fascinating possibilities exist.
I emphasize it, he writes:
It is not reasonable, given (...) the similarity of architectural and hardware requirements, to devise separate and unique spaces for listening to original music, to reproduced music or for viewing film and/or video. I assume, therefore, a single, multi-use space.
He expects this space to be used for conventional music listening as well as other uses. Indeed, both high end and mediocre North American domestic listening spaces are increasingly put to exactly the purposes as described by Moulton. This is the reality. The surround sound room, is the multimedia room, is the domestic listening room
Also, on the same page:
Based on personal experience with numerous purpose-built and adapted home theatre and studio-type listening rooms, I suggest that the minimum size for a successful high-quality room might be 20’ long, 15’ wide and 8’ high.
Note that's the minimum size he recommends.
Also, on that page he says he says, (and I can't help it that this appears to outrage you and that is the impression I get from extraordinary tone of your well crafted, near idiomatic English):
loudspeakers need to be at least 6’ from the nearest wall or else installed in the wall;
As I pointed out, rather reasonably, I thought, near the start of this discussion, which was brought up by your posting of an interview with Moulton, use of omni directional speakers, especially like those he uses in a small listening room, (a relative thing, 'small'), would require placing them well away from the wall, especially if those speakers as he says, need to have broad and uniform lateral high frequency dispersion.
It's an interesting article. Folk might want to read it:
http://www.moultonlabs.com/more/design_considerations_for_an_idealized/P1/
Moulton's domestic listening room is in fact a clone of his studio which is set up as a surround sound room. In an another place he writes:
At present, I am using loudspeakers (designed by Manny LaCarrubba, incidentally) that are essentially flat from 30 Hz. to 17 kHz. over AT LEAST 180° laterally and 30° vertically. I use them in a large (8000 cubic feet) reverberant (.5 sec. RT60) room for production purposes (production, postproduction and mastering, plus research) as speakers of choice. They are configured as a 6-channel discrete playback system (including center and overhead speaker). Because of this configuration, the angle between L and R is in fact 90° at the sweet spot, and there is also a center speaker in place.
http://www.paudio.com/Pages/learning_thread_loudspkrs.html
oh please not that "fat lady" and so on again...
I have read that Toole's marketing literature already and long ago
That is PRECISELY why I'm very interested in what is in his more recent scientific paper, which nobody knows and nobody seems to want to know
I told you how to get hold of it. (Probably at no cost since you are a graduate student).
and even Dr Geddes admitted that "At greater than about 2 ms. the effects transition more to coloration than image"
As I said , if you place, an omnidirectional speaker close to a wall, then you will get colouration, among other highly undesirable effects.
Colouration is distortion. Colouration is a information added to the signal after it leaves the loudspeaker. Colouration effects are created in that reflection time after 2 ms and about out to about10 -15 ms, and they may be avoided through attenuation by placing absorption on the surfaces nearby the speaker, or by attenuating them through its distance from those surfaces or by attenuation by falling power response off axis. Such attenuation will allow the direct signal to mask this distortion rather than the distortion to mask components of the direct signal.
I think distortion is something to be avoided and I'm sure we both agree about that - you said many pages back in this thread accuracy was to be striven for, and I agree.
Like everyone else who's sharp
What are You suggesting?
I sincerely meant he's very intelligent. Probably more intelligent than me. I feel comforted his prescription for speaker placement agrees with my analysis of his situation. He gives up nasty early reflections in the less than 10 -15 ms window and gets a nice, lively reverberation field instead.
I don't know what's going on with you but discussion with you is reminiscent of a combination of the worser aspects of student common room talk and facing a really nasty lawyer in a N American courtroom:
You clearly don't know what You are talking about
...
maybe I was not listening to You carefully enough but fortunately I have listened to omnis and also to relatively close to wall omnis (read: much less the 6 feet away from walls) many times, and not only me, and I have to say say that all You have to say about this only proves that You haven't
Moulton seems to think he has good reasons for specifying a six foot distance and he's a real pro and he listens to them a lot. So perhaps you should put your mind to thinking up reasons he specifies the distance, other than the ones I provided. No?
You seem to imply I don't know what it is I'm reading, or perhaps something else. The title of Moultons paper is indeed Design considerations for an idealized domestic surround sound listening space .
You ask:
why have You omitted these two words: "surround sound "
I didn't think they were important, Moulton certainly doesn't. And to support this, we should look at what he says.
In the second paragraph, first page of that article Moulton writes:
Criteria for a high-quality domestic listening environment
Multiple uses
It is not reasonable, given the investment involved and the similarity of architectural and hardware requirements, to devise separate and unique spaces for listening to original music, to reproduced music or for viewing film and/or video. I assume, therefore, a single, multi-use space. I have even envisaged the integration of live acoustic music performance as well, but that implementation remains beyond the scope of this paper. Some fascinating possibilities exist.
I emphasize it, he writes:
It is not reasonable, given (...) the similarity of architectural and hardware requirements, to devise separate and unique spaces for listening to original music, to reproduced music or for viewing film and/or video. I assume, therefore, a single, multi-use space.
He expects this space to be used for conventional music listening as well as other uses. Indeed, both high end and mediocre North American domestic listening spaces are increasingly put to exactly the purposes as described by Moulton. This is the reality. The surround sound room, is the multimedia room, is the domestic listening room
Also, on the same page:
Based on personal experience with numerous purpose-built and adapted home theatre and studio-type listening rooms, I suggest that the minimum size for a successful high-quality room might be 20’ long, 15’ wide and 8’ high.
Note that's the minimum size he recommends.
Also, on that page he says he says, (and I can't help it that this appears to outrage you and that is the impression I get from extraordinary tone of your well crafted, near idiomatic English):
loudspeakers need to be at least 6’ from the nearest wall or else installed in the wall;
As I pointed out, rather reasonably, I thought, near the start of this discussion, which was brought up by your posting of an interview with Moulton, use of omni directional speakers, especially like those he uses in a small listening room, (a relative thing, 'small'), would require placing them well away from the wall, especially if those speakers as he says, need to have broad and uniform lateral high frequency dispersion.
It's an interesting article. Folk might want to read it:
http://www.moultonlabs.com/more/design_considerations_for_an_idealized/P1/
Moulton's domestic listening room is in fact a clone of his studio which is set up as a surround sound room. In an another place he writes:
At present, I am using loudspeakers (designed by Manny LaCarrubba, incidentally) that are essentially flat from 30 Hz. to 17 kHz. over AT LEAST 180° laterally and 30° vertically. I use them in a large (8000 cubic feet) reverberant (.5 sec. RT60) room for production purposes (production, postproduction and mastering, plus research) as speakers of choice. They are configured as a 6-channel discrete playback system (including center and overhead speaker). Because of this configuration, the angle between L and R is in fact 90° at the sweet spot, and there is also a center speaker in place.
http://www.paudio.com/Pages/learning_thread_loudspkrs.html
oh please not that "fat lady" and so on again...
I have read that Toole's marketing literature already and long ago
That is PRECISELY why I'm very interested in what is in his more recent scientific paper, which nobody knows and nobody seems to want to know
I told you how to get hold of it. (Probably at no cost since you are a graduate student).
and even Dr Geddes admitted that "At greater than about 2 ms. the effects transition more to coloration than image"
As I said , if you place, an omnidirectional speaker close to a wall, then you will get colouration, among other highly undesirable effects.
Colouration is distortion. Colouration is a information added to the signal after it leaves the loudspeaker. Colouration effects are created in that reflection time after 2 ms and about out to about10 -15 ms, and they may be avoided through attenuation by placing absorption on the surfaces nearby the speaker, or by attenuating them through its distance from those surfaces or by attenuation by falling power response off axis. Such attenuation will allow the direct signal to mask this distortion rather than the distortion to mask components of the direct signal.
I think distortion is something to be avoided and I'm sure we both agree about that - you said many pages back in this thread accuracy was to be striven for, and I agree.
Like everyone else who's sharp
What are You suggesting?
I sincerely meant he's very intelligent. Probably more intelligent than me. I feel comforted his prescription for speaker placement agrees with my analysis of his situation. He gives up nasty early reflections in the less than 10 -15 ms window and gets a nice, lively reverberation field instead.
I don't know what's going on with you but discussion with you is reminiscent of a combination of the worser aspects of student common room talk and facing a really nasty lawyer in a N American courtroom:
You clearly don't know what You are talking about
...
maybe I was not listening to You carefully enough but fortunately I have listened to omnis and also to relatively close to wall omnis (read: much less the 6 feet away from walls) many times, and not only me, and I have to say say that all You have to say about this only proves that You haven't
Moulton seems to think he has good reasons for specifying a six foot distance and he's a real pro and he listens to them a lot. So perhaps you should put your mind to thinking up reasons he specifies the distance, other than the ones I provided. No?
I'd like to bring this thread back to the top and pose a question that I've had in my mind for some time but haven't quite been able to word well. I still don't think I can but here's a shot.
For the most part there is a consensus that in order for a loudspeaker to sound good the sound it radiates in every direction needs to have a smooth and either flat or downward tilted frequency response, and the first copy of the sound coming in must be flat or slightly downward tilted. One thus designs a speaker such that the family of FRs in all directions are optimized together, with the listening window being approximately flat and the off-axis responses downward tilted. My question pertains to the off-axis responses that are downard tilted.
What I'm trying to ask is when we hear those reflections coming in, do we "hear" the general linear trend of the FR and then local anomolies are the deviations from that, or is it something else? Wow, unless you already have thought of this, it is probably not clear what I mean by that. A motivating example might help things. Suppose we use a woofer down low mated to a constant directivity device such as a horn. Off-axis the response is flat until the woofer narrows and then it is mated smoothly to the CD device where the directivity of the piston matches that of the device. There are no discontinuities in the response curves, and they are downward tilted, but what I am asking is do we "hear" the response in general as the line that would be fitted to the data and then deviations from that line are percieved as anomilies in the response just as anomolies from flat are heard in the listening window, or how exactly do we percieve this? I've attached a (crude) picture that may clarify what I am trying to ask. Maybe I can phrase this another way- is it bad for the off-axis responses to be concave up in any region? That doesn't fully capture what I am trying to ask but it is close.
I suppose what I really should do is just hook up a pair of headphones to an EQ and test this out, but I'm curious what you all have to say.
On another note, I'd like to thank everyone for the thought-provoking conversation so far. I have been experimenting with speaker placement after reading these posts, keeping in mind the effects of early reflections, and although I don't have any concrete conclusions, I certainly have some general ideas and have added a few more items to the set of considerations in designing and setting up a sound system.
For the most part there is a consensus that in order for a loudspeaker to sound good the sound it radiates in every direction needs to have a smooth and either flat or downward tilted frequency response, and the first copy of the sound coming in must be flat or slightly downward tilted. One thus designs a speaker such that the family of FRs in all directions are optimized together, with the listening window being approximately flat and the off-axis responses downward tilted. My question pertains to the off-axis responses that are downard tilted.
What I'm trying to ask is when we hear those reflections coming in, do we "hear" the general linear trend of the FR and then local anomolies are the deviations from that, or is it something else? Wow, unless you already have thought of this, it is probably not clear what I mean by that. A motivating example might help things. Suppose we use a woofer down low mated to a constant directivity device such as a horn. Off-axis the response is flat until the woofer narrows and then it is mated smoothly to the CD device where the directivity of the piston matches that of the device. There are no discontinuities in the response curves, and they are downward tilted, but what I am asking is do we "hear" the response in general as the line that would be fitted to the data and then deviations from that line are percieved as anomilies in the response just as anomolies from flat are heard in the listening window, or how exactly do we percieve this? I've attached a (crude) picture that may clarify what I am trying to ask. Maybe I can phrase this another way- is it bad for the off-axis responses to be concave up in any region? That doesn't fully capture what I am trying to ask but it is close.
I suppose what I really should do is just hook up a pair of headphones to an EQ and test this out, but I'm curious what you all have to say.
On another note, I'd like to thank everyone for the thought-provoking conversation so far. I have been experimenting with speaker placement after reading these posts, keeping in mind the effects of early reflections, and although I don't have any concrete conclusions, I certainly have some general ideas and have added a few more items to the set of considerations in designing and setting up a sound system.
Attachments
Rybaudio said:
In my opinion your premiss is wrong and hence the question has no answer. I DO NOT agree that an upward or downward tilt to the FR either on axis or off axis is acceptable. I claim that all the curves must be smooth AND flat and that they must have lower amplitudes as they progress off axis. This then makes your question mute doesn't it?
I'm curious as to how you could possibly contend that given you are promoting a loudspeaker that has a HUGE spectral tilt beyond 30 or so degrees off axis. If your waveguide has a 90 degree beamwidth, then at 45 degrees off axis you have a 6 dB downward tilt in the response of the system. Because your midwoofer is omnidirectional, at least in the front hemisphere, below about 500 Hz, the tilt gets more and more as you go further off axis.
It sure would be nice if we could have a speaker such as the one you are referring to, but other than a large array of small drivers, such as a large section of a spherical cap of densely packed very small drivers, is this practically possible? It seems to me that we have to make comprimises, and my question here is about one of those comprimises, not about what would be the ideal loudspeaker.
It sure would be nice if we could have a speaker such as the one you are referring to, but other than a large array of small drivers, such as a large section of a spherical cap of densely packed very small drivers, is this practically possible? It seems to me that we have to make comprimises, and my question here is about one of those comprimises, not about what would be the ideal loudspeaker.
I have attached a copy of the picture you attached in the other thread of the set of responses for your Summa system. It certainly looks to me like the curves are more and more downward tilted as you go further and further off-axis. For example, in the furthest curve (90 deg?) there is a ~26 dB drop from 100 Hz to 10 kHz. Is that not a spectral tilt? Maybe we are working under different definitions of spectral tilt.
Attachments
Rybaudio said:
This "tilt" as you call it, implies "beaming" which the Summa does do to a certain extent. But you must consider that virtually any other speaker available will "beam" even more than this. It was your comment "a HUGE spectral tilt beyond 30 degrees " that got my attention, simply because the "spectral tilt" of the Summa is less than virtually any other loudspeaker made. So where does your comment come from?
You obviously have not seen polar maps for many other speakers if you can think that the Summa has a "HUGE (Bold in your original) spectral tilt. You cannot use 100 Hz as a basis, since that is below the waveguides usage. Use 1000 Hz. to 10 kHz and compare this difference with any direct radiator loudspeaker. This is the frequency range that matters and in this range I would put the Summa up against any loudspeaker that you want to name for "spectral flatness" across the polar range.
Perhaps we are working from different deffinitions, Mine are based on what matters subjectively not some "concept" of perfection.
Rybaudio said:
that is very easy to achieve with an omnidirectional design and very difficult to achieve with conventional boxes with forward firing drivers
Rybaudio said:
the point is that we NEVER hear those reflections (approximately 2 ms<50 ms) as such
what we hear is total SOUND and in normal room and with normal speaker placement those reflections above 2 ms contribute to the perceived level of sound (SPL) that is amplitude/frequency response and to nothing more
perhaps it would be even better to say "time-locked" instead of "phase-locked"
and IF off-axis response is uneven and/or there are nonlinear sound absorbers on room walls then we get also coloration because not all frequencies are amplified equally by the room reflections
best,
graaf
gedlee said:
That comment was in response to your comment that "I DO NOT agree that an upward or downward tilt to the FR either on axis or off axis is acceptable." I just thought it was strange that you would say that given the design of your speaker. I'm guessing that you meant that comment to apply strictly to the 1-10 kHz region.
If the 1-10 kHz region is considered, yes I suppose it is less than virtually any speaker out there. I was thinking globally (100-10k), where the wide radiation of the dome tweeter would give less tilt. For example, if you mated a 1" dome with a 4" midrange at 2.5 kHz and that to a 15 at 400 Hz you would have a system that at 45 degrees off axis was flat to about 5k and then gently sloped down to about -3dB at 10k, which corresponds to less "global" spectral tilt than the Summa (-3 dB as opposed to -6 dB), but more tilt in the 1-10k region. I guess I am just used to thinking of the full 100-10k and this affects the language I use to describe the behavior of the speaker.
gedlee said:
I have worked for a Harman company on a couple occasions and had access to their measurement database. Does 500+ speakers count? Sorry to be a wise-a$$
In all honesty though, I only examined maybe 200 of them. I made the HUGE comment just because you said no upward or downward tilt is acceptable, and compared to flat, a 6 dB tilt seems like quite a bit to me (imagine a speaker that had that slope on-axis).gedlee said:
Sure; save for a really small tweeter that is omnidirectional past 10k, it looks as consistent as can be to me. My main concern is this range in the context of the entire spectrum. That gets back to the original question of mine, and if I'm understanding you correctly, your answer is to forget about the global linear trend and make sure 1-10k is flat, even if it means the power response of the speaker has a shape like that I posted above which is not a downward sloped line but a couple regions that are flat connected by a slope. For example, in your speaker your off-axis responses from 1-10k are flat, as opposed to putting a say 2-3 dB boost in the 1.5-2k region (not that you would want to do that anyway because it would put a sizable peak on axis) so that the off-axis responses more closely fit a downward sloped line in the context of the response from 100-10k Hz? Am I being clear what I am asking? Maybe it just doesn't make any sense.
I guess I just see closer to a line as smoother. If I remember correctly, Toole and Olive put out a paper in 2004 regarding a metric they developed for predicting subjective preference based on their measurements (direct, list wind, first refl, power, dir ind), and fits of these curves to lines (good if the curve fits a line well, and good if the line is shallowly sloped, but the fits were global). Maybe I just read the paper, heard them talk about it, and have continued to think that way since. I'll have to think about this more and take it into account in my future observations. Thanks.
graaf said:
I'm not sure that it was clear what I meant by "hear" in that case. I did not mean that I can discern the sound as a reflection. I meant that it affects my perception of the overall tonality of the system, and based on your response I think we agree on that.
My question in the above post is essentially this- we can't make the off-axis response of the speaker globally flat. What shape should it have? Dr. Geddes as far as I can tell is reccomending that the highest priority be given to the 1-10k region where we want the off-axis response be as flat as possible, and to drop off amplitude wise as we go further and further off-axis. What I was trying to point out is that this leads to a power response that falls and then has a sort of plateau from 1-10k. My question is when we hear this do we percieve it compared to a downward tilted line (as in the above drawing) or not?
Rybaudio said:
surely we do
Rybaudio said:
my answer is that it is not at all necessary
what we need is "flat enough" plus much more important horizontal and vertical polar response "general pattern" plus most important "correct positioning" of the loudspeakers in the room
Rybaudio said:
perhaps something like that: http://www.timedomainusa.com/
or very similar Linkwitz Pluto
or perhaps something in Stig Carlsson style
post scriptum
and perhaps we need something like this: http://www.enjoythemusic.com/magazine/equipment/0705/behringerultracurve2496.htm
and perhaps we need something like this: http://www.enjoythemusic.com/magazine/equipment/0705/behringerultracurve2496.htm
post post scriptum
it is important not to fear the cheap brand name "Behringer"
DEQ and DCX from them is pure hi-end when put between CD-transport and DAC
check this:
http://www.overkillaudio.com/products.html
http://www.6moons.com/audioreviews/overkill/encore.html
quite impressive, isn't it?
it is important not to fear the cheap brand name "Behringer"
DEQ and DCX from them is pure hi-end when put between CD-transport and DAC
check this:
http://www.overkillaudio.com/products.html
http://www.6moons.com/audioreviews/overkill/encore.html
quite impressive, isn't it?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.