I could not disagree with you more. And I would suggest that if "it sounds good to me" is your bottom line, then you are in the wrong thread - or I am. How something sounds to any given individual is not only irrelavent, but pointless. We are talking about the science of reproduction here based on well founded principles and psycho-acoustic research. We are not talking about handbags here and preference is not the point. (Well it is to some of course, but most of the interesting discussion here involves perception, not preference.)
No, there is no "Holy Grail of perfect loudspeaker polar response", but there certainly is "better" and "worse". The lack of "perfection" is no justification to stop seeking it.
Here is a typical preference thread:
"I like this ...."
"Well I like this ..."
"Oh, I love that too ..."
"But I like this ..."
"JC likes this ..."
"JC really knows his stuff ..."
And on and on. Thanks but I'll stay out of that kind of thread.
Last edited:
Of course low midrange affects imaging in a small room. But it will become obvious only for those blessed souls who have done extensive listening experiments in a small room with high directivity sources like dipole line arrays, like me.
Almost for the last 10 years I've had dipole line arrays (below about 1kHz) and as well and dipoles and monopoles side-by-side in my small room, and the result is clear.
Below about 1 kHz in a small room (i.e. a living room) the imaging with monopoles is poor, dipole are better but dipole line arrays are best.
Below 500 Hz the difference gets even bigger.
This follows directly from the directivity pattern of the speakers below 1 kHz.
It is not a question whether it is practical to use high directivity sources below 1 kHz, but it is the way it is in terms of performance.
- Elias
Yes interesting
The ideal cardioid source is a nature of constant directivity, while the KEF is less so. Thus the results suggest it would be a bad idea to use CD speakers toed in in front of the listener as this causes coloration due to strong contralateral reflection. The optimum would be very opposite from what some people are using
- Elias
Yes interesting
The ideal cardioid source is a nature of constant directivity, while the KEF is less so. Thus the results suggest it would be a bad idea to use CD speakers toed in in front of the listener as this causes coloration due to strong contralateral reflection. The optimum would be very opposite from what some people are using
- Elias
The KEF is also probably more directional at mid/high frequencies. Crossing a fairly directional speaker (dipole, cardioid) has not worked for me. In that situation the contralateral reflection is pretty strong while other reflections are mostly absent.
Me too - with an AKG shotgun mic. Also more HF content than I expected. In my case it was all down pretty low, but a decent mirror of the direct sound.
Which got me to wondering: If one wanted to measure the sound at the front wall, what would be a good approach? You could put a PZM mic right there to see what's hitting the wall. Or another type of mic very close. But what might be a good way to compare it to the direct sound? Maybe normalize it to the 1M response on axis?
What do you think? Is there any standard for this?
Gedlee,
Not taking offense to your comment to me I will add to what you are saying. Back to science here then. I love when I read simulating a cardioid or such nonsense. Is it cardioid or isn't it, how are we simulating something to test that isn't what we are testing? What I should really be saying is that to me anyway, unless the test is specific, an exact speaker implementation with exacting knowledge of that device the test is flawed. Going to science I don't like and have been taught not to make to broad of a generality. No two waveguides made by two different methods will have the exact same polar response, so I would have to know exactly what that one device was doing to see the real effects to a specific test. I can no more test one of your elliptical waveguides and say that a JBL or EV horn will give me the same results in the same room. Same for a pair of direct radiators from two different manufacturers or any other implementation. There are to many variables going on here. Change one variable at a rime and I will be able to make sense of the changes and see the resultant. But to willy nilly say that all speakers in a class perform the same just aint so. So all we are speaking here is in generalities it seems.
Not taking offense to your comment to me I will add to what you are saying. Back to science here then. I love when I read simulating a cardioid or such nonsense. Is it cardioid or isn't it, how are we simulating something to test that isn't what we are testing? What I should really be saying is that to me anyway, unless the test is specific, an exact speaker implementation with exacting knowledge of that device the test is flawed. Going to science I don't like and have been taught not to make to broad of a generality. No two waveguides made by two different methods will have the exact same polar response, so I would have to know exactly what that one device was doing to see the real effects to a specific test. I can no more test one of your elliptical waveguides and say that a JBL or EV horn will give me the same results in the same room. Same for a pair of direct radiators from two different manufacturers or any other implementation. There are to many variables going on here. Change one variable at a rime and I will be able to make sense of the changes and see the resultant. But to willy nilly say that all speakers in a class perform the same just aint so. So all we are speaking here is in generalities it seems.
You clearly haven't read Bech's papers, have you? His experiment was very well prepared and executed. Please read them and then return with your comments, but don't comment on something you have not read.
Glad you asked, they're very interesting material
Timbral aspects of reproduced sound in small rooms. I
Timbral aspects of reproduced sound in small rooms. II
There's a paper on spatial aspects as well
Spatial aspects of reproduced sound in small rooms
Timbral aspects of reproduced sound in small rooms. I
Timbral aspects of reproduced sound in small rooms. II
There's a paper on spatial aspects as well
Spatial aspects of reproduced sound in small rooms
a_tewinkel,
Thank you for those references. I would much rather read them myself than have a second hand interpretation of the paper. I will comment after I read them and digest the information.
Thank you for those references. I would much rather read them myself than have a second hand interpretation of the paper. I will comment after I read them and digest the information.
There are simulations and then there is reality. They are never the same, but then no one doing simulations should expect them to be. Simulations get us "in-the-ballpark" and this saves a lot of time over "cut-and-try". The better the simulation the more time is saved. I have spent my life doing this and I know that sims are never perfect, but then tests are never perfect either.
"Perfection is the enemy of good" - something every engineer needs to understand.
Your previous post sounded a lot like "Speakers can't ever be perfect so why try?", which I have heard way too often. I don't buy that argument. Perfection is the unobtainable, but there certainly is "better" and "best", and it makes no sense to run away from this reality. One CAN define "better" and "best" subjectively, but if you have been in this business as long as I have you will realize that defined subjectively there really isn't an answer - its not a stable factor. Its different for everyone. But defining them in some quantitative manner and we CAN get to "better" and "best". It actually becomes possible to get somewhere and not just go around in circles.
Unfortunately this is what we're (still) doing, going round in circles, because there isn't any meaningful standard from which "better" and "best" could be derived from. It's not enough if one manufacturer says "I am the standard" - or maybe it is
Last edited:
NO: "I AM THE Standard."
This is a good point. It might be good to review the aspects of this discussion that are clearly defined and supported by data and those that are not. For example, I don't think that anyone would disgree that "it would be better if the floor and ceiling reflections were surpressed", Hence "best" would be that design which does this. Of course there are many other factors and they have to be weighted in subjective importance, but this is not impossible. I have my rankings and I try and justify those based on solid principles (although not everyone agrees). So I have no trouble with "better" and "best".
Where I come loose is adding the "if it sounds good it is good" factor. Because, as you know, this has no meaning and never can.
In rereading the Bech paper it became apparent that his results are room specific. In other words the floor and ceiling were the most audible in that room. His room had very high wall absorption and very low floor and ceiling absorption - the exact opposite of what I do and recommend. This makes it difficult to apply his results. What is clear in the paper is how the image shift has much lower jnd than the coloration. It is image that I am most concerned with, I suppose since it is more sensitive to the speaker and room design than just timbre. Once coloration has been solved, imaging becomes the key aspect.
Indeed, and the simulated source was pretty far from the front and side walls (or at least further than would be the case in most European living rooms).
Gedlee and Tewinkel,
Now I think we are more on the same page going forward. Tewinkel I have the papers and will print them to read the whole experiment. At the same time I think that we can say that the results that Bech's analysis elucidated are room specific to his experiments. What most of us seem to believe, perhaps speaking out of place here, is that the ceiling and floor reflections do have a measurable and sometimes deleterious effect on the imaging of the sound field just as the side wall reflections do also. How we weight those contributions is something that I have not studied so can not make a judgement call. At the same time I think that we should make certain judgements based on as much fact as possible. Pick a particular loudspeaker implementation in each category such as horn loaded, dipole and such and use that particular device as a baseline for comparison reasons only. Let's say we use Earl's waveguides as the base for that as he would know the polar response of the devices in question and pick a known dipole and do the same. Then we can scientifically speak of the changes made. Each reference type would remain the same throughout the discussion, that takes the I like my sound better statements out of the discussion. We need standards to do this kind of work or it is just a discussion of opinions. We all know what the best sound is in a room, a live performance. Then it gets back to opinion again as one person like one performer and another likes someone else, and someone else say I like the Martin guitar better than a so and so guitar. We need to take some of the variation out of the equation to discuss this further. Anyone else have an opinion on this?
Now I think we are more on the same page going forward. Tewinkel I have the papers and will print them to read the whole experiment. At the same time I think that we can say that the results that Bech's analysis elucidated are room specific to his experiments. What most of us seem to believe, perhaps speaking out of place here, is that the ceiling and floor reflections do have a measurable and sometimes deleterious effect on the imaging of the sound field just as the side wall reflections do also. How we weight those contributions is something that I have not studied so can not make a judgement call. At the same time I think that we should make certain judgements based on as much fact as possible. Pick a particular loudspeaker implementation in each category such as horn loaded, dipole and such and use that particular device as a baseline for comparison reasons only. Let's say we use Earl's waveguides as the base for that as he would know the polar response of the devices in question and pick a known dipole and do the same. Then we can scientifically speak of the changes made. Each reference type would remain the same throughout the discussion, that takes the I like my sound better statements out of the discussion. We need standards to do this kind of work or it is just a discussion of opinions. We all know what the best sound is in a room, a live performance. Then it gets back to opinion again as one person like one performer and another likes someone else, and someone else say I like the Martin guitar better than a so and so guitar. We need to take some of the variation out of the equation to discuss this further. Anyone else have an opinion on this?
Last edited:
Like I already said (do you guys read all the posts?) I don't know where it came from.Which research is this? Source please
I did however find the interview, which is from 2004. And I was wrong. I thought the ceiling reflection was said to be the most detrimental and floor the second worst. It was the other way around. Anyway, here's Peter's own words:
Gedlee:
Do you have an impulse response or a ETC showing high gain reflections arriving directly from the front wall and absorbed afterwards with your speakers?
I would state that, in a typical situation, the reflections that would have highest priority reduce would be the floor, ceiling and front wall.
I think it is also likely that in some situations the ceiling may have a larger effect while in others the front wall may have a larger effect, depending on the room height, listening distance, speaker placement and speaker directivity.
Side wall reflections should be reduced if they are very early, < 5 ms (more or less), otherwise it may be a matter of preference (as Toole suggests).
I think this shows that the optimal solution will be different depending on the situation, especially if a listener wants image broadening etc. For example a speaker may have a preferred amount of early reflections and image broadening in one room, while it may have too much of it (and added colouration) in another (smaller) one.
If strong direct sound and thus weak side wall reflections are desired, for example for monitoring, the solution is probably fairly straightforward (and fairly directional both vertically and horizontally).
I think it is also likely that in some situations the ceiling may have a larger effect while in others the front wall may have a larger effect, depending on the room height, listening distance, speaker placement and speaker directivity.
Side wall reflections should be reduced if they are very early, < 5 ms (more or less), otherwise it may be a matter of preference (as Toole suggests).
I think this shows that the optimal solution will be different depending on the situation, especially if a listener wants image broadening etc. For example a speaker may have a preferred amount of early reflections and image broadening in one room, while it may have too much of it (and added colouration) in another (smaller) one.
If strong direct sound and thus weak side wall reflections are desired, for example for monitoring, the solution is probably fairly straightforward (and fairly directional both vertically and horizontally).
Unless you are talking about unusual heigh ceiling, I would say no. A few milliseconds in arrival time of difference doesn't really change much.
Sloping ceiling. Depends on the height and how it slopes. In most cases I believe you will get same the amount of early reflections. I have treated a few rooms with sloping ceiling. Seemed to be the same to me.
I almost never see that frontwall needs to be treated for specular reflections. Not if the rear of the room is correctly treated. With dipoles, one would definetly need to though.
thanks again Omholt!
VERY well said! I cannot agree more!
...or a floor coupled up-firing speaker?
ceiling reflection in case of such a speaker is very different from a ceiling reflection which Mr Lyngdorf considers second worst
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.