So you've never heard "room gain" being applied to anything other that pressurisation gain ? So you would consider a room that has no pressurisation gain (or is driven by a dipole for example) to have no room gain ?
Perhaps just a case of semantics. To me room gain encompasses all the effects that cause an increase in bass response compared to the free field response of the speaker at the same measurement distance. The cause doesn't really matter, the result does.
There is more than just pressurisation and standing wave modes that increase the bass response in a room over free field though - the room boundaries constrain the solid angle of radiation thus increasing the SPL at the listening point. All three effects combine together.
Imagine a shoebox shaped room with the speakers at one end and the far end wall near the listener is missing but all other walls are present. Is there any pressurisation gain in this room ? Nope, because an entire wall is missing.
Are there any modes ? Yes, although the length mode wont be present because there are no parallel boundaries in that axis for the wave to bounce back and forth repeatedly from. Width and height modes will still be present though.
Do the remaining modes explain all the increase in bass response over free field conditions ? No, because from the perspective of the listener within the room boundary the wave from the speaker can't expand into full space until it exits the open wall at the back of the room - within the "room" the SPL will be increased due to the constraining effect of the walls, and this will apply even below the lowest (remaining) room mode.
Below the lowest mode all the room reflections will arrive constructively and give maximum gain due to their long wavelength, but once you get into the modal region some of the reflections start to arrive out of phase and start becoming destructive.
The net result is that you get maximum coherent "gain" from the wall reflections below the modal region (independent of pressurisation) and on average less gain within the modal region as the reflections become incoherent - thus you inherently get a tilted down bass response which falls with increasing frequency in a room, which was my point.
(Another analogy is the transition between coherent addition and power addition of two widely spaced speakers - at low frequencies they will add to 6dB as they are in phase but at high frequencies they will tend to add to 3dB at anything other than perfectly matched distances due to dense comb filtering)
It's easy enough to prove with a reflection/image based room response simulator such as the FRD consortium spreadsheet - you can set the reflectivity of an entire wall to 0 and whilst the bass response certainly changes a lot (and one of the modes is eliminated) the general downwards trend in bass with increasing frequency remains, albeit with less slope. In fact even with two walls eliminated it does, depending on their orientation to each other.
One also only has to measure the response in typical rooms to empirically show that rooms almost universally cause a tilted down bass response relative to the actual nearfield response of the speaker.
Last edited:
That is correct
These effects are all already encompased in the calculation of the room modes and should not be considered as a seperate effects. There have been lots of mistakes made in this regard in the literature. Any and all effects of the room boundaries are already accounted for in the normal modes solution. Many claim that the sources "free field" response needs to be added to the room mode response, but this is incorrect. This "free-field" or direct field response is already present in the modal series solution, but convergence is slow. This effect is well described by Morse in Vibration and Sound in his discussion of Eq 20.4. He shows how the direct field term is contained in the normal modes solution and how it can actually be seperated off and defined independently, which, of course, requires a new normal modes series solution. This new solution now converges much faster than the normal one. A very neat trick.
It pretty exciting to be in this hobby for 30 or 35 years and come across such great info like multiple sub placement threads.
I'm in the middle of a project that will take my subs to the next level. I have four colocated 15's in the front center of my room that are way to heavy and way too large to relocate them. My only way to fill in the nulls at my LP is locating what i call "tuning" subs mid wall on the other three walls.
I built one already and tested it. It works perfect. I dragged in two other subs from other systems in my home to verify their effect on the listening position, and dont you know it, i now have a workable solution to the huge depression at the LP centered around 55hz.
My plan is to build two more tuning subs for a total of three that will surround the LP, sides and back, with the front colocated subs to fill in the front wall.
The whole thing will be EQ'd with a single MiniDSP.
Here is one tuning sub, it an RS Dayton 12" and a Dayton 240 plate amp
Here are my colocated 15's
I'm in the middle of a project that will take my subs to the next level. I have four colocated 15's in the front center of my room that are way to heavy and way too large to relocate them. My only way to fill in the nulls at my LP is locating what i call "tuning" subs mid wall on the other three walls.
I built one already and tested it. It works perfect. I dragged in two other subs from other systems in my home to verify their effect on the listening position, and dont you know it, i now have a workable solution to the huge depression at the LP centered around 55hz.
My plan is to build two more tuning subs for a total of three that will surround the LP, sides and back, with the front colocated subs to fill in the front wall.
The whole thing will be EQ'd with a single MiniDSP.
Here is one tuning sub, it an RS Dayton 12" and a Dayton 240 plate amp
Here are my colocated 15's
Last edited:
Dun dun dun... Thread revival!
I'm planning a multi-sub arrangement with dipole subs but I'm not sure of positioning: Should I still arrange them to excite the most amount of nodes as with monopoles or should I try to excite the least since they are dipoles?
If I've understood correctly the excite most if they are 45 degrees to the walls and least if I place them parallel?
Thanks in advance,
Olle
I'm planning a multi-sub arrangement with dipole subs but I'm not sure of positioning: Should I still arrange them to excite the most amount of nodes as with monopoles or should I try to excite the least since they are dipoles?
If I've understood correctly the excite most if they are 45 degrees to the walls and least if I place them parallel?
Thanks in advance,
Olle
Place them directly behind the listening position:
Comparison of different near field and far field subwoofer configurations
There are issues with bass reproduction in domestic settings which are all but unsolvable. Therefore, you need to address the problems in a statistical/decision approach.
It must strike terror into the hearts of the wannabee engineer kind of person, but the right approach is to be as heterogeneous as possible in your woofing, in terms of the kinds of housing, the locations, etc.
And read Nate Silver's best seller on statistics of prediction.
Ben
It must strike terror into the hearts of the wannabee engineer kind of person, but the right approach is to be as heterogeneous as possible in your woofing, in terms of the kinds of housing, the locations, etc.
And read Nate Silver's best seller on statistics of prediction.
Ben
Last edited:
Ok, what makes me want to test dipoles is to see if there is less bass leaking to surrounding rooms. I want to have the setup apartment friendly so reduction of sound for the neighbours is more important than full volume down at 20 hz. From 30 hz and upwards there there will still be enough headroom to get complaints so the only sacrifice would be the really low bass but this is a music setup so should be fine.
That's not accurate. The impact of the angle itself can't be stated without complete context of its location in the room.
To maximally excite a given mode, simply locate the dipole at the place of a modal "hump" in the room, with the dipole axis matching the axis of the mode. For a perfectly undamped rectangular room, this would require 3 dipoles, located at the midpoint of each dimension l, w, h. The one for the vertical dimension would have to point toward floor and ceiling.
. . . Or just park a single monopole in the corner.
The O'toole/devantier paper suggests mid-wall ocations, but for completely opposite reasons!-- placement of monopole/omni subs at these locations actually suppresses the resonance of a given axial mode. A dipole has exactly the opposite effect at THAT location.
I believe that dipoles are better used (in Geddes approach) as fill subs.
All-dipoles can be used, but some of the fundamental benefit of his method will be lost; the Geddes method is the most efficient because that first corner sub provides a high-output foundation, with subsequent subs primarily functioning to smooth out the response. Dipoles can be very effective for this as long as the user is aware that he/she has added a variable to the mix (polar pattern with phase flip). The way to address this is to optimize dipole orientation at a given location BEFORE making electrical phase and volume adjustments.
--Mark
Last edited:
Trust me, dipoles won't work any better in this regard than anything else. For a given SPL in the room, the leakage will be the same regardless of the source type. Dipoles just have less output than a monopole so there is less SPL and hence less leakage (apparantly, but not actually).
It is true that dipoles want to be closer to the center of the room and away from the walls (there is a lower modal velocity near the walls), while monopoles tend to want to be nearer the walls where there is always a pressure maximum.
Hmm.
One of the things I want to test is to remove room gain. How important is it for all the subs to work in phase or close to at 20-40 hz? An idea would be to test to instead of having 3 dipoles use 2 monopoles and a dipole as HF sub where each of the monopoles works in inverted phase. That should solve that and they should be able to be spaced far apart enough such that fairly little efficiency is lost due to cancellation.
One of the things I want to test is to remove room gain. How important is it for all the subs to work in phase or close to at 20-40 hz? An idea would be to test to instead of having 3 dipoles use 2 monopoles and a dipole as HF sub where each of the monopoles works in inverted phase. That should solve that and they should be able to be spaced far apart enough such that fairly little efficiency is lost due to cancellation.
Last edited:
So you're saying there's no difference between 4pi, 2pi, etc? Or were you referring to something else?
Rooms are neither 2 Pi or 4 Pi, they are modal, and that is a completely different thing. The 2 Pi and 4 Pi arguments do not apply to real rooms.
There is a DC mode in a perfectly sealed room, that is sometimes called "room gain", but real rooms are too leaky to see this effect.
Was just looking at the Gedlee sub and noticed a couple things:
1) He's really not messing around with that sub driver! If I'm not mistaken the 12TBX100 has the same ginormous motor as it's big brother the 15TBX100. Parts Express doesn't stock it, likely because most people opt for the lower priced options, which don't have this wildly over-engineered motor
2) I have no idea how he's selling the sub so cheaply, considering the driver is $300 at usspeaker. Considering the parts cost, I'd expect the sub to cost $2000-$3000
1) He's really not messing around with that sub driver! If I'm not mistaken the 12TBX100 has the same ginormous motor as it's big brother the 15TBX100. Parts Express doesn't stock it, likely because most people opt for the lower priced options, which don't have this wildly over-engineered motor
2) I have no idea how he's selling the sub so cheaply, considering the driver is $300 at usspeaker. Considering the parts cost, I'd expect the sub to cost $2000-$3000
One correction: The motor is not "over-engineered" it is optimal. the voice coil is enormous and the magnet structure also needs to be. All this leads to huge thermal capacity and a sound quality unlike any other. The 12TBX100 and the 15TBX100 have identical motor structures, just different cones. To me, they both represent the best woofers available - bar none.
That the "brand names" charge ridiculous prices is quite true. People pay it so they do it!
Its about thermal modulation, not SPL.
That is true as well. My speakers are loafing at normal levels and as such the thermal modulation is negligible. Small two ways are maxed out at even reasonably high SPls and even if they are not totally distorted, they are thermally modulating and losing dynamics as a result.
This whole concept is simply not understood.
Last edited:
