Of course this is a subjective opinion: Two subs placed in the center of two opposing walls are really good sounding for me. Only a little eq is needed to get a flat response and lots of output in a reasonable sized listening area. I tried random placement as well, but it didnt seem to achieve so good results. The ease of use of these symmetrical placements may be beneficial here.
Maybe i should add that i have only one amp for the subs, so i couldnt adjust phase individually for each driver. Maybe thats a reason why random placement didnt work well.
Maybe i should add that i have only one amp for the subs, so i couldnt adjust phase individually for each driver. Maybe thats a reason why random placement didnt work well.
Hi Jacobus,
I measured them once but do not know if calibrated or not. So I better don't post any diagrams. But I'll definitely measure them again when the whole system is finished.
The subs are about that size:
An externally hosted image should be here but it was not working when we last tested it.
Best, Markus
MaVo said:
I've said many times that there aren't and cannot be any hard and fast rules as everything is room dependent and no two rooms are identical. The more subs there are, regardless of location, the better.
Mathematically I can show that symetrical locations will not work as well as nonsymmetrical ones. I tried to explain why that is.
"The benefits of the one or two pairs on opposite mid-walls has nothing to do with numerous uncorrelated sources and everything to do with the unique characteristics of those locations."
Unfortunately Noah's statement is incorrect since the "unique characteristics" are exactly what makes the sources "correlated".
Don't get fixated on location as its always a secondary effect to the number. Being able to adjust the magnitude and phase to each sub, IME is also more important than location. I simply put the subs whereever its convenient and then adjust them for those locations.
"Unfortunately Noah's statement is incorrect since the "unique characteristics" are exactly what makes the sources "correlated".
Don't get fixated on location as its always a secondary effect to the number."
You seem to be insisting that the only mechanism possible is randomness of sources.
Do you disagree that two subs on opposite parallel walls have out of phase output at freq of odd-order modes?
If not, how is that random?
Don't get fixated on location as its always a secondary effect to the number."
You seem to be insisting that the only mechanism possible is randomness of sources.
Do you disagree that two subs on opposite parallel walls have out of phase output at freq of odd-order modes?
If not, how is that random?
noah katz said:
I never look at the LF room problem using the assmued modes of a rectangular room as these are too idealized for the real world. I look at ALL rooms as having a basically random location of modes and I would suggest that to do otherwise is not useful. Real rooms have furniture, doors, walls that arn't rigid, AC vents, a whole array of random features that tend to randomize the modes. Now to the extent that two sources are symmetric, then they will be correlated and correlated sources don't act as independent sources.
The answer to your question is "No I don't disagree" within this limited context, but that has no bearing on the situation. It only shows that these two sources are highly correlated.
Perhaps where we disagree is that you seem to be taking the classical position that cancelling the modes is what one seeks, but I take the opposite position, that exciting the modes is what we seek. But more than just exciting them, its important that they not be excited in exactly the same way. The wall placement that you suggest tends to cancel some modes and highly accentuate others. Not the way to achieve maximal flatness of response.
Point taken on real vs. ideal.
"Perhaps where we disagree is that you seem to be taking the classical position that canceling the modes is what one seeks, but I take the opposite position, that exciting the modes is what we seek."
Given the wide freq spacing of the lower room modes (a whole octave between the first two), how could exciting them give smoother response than not exciting them?
"The wall placement that you suggest tends to cancel some modes and highly accentuate others."
A valid point. In my case, there's no other choice than wall location (actually right behind the wall), so I might as well get the cancellation that I can.
Actually I don't know yet how well it will work; only the sub on the back wall is done, and it's near a doorway opening into the rest of the house.
I'm hoping for a nice improvement from elimination of the 3rd mode at 75 Hz.
"Perhaps where we disagree is that you seem to be taking the classical position that canceling the modes is what one seeks, but I take the opposite position, that exciting the modes is what we seek."
Given the wide freq spacing of the lower room modes (a whole octave between the first two), how could exciting them give smoother response than not exciting them?
"The wall placement that you suggest tends to cancel some modes and highly accentuate others."
A valid point. In my case, there's no other choice than wall location (actually right behind the wall), so I might as well get the cancellation that I can.
Actually I don't know yet how well it will work; only the sub on the back wall is done, and it's near a doorway opening into the rest of the house.
I'm hoping for a nice improvement from elimination of the 3rd mode at 75 Hz.
markus76 said:
Markus,
thanks for the answer. What about the maximum output of these subs? you have an idea? do they run out of air long before the Nathans will do especially considering the boosting? I heard good things about the peerless drivers, I'm very leaning to the option of building 2 subs with them.
Jacobus
noah katz said:
Noah
Boy I wish I had the energy to discuss why one wants to excite the modes instead of cancel them, but this has been discussed so many times before and its not a short "lecture". IF I were to do this (again) I would want it as a seperate thread since its not germain here.
Make no mistake about it this is a very very important issue, and one where I have been arguing with seemingly the whole world about, so don't feel alone, but please, not here.
"I'm hoping for a nice improvement from elimination of the 3rd mode at 75 Hz." - I'd expect this to make it sound boomy.
Jacobus said:
Another advantage of multiple subs is that they don't need to be very powerful. One achieves sufficient output through quantity.
But I would say that in a sub the driver particulars and not very important if they are important at all. In a direct radiator distortion might be an issue at very high levels, but in the bandpass that I use the acoustic LP filtering makes even this distortion irrelavent.
Nonlinearity perception is a bandwidth issue. Its the modulation of the upper frequencies by the lower ones that is audible if the distortion is audible at all. A limited bandwidth device, such as a sub, has an inherent advantage in this regard and with a bandpass its simply not even a consideration. I'd just use Paudio drivers - total junk, but it doesn't matter.
"Boy I wish I had the energy to discuss why one wants to excite the modes instead of cancel them"
Could you point me to where the case is made that exciting widely spaced low-order modes gives smoother response?
I've not seen this before, and I thought I was reasonably well-read on the subject.
I realize that it's not an issue at higher freq where mode density is high.
"I'd expect this to make it sound boomy."
At this point I can only guess - do you mean having the mode or canceling it?
Could you point me to where the case is made that exciting widely spaced low-order modes gives smoother response?
I've not seen this before, and I thought I was reasonably well-read on the subject.
I realize that it's not an issue at higher freq where mode density is high.
"I'd expect this to make it sound boomy."
At this point I can only guess - do you mean having the mode or canceling it?
So this would be the purpose of a decorrelation filter?
Given the autocorrelated nature of of music signals, symmetrical sub placement should work against this purpose since, in effect, it "filters" for correlation - yielding cancellation and reinforcement effects.
Not insuperable difficulty but an added complication since the goal is smooth distribution of as many modes as possible.
Thus your earlier comment:
I think in slogans:
The object is to fill in the blanks between the room modes.
Given the autocorrelated nature of of music signals, symmetrical sub placement should work against this purpose since, in effect, it "filters" for correlation - yielding cancellation and reinforcement effects.
Not insuperable difficulty but an added complication since the goal is smooth distribution of as many modes as possible.
Thus your earlier comment:
I think in slogans:
The object is to fill in the blanks between the room modes.
I mean cancelling it. This will leave a dead zone flanked by what will likely be very pronounced modes or sets of modes. The bass will sound dull or missing unless you turn it up, but then it will sound boomy. With a hole like this in the response there would not be much that you could do to make it sound right.
My whole goal in LF design is to maximize mode excitation, interaction and damping. This is the way to get bass that sounds right.
You must not have read much of my work as I drive this point home (that modes are our friends) as often as I can. It's in both of my books and my Summa white paper.
My whole goal in LF design is to maximize mode excitation, interaction and damping. This is the way to get bass that sounds right.
You must not have read much of my work as I drive this point home (that modes are our friends) as often as I can. It's in both of my books and my Summa white paper.
FrankWW said:
In some major studies that I did to try and find out why bass sounds so good in big rooms and typically so bad in small ones I found that the large room has its LF signals decorrelated by the reflections and reverberation (modes in essence). Everybody knows that reveberation is uncorrellated with the direct sound. At LF's the period of the waves and certainly the time it takes for our hearing to react and recognize a LF signal are actually on the order of the reverb time in a small room. Thus there is no decorrelated signal for our hearing to process at LF in a small room, its basically completely correlated. As I studied this situation I came to realize that it was the modes in the large room that decorrelated the LFs and extended the time of their presence thus yielding a psychoacoustic gain in their perception.
I thought that if we could do this electronicaly at LF for the subs then maybe we could get the sound more like the large room. Turns out that the filter to do this is called a decorrelation filter - although I did not recognize that at first. Thats why I now look at the whole LF problem as one of correlated versus uncorrelated sources. To me this is the important thing. Thats how I realized that random placement would be advantageous over symmetrical placement. This has been shown to be true both theoretically as well as subjectively.
"I mean canceling it. This will leave a dead zone..."
Ah, a light bulb just went on - for some reason I've been thinking that only the modal peaks/nulls would be eliminated, not that freq altogether.
You're right, there will certainly be holes in the response at that freq.
However, these will be maximum only right at the wall and there will be progressively less cancellation moving away from the walls.
I'm thinking in my situation where the seats are relatively central, the slight cancellation will be a lesser evil than the modal peaks and nulls.
Oops, sorry, you said you didn't want to continue with this here, but the the wheels started turning.
Thanks, good food for thought.
"You must not have read much of my work as I drive this point home (that modes are our friends) as often as I can. It's in both of my books and my Summa white paper."
I've read the chapters online and the Summa white paper, but it's been awhile; I'll revisit.
"In some major studies that I did to try and find out why bass sounds so good in big rooms and typically so bad in small ones..."
I don't mean to be tiresome, but it seems that the simple explanation is that larger rooms have higher mode density at low freq.
Ah, a light bulb just went on - for some reason I've been thinking that only the modal peaks/nulls would be eliminated, not that freq altogether.
You're right, there will certainly be holes in the response at that freq.
However, these will be maximum only right at the wall and there will be progressively less cancellation moving away from the walls.
I'm thinking in my situation where the seats are relatively central, the slight cancellation will be a lesser evil than the modal peaks and nulls.
Oops, sorry, you said you didn't want to continue with this here, but the the wheels started turning.
Thanks, good food for thought.
"You must not have read much of my work as I drive this point home (that modes are our friends) as often as I can. It's in both of my books and my Summa white paper."
I've read the chapters online and the Summa white paper, but it's been awhile; I'll revisit.
"In some major studies that I did to try and find out why bass sounds so good in big rooms and typically so bad in small ones..."
I don't mean to be tiresome, but it seems that the simple explanation is that larger rooms have higher mode density at low freq.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.