Patrick,
The "jury" has long known that room modes dominate the low frequency response in small rooms.
You are, after twenty years, discovering that fact, and that locating the sub in various locations will change how it interacts with the room modes.
Why you call something a sub that rolls off at more than 12 dB per octave below 100 Hz is another issue 😉..
There are a number of room mode calculators on line, you may be interested in entering your room dimensions in and seeing the results.
Art
Three stack subs with the middle one facing backwards...20dB reduction in rear radiation.
http://www.dbaudio.com/fileadmin/docbase/TI330_1.2E.PDF
Dan.
http://www.dbaudio.com/fileadmin/docbase/TI330_1.2E.PDF
Dan.
Dan,
Unfortunately, in a small room the cardioid array won't exhibit much, if any, reduction in rear radiation.
Outdoors, or in rooms large compared to wavelength, the reduction of rear radiation can be of great benefit.
Art
Art - when I used OB bass - off axis it gave the impression of distortion like a BassMaxx harmonic generator - although OT, I really liked your parabolic long throw speaker - another OT
~1955 - if one had 12 spare K15
~1955 - if one had 12 spare K15
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Cardiod:
Two 12" woofers are mounted in 14"x14"X21" boxes; one is sealed the other has back left off, forming 'U' frame dipole.
When monopole is stacked on top of dipole, very cardioid pattern is result.
With monopole only operating, room modes load to monopole, with dipole only operating room modes load to dipole; operating together room modes load as cardioid.
Is this beneficial in 17' x 25' x 8' room? Modal spacing is sparse <125Hz.
With above woofers EQ was done to make each flat:
Monopole:
'U' frame dipole:
Response with both monopole and dipole running together:
Measurements were done at 5' in front of stack at listening position. Sweet spot below 125Hz is about 5' in diameter for all operating conditions.
Integrated with mains (with monopole woofers) at 80Hz and summed response sent to woofers, sound is identical at listening position regardless of monopole, dipole, or cardioid operation.
Single greatest improvement was equalization of elements to flat.
Dipole operation is waste of driver excursion, in my opinion.
In big room cardioid pattern is quite useful, and implementation without wasteful dipole is way to go.
Two 12" woofers are mounted in 14"x14"X21" boxes; one is sealed the other has back left off, forming 'U' frame dipole.
When monopole is stacked on top of dipole, very cardioid pattern is result.
With monopole only operating, room modes load to monopole, with dipole only operating room modes load to dipole; operating together room modes load as cardioid.
Is this beneficial in 17' x 25' x 8' room? Modal spacing is sparse <125Hz.
With above woofers EQ was done to make each flat:
Monopole:
An externally hosted image should be here but it was not working when we last tested it.
'U' frame dipole:
An externally hosted image should be here but it was not working when we last tested it.
Response with both monopole and dipole running together:
An externally hosted image should be here but it was not working when we last tested it.
Measurements were done at 5' in front of stack at listening position. Sweet spot below 125Hz is about 5' in diameter for all operating conditions.
Integrated with mains (with monopole woofers) at 80Hz and summed response sent to woofers, sound is identical at listening position regardless of monopole, dipole, or cardioid operation.
Single greatest improvement was equalization of elements to flat.
Dipole operation is waste of driver excursion, in my opinion.
In big room cardioid pattern is quite useful, and implementation without wasteful dipole is way to go.
Those stacked delayed woofers and other experiments work if there is no room or a vey very large one. A room always has modes and reflections. Cardioid woofers done the way that Barleywater says, do work in small rooms, there are many practical examples to prove that -or at least they sound different. But I believe that the reason for difference lies above 100Hz.
Room measurements and analyziing them in the range below 300Hz is extremely difficult. You can never trust a single measurement. Every measurement contains reflections and room modes. Geddes' multisub approach is the best way to overcome these, by mixing and distributing these differencies at different locations in the room are minimized and randomized.
A cardioid pattern (and other high-directivity systems) shows it's best assets in the range 150-1000Hz, up to my reasoning. It is because this range of wave length is "foundation" of most sounds and with typical small room speaker placement, most of reflection-nulling and summing occurs there (front-wall, floor, side-walls, perhaps even ceiling)
Room measurements and analyziing them in the range below 300Hz is extremely difficult. You can never trust a single measurement. Every measurement contains reflections and room modes. Geddes' multisub approach is the best way to overcome these, by mixing and distributing these differencies at different locations in the room are minimized and randomized.
A cardioid pattern (and other high-directivity systems) shows it's best assets in the range 150-1000Hz, up to my reasoning. It is because this range of wave length is "foundation" of most sounds and with typical small room speaker placement, most of reflection-nulling and summing occurs there (front-wall, floor, side-walls, perhaps even ceiling)
Last edited:
A practical example of room measurements. Similar downfiring woofers+dipole mids at similar distance to front and side walls. Mic at one spot, distance to speaker equal. 300ms gating, 1/48 smoothing. xo LR4 at 130Hz
Lowest room mode at 40Hz, below that response is even and bass response is boosted to straight line!
Room modes between 40-150Hz are different because back-wall is not symmetric (left side opens to corridor). Moving the mic 30cm or more to any directions will give a very different result.
Fortunately humans have brains that do a wonderful job to sum and process the sound perceived by left and right ear!
Lowest room mode at 40Hz, below that response is even and bass response is boosted to straight line!
Room modes between 40-150Hz are different because back-wall is not symmetric (left side opens to corridor). Moving the mic 30cm or more to any directions will give a very different result.
Fortunately humans have brains that do a wonderful job to sum and process the sound perceived by left and right ear!
Attachments
Patrick - have you looked at these subs?
M3D-Sub : Directional Subwoofer
We use them at work everyday, very good sub with front facing 18" drivers and rear facing 15" drivers. Another way of doing cardioid. It works.
M3D-Sub : Directional Subwoofer
We use them at work everyday, very good sub with front facing 18" drivers and rear facing 15" drivers. Another way of doing cardioid. It works.
Pano,
"It works" in acoustically large spaces, but in small spaces, the boundary reflections will reduce or eliminate the cardioid pattern.
As Meyers mentions in the spec sheet, the response depends on loading conditions and room acoustics, and is not applicable to small rooms.
The rear-facing transducers are separately driven from the front drivers (four channels of amplification per cabinet) by a phase manipulation circuit designed that the rear transducers’ output reinforces forward low-frequency energy while canceling energy emanating from the rear.
For the phase manipulation (time) relationship of the front and rear output to remain intact and cancel rear LF energy, the cabinet needs to be some distance away from boundaries, as reflected waves will not be in the same phase as the cabinet output. The closer to a rear wall, the closer the reflected rear energy is in level compared to the front output, so the lesser (and more ragged) the rear cancellation.
I would be curious to see the response of the M3D-Sub at various locations in a small room (dimensions under 25 feet or 8 meters) placed varying distance from the rear wall compared to a standard dual 18" bass reflex.
Art
How about an enclosure with a front firing woofer and a rear facing woofer, but with a few inches of insulation in front of the rear woofer? I've seen similar treatments done to the rear of front firing open baffle and passive resistive enclosures too add a delay which will result in cardioid response. Is there any reason why the above wouldn't work, or would it be too crude?
On a related note...the rear firing woofer, aside from giving a cardioid type response, does it add to the overall gain/SPL at all?
On a related note...the rear firing woofer, aside from giving a cardioid type response, does it add to the overall gain/SPL at all?
1) Don't know what you mean by "passive resistive enclosures" or how they could "add a delay which will result in cardioid response".
In the range below 100 Hz direct radiator woofers are omnidirectional.
Orientation (facing front or rear) makes no difference in the radiation pattern as long as the woofers are within 1/4 wavelength of each other.
2) Regarding typical "cardioid" subs like the M3D, the rear facing woofers are in a separate enclosure, reverse polarity from the front, and delayed by the time of flight spacing difference.
The output of the rear facing speakers are (mostly) in phase with the front facing speakers in the forward direction, and do add to forward gain, while causing cancellation to the rear when placed away from reflective boundaries.
The monopole/dipole combo sounds interesting. Basically the idea seems to be that the dipole's rear radiation cancels the monopole's rear radiation and adds to the forward radiation. I think it would be interesting to set up an actual experiment to find out how well this actually works
On the Meyer sub, the rear facing 15s are behind some sort of "insulation". I do not know if they add SPL.
Depending on the distance between the woofers relative to wavelength, the rear one may either add or subtract SPL. Same deal as with dipoles 6 dB/oct roll-off.
From what I have seen with simulations it does seem to add a tiny bit at HF 1dB or so. At LF it subtracts 1.5-2dB. At least with my suboptimal sims.
Stack of 3 TH118 with the middle one facing back.
With and without the rear facing delayed cabinet.
Attachments
Above sims are free space; in room approaching moderately big living room, with each 3' from front wall, pattern at 6-12' is virtually identical, except that monopole gets gain from front wall reflection.
- Status
- Not open for further replies.