In thinking about this I think the problem comes from my tendency to only talk about modal issues in the steady state and to only talk about transient effects and singular reflections above the Schroder Frequency. This means that a 1/4 wave cancellation does not occur in the modal region, because in the "steady state" it has to go on and contact all the other walls becoming a "mode". So the points about 1/4 wave cancellation have no meaning in the modal region. And I think that discussing anything "transient" in the modal region is pointless because there can't really be a transient at these frequencies or at least we cannot perceive any effects at these frequencies as transients. Below 200 Hz the ear is becoming a very long time integrator in order to even perceive these long period tones.
At 100 Hz, the period is 10 ms. It takes at least this long for our hearing to even detect that such a tone is present let alone any ability to resolve it or reflections of it. We hear LFs in the steady state, thats pretty obvious.
In the geometrical acoustics region the 1/4 wave discussion is a valid concept, but then it's no different than treating any other reflection regardless of its frequency domain implications. And these discussions have already been well hashed out.
At 100 Hz, the period is 10 ms. It takes at least this long for our hearing to even detect that such a tone is present let alone any ability to resolve it or reflections of it. We hear LFs in the steady state, thats pretty obvious.
In the geometrical acoustics region the 1/4 wave discussion is a valid concept, but then it's no different than treating any other reflection regardless of its frequency domain implications. And these discussions have already been well hashed out.
The discussion of multiple subs is to effectively address room issues in the modal region based on the assumption that above the modal region room is no longer an issue. But I have just became battled with the problem just above the modal region.
When measuring the midwoofers at 1.5 metre distance using 11ms and 22ms window the response is quite flat and smooth (+/-3dB at most) from 100Hz to 1kHz.
When measured at actual sitting position with the speaker at actual speaker position with the same 11ms and 22ms window, there is a consistent broadband 10dB null from 300 Hz to 500Hz. The speaker drivers are 1.8 metre from the front wall (10ms) and effectively has no side walls (far exceeding the 11ms window).
I would not expect the null being so deep and so broad. This null may even be worse than the peaks and dips in the modal region because our ears are more sensitive from 300Hz to 500Hz. There is no way the speakers can sound right with such a large broadband null.
What would cause such a broadband null just above the modal region? I have laid 60cm wool batts and pillows on the floor, rear walls, etc, and the null is still there. I also did all sort of room simulation modellings and have not found what causes it.
So it does not seem to be correct for me to assume that room is not an issue above the modal region.
Regards,
Bill
When measuring the midwoofers at 1.5 metre distance using 11ms and 22ms window the response is quite flat and smooth (+/-3dB at most) from 100Hz to 1kHz.
When measured at actual sitting position with the speaker at actual speaker position with the same 11ms and 22ms window, there is a consistent broadband 10dB null from 300 Hz to 500Hz. The speaker drivers are 1.8 metre from the front wall (10ms) and effectively has no side walls (far exceeding the 11ms window).
I would not expect the null being so deep and so broad. This null may even be worse than the peaks and dips in the modal region because our ears are more sensitive from 300Hz to 500Hz. There is no way the speakers can sound right with such a large broadband null.
What would cause such a broadband null just above the modal region? I have laid 60cm wool batts and pillows on the floor, rear walls, etc, and the null is still there. I also did all sort of room simulation modellings and have not found what causes it.
So it does not seem to be correct for me to assume that room is not an issue above the modal region.
Regards,
Bill
In that frequency region, it is not a modal issue. It has to be a reflection from somewhere. The key is to look at the time domain to determine what the timing of the reflection is. Then you should be able to find the culpret. I'm betting on floor or ceiling. Thats the same problem that I had until I killed the ceiling reflection.
My post may have been misleading in that what I mean is that only a steady state approach makes any sense in the modal region. In the steady state ALL of the walls have to be considered. There is just no other way to look at the problem. Above the modal region we can consider individual reflections, but not at LFs.
If, somehow (like outdoors), you could set up a steady state situation with only one reflection then you could talk about what happens, but not in a room, its just not a useful concept or discussion.
And, of course, you can always hypothesize about what would happen in a hypothetical situation that is not realistic, but what's the point in that?
If, somehow (like outdoors), you could set up a steady state situation with only one reflection then you could talk about what happens, but not in a room, its just not a useful concept or discussion.
And, of course, you can always hypothesize about what would happen in a hypothetical situation that is not realistic, but what's the point in that?
Earl,
you say that the 1/4 wave dip "doesn't occur" below the Schroeder frequency. Have you tried the basic practical experiment which allows to verify (or disprove) this theory? I just wonder because I can measure this dip in the back of my room. Also it is been mentioned many times by control room designers and the likes.
The fact that every room has more than one boundary just doesn't imply that a single boundary in a room doesn't have an effect on its own.
Greets,
Jules
you say that the 1/4 wave dip "doesn't occur" below the Schroeder frequency. Have you tried the basic practical experiment which allows to verify (or disprove) this theory? I just wonder because I can measure this dip in the back of my room. Also it is been mentioned many times by control room designers and the likes.
The fact that every room has more than one boundary just doesn't imply that a single boundary in a room doesn't have an effect on its own.
Greets,
Jules
You are simply getting things all mixed up. I never said that the 1/4 wave dip doesn't occur, its just that it isn't reasonable to talk about such a thing below the Schroeder frequency.
How would you test your claim? Think about it. How could you issolate a single wall reflection in a small room in the modal region? Gate it? How do you gate below 100 Hz? OK, don't gate it (because its not possible at those frequencies) then you have all the reflections and the dip is really a mode.
You're just not thinking this through and thinking about the problem as if its a HF one where geometrical acoustics applies. A classic mistake.
As to what "others" say, it isn't the first time I don't agree with the "experts". Many of the "experts" are still saying that multiple subs won't work and that modes are bad things that should be eliminated. The vast majority of information out there is simply wrong, so of course I don't agree with it.
How would you test your claim? Think about it. How could you issolate a single wall reflection in a small room in the modal region? Gate it? How do you gate below 100 Hz? OK, don't gate it (because its not possible at those frequencies) then you have all the reflections and the dip is really a mode.
You're just not thinking this through and thinking about the problem as if its a HF one where geometrical acoustics applies. A classic mistake.
As to what "others" say, it isn't the first time I don't agree with the "experts". Many of the "experts" are still saying that multiple subs won't work and that modes are bad things that should be eliminated. The vast majority of information out there is simply wrong, so of course I don't agree with it.
gedlee said:
gedlee said:
That to me looks slightly contradictory, but I'm glad we can now agree that this effect does occur. If it does occur, it is an example of a non-modal effect below the Schroeder frequency. And that's why I could only deduce from your paper that either it is an incomplete description of the situation within a room or your definition of mode is unusual.
This will sound like a broken record to you anyway. But maybe you can now see where I'm coming from and I don't think I'm the only one who had this question reading your paper.
Greets.
Earl,
I think you are thinking too much about modes and not about the physics. Remember that if a frequency is corresponds to a mode then it simply means a standing wave solution exists for that frequency. However, if the frequency is not a mode then the wave still bounces back and fourth off the walls and decays without generating a standing wave. Still, assuming plane wave radiation and steady state sine wave excitement, the wave reflected off the wall will be 180 degrees out of phase with the incident wave when the distance is 1/4 wave length. And if we add subsequent reflections, the reflection of those reflections will also be 180 degrees out of phase when the distance off the wall is 1/4 wave length. The problem in 3-dimension becomes more complex because reflections from side wall and floor and ceiling may alter the depth of the null.
Here is a simulation when only modes in 1-D are considered and the "mic" is 1 meter off the wall.
Clearly a null at 84 Hz
I think you are thinking too much about modes and not about the physics. Remember that if a frequency is corresponds to a mode then it simply means a standing wave solution exists for that frequency. However, if the frequency is not a mode then the wave still bounces back and fourth off the walls and decays without generating a standing wave. Still, assuming plane wave radiation and steady state sine wave excitement, the wave reflected off the wall will be 180 degrees out of phase with the incident wave when the distance is 1/4 wave length. And if we add subsequent reflections, the reflection of those reflections will also be 180 degrees out of phase when the distance off the wall is 1/4 wave length. The problem in 3-dimension becomes more complex because reflections from side wall and floor and ceiling may alter the depth of the null.
Here is a simulation when only modes in 1-D are considered and the "mic" is 1 meter off the wall.
An externally hosted image should be here but it was not working when we last tested it.
Clearly a null at 84 Hz
markus76 said:
What do you think?
Would you rather continue the argument as to whether 1/4 wave cancellation is a real effect or not and whether it is a transient or steady state result? There seems to be a growing propensity to argue about things here for the sake of who is right or wrong rather that make the effort to understand and demonstrate the physical reality of the situation. If you don’t understand the simple case how would you ever gain any insight into the more complex situation?
Yes Earl, correct. The point of my post is to encourage thought. The questions which were bouncing around were whether 1/4 cancellation occurred, whether it was a steady state or transient result, and how could a single wave in the modal region be looked at. The calculation I presented shows the effect when only considering a single axial direction, say the X direction. The listener or mic would be in a YZ plane 1/4 wave length (1 M in this case) from the wall opposite the speaker. The contributions of the response from the axial modes in the X directions show the 1/4 wave null resulting from reflections off the walls which are YZ planes. This null may also appear when modes in the Y or Z direction only are considered and the mic is in an XZ or XY plane 1/4 wave length from the appropriate wall.
Going back to the original problem where the mic is 1/4 wave length from a wall in an YZ plane, when we introduce the remaining axial modes in the Y and Z directions, and tangential and oblique modes, then it becomes a matter of how much do these modes contribute to the response at the 1/4 wave frequency associated with the listening position and the wall 1/4 wave length behind it. Under these conditions the presence or absence of the 1/4 wave null, and its depth will be a function of where exactly in the YZ plane the mic is located as these additional modes can fill in the null at different points in this plane. Add in differences in wall damping, .... and the problem becomes more complex and difficult to predict. But the possibility of a 1/4 wave null is real and it is a steady state, modal phenomena.
That is the nice thing about a simple analytical tool like the one I have. Perhaps I can not predict the exact result for the in room response, but I can ask what if questions and look at what can or can not happen and quickly gain insight as to where problems might arise, or what the sources of observed problems are.
Going back to the original problem where the mic is 1/4 wave length from a wall in an YZ plane, when we introduce the remaining axial modes in the Y and Z directions, and tangential and oblique modes, then it becomes a matter of how much do these modes contribute to the response at the 1/4 wave frequency associated with the listening position and the wall 1/4 wave length behind it. Under these conditions the presence or absence of the 1/4 wave null, and its depth will be a function of where exactly in the YZ plane the mic is located as these additional modes can fill in the null at different points in this plane. Add in differences in wall damping, .... and the problem becomes more complex and difficult to predict. But the possibility of a 1/4 wave null is real and it is a steady state, modal phenomena.
That is the nice thing about a simple analytical tool like the one I have. Perhaps I can not predict the exact result for the in room response, but I can ask what if questions and look at what can or can not happen and quickly gain insight as to where problems might arise, or what the sources of observed problems are.
gedlee said:
I'll do some when I'm back home in about a week.
@ john k...
Thanks for chiming in. The possibility of a deep null exists, but other reflections occur at other boundaries, and they combine in and out of phase to support or reduce the 1/4 wave dips. To create a deep null, the opposing wave fronts must be nearly identical in level. It takes very little contribution from an errant reflection arriving from somewhere else to disturb the balance needed for a total cancellation.
The only thing I wonder is why you call it a modal phenomenon, although the phenomenon is not related to resonance. Is every frequency below the Schroeder frequency a "modal frequency"? I'd say no. Only some of the frequencies are and they are determined by the room dimensions. The 1/4 wave dip was meant as an example for the existence of "non-modal" dips and peaks in every room.
Greets
john k... said:
I think it's always good to educate people. But what people tend to take away from these theoretical views is that there are optimal room dimensions, 38% placing rules, 1/4 wavelength sub placement rules, etc. In real rooms this all doesn't work if it's optimal sound reproduction that you're after. Only measurements can help. Software simulation of real rooms is possible but it's just not available (yet).
Best, Markus