or Connect
AVS › AVS Forum › Audio › DIY Speakers and Subs › Double Bass Array (DBA) - The modern bass concept!
New Posts  All Forums:Forum Nav:

Double Bass Array (DBA) - The modern bass concept! - Page 3

post #61 of 321
what happens when we have more than 4 drivers?
the spacing between them is descreased thus the spacing VS walls and floor/ceiling also decreases?

With 16 time 18", i'd probably be looking at the complete front wall though
ahhaha nice image in my head right now


is there any problem with those drivers used in Stereo ?

is there any theory on that. i just want to understand why this specific spacing gives a wavefront ( maybea drawing with wavefront ?? )

what are the CONS of this technique ?

the amplitude of the front to rear modes will be more than usual since it will take all of the power from the other dimensions wich are not present??
post #62 of 321
Thread Starter 
Quote:
Originally Posted by JinMTVT View Post

what happens when we have more than 4 drivers?
the spacing between them is descreased thus the spacing VS walls and floor/ceiling also decreases?

Yes, as you can see in the picture below, the distance between the wall has to be 1/2 of the distance between the drivers itself to form a planar wave. The source (S0) is mirrored on the side wall (S1) which leads in addition to the driver grid to a waveform of an infinite plane source.

So the formular for the positioning on the front and back wall is

pX = (2 * n + 1) * widthWall / (2 * numDriversX)
pY = (2 * n + 1) * heightWall / (2 * numDriversY)

n = 0, 1, 2, 3,...

LL
post #63 of 321
Just saw this thread revived. To help anyone having trouble visualizing the layout, and what is going on here, I'll try expressing it differently. It's easier if you think only in the horizontal plane, as if you are looking down from above. Everything is the same in the vertical direction.

I've attached a drawing to help show the mirror effect.

As to why the spacing is critical... moving a speaker to the right, moves its reflection to the left, and would lead to inconsistent spacing between the drivers and their reflections:
Instead of this: _x__x_|_x__x_|_x__x_
You'd get this: x____x|x____x|x____x
LL
post #64 of 321
I get the front wall. As for the back wall, the woofers are time delayed so their cones are effectively moving back the same distance as the front woofers did, thus they are essentially catching the same air thats thown at them. A poor analogy is what the winning team of an egg thowing contest would do. One person gently throws the egg with two hands, and ever so gently, with the same motion and in the same direction, the other person catches the egg. Those who have the same catching and releasing motion can essentially go on forever.
post #65 of 321
not sure i get it there
if i use your formula for 4 drivers on X with wall width of 10 units
i get positions at : 2.5-5-7.5-10
???
am i dumb again! ?
post #66 of 321
Thread Starter 
Quote:
Originally Posted by JinMTVT View Post

not sure i get it there
if i use your formula for 4 drivers on X with wall width of 10 units
i get positions at : 2.5-5-7.5-10
???
am i dumb again! ?

I don't know if you are dumb, but you definitely didn't used the formular!

OK, here we go (widthWall = 10, numDriversX = 4):

n = 0:
pX1 = (2 * 0 + 1) * 10 / (2 * 4) = 1.25

n = 1:
pX2 = (2 * 1 + 1) * 10 / (2 * 4) = 3.75

n = 2:
pX3 = (2 * 2 + 1) * 10 / (2 * 4) = 6.25

n = 3:
pX4 = (2 * 3 + 1) * 10 / (2 * 4) = 8.75
post #67 of 321
lol ....
i was doing 2 * (n+1) instead of 2*N+1
my bad ...thanks very much for the nice formulas



Now onto the real beast,

what will happen if the side walls aren't paralell ??
would it work ??
post #68 of 321
ok
LAST QUESTION PLEAS PEOPLE!

i need answeres to that so i can be done with my new house design
( well at least the HT room )

Is there any problem with plaanr waves if the side walls aren't paralell ?
is it a requirement ?
post #69 of 321
For optimum performance it is necessary. Then again, for optimum performance, there should be no bass traps and no furniture in the room. Deviations from perfection result in less than ideal performance. If you room is relatively rectangular, I would say you're probably fine. If it is far from rectangular, the results may be far from perfect. Ultimately nobody knows how well your room will do until you try and take measurements.
post #70 of 321
Only guessing here: assuming they are remotely close to parallel, I wouldn't think it would have a big effect if you're doing the passive absorbtion at the rear. I would expect a much bigger difference if you were doing the active time-delayed phase-reversed subs at the rear of the room (since the rear wall would then be a different size, requiring different placement distances, etc...)
post #71 of 321
thinking of it...it is probably a similar wave front as experienced within most horns,
wich doesn't have paralell walls ??
so i guess that it would work to some extent!

i'll have to see about the rear of the room, i'll probably try to absorb most of it
to get rid of the longitudinal mode as much as possible
it will also help achieve my nearfield listening space

i guess that with such a planar wave, nearfield like loss of sound per distance applies?
post #72 of 321
No. At least in theory, there would not be any loss of sound per distance. (at least in a perfectly executed DBA with straight sides)

In point source examples, the energy being produced, i.e. the wavefront, occupies a larger and larger sphere as it radiates out. Fixed amount of input energy, stretched over a larger and larger surface - thus the decrease in energy in any one small part of that spherical wave.

Line arrays radiate out in a cylindrical wave, therefore less surface area per distance and better retention, if you will, of that original energy.

The DBA produces a plane wave - which continues to occupy the exact same amount of surface area as it travels through the room - therefore no decrease in local intensity/energy level.
post #73 of 321
"The DBA produces a plane wave - which continues to occupy the exact same amount of surface area as it travels through the room - therefore no decrease in local intensity/energy level."

Since the entire front wall is not occupied by drivers, this is certainly not so near it.

It would be interesting to see at what distance from the front wall the drivers' output is more or less fully integrated.
post #74 of 321
Noah... that is an interesting thought. It is to 3-D as the line array is to 2-D. There is a certain area one must be in where you are effectively in the near field response but comb filtering is minimized. I suppose it's a function of the spacing between the woofers (the greater the spacing, the lower the low pass filter and the further away you must be).
post #75 of 321
Do you guys have any sources for all that info ?
i'd like to read more papers ( not in german please ) about planar waves
and DBA setup ...



What would it do if the SUB array would be used for STEREO bass ???
would it work the same ?
i guess not ..
post #76 of 321
Thread Starter 
Quote:
Originally Posted by noah katz View Post

Since the entire front wall is not occupied by drivers, this is certainly not so near it.

It would be interesting to see at what distance from the front wall the drivers' output is more or less fully integrated.

I measured my DBA and placed the microphone in the middle of the room's width and changed the distance to the front. The graph shows the distances 1m, 2m, 3m and 4m at 1/3 octave filtering. The room's length is 5.6m.
As you can see, the frequency response and the overall level stay nearly constant. Only in the upper bass where the driver grid is not dense enough differences occur. But these are also pretty small in comparison to single woofers.




@JimMTVT
Quote:


What would it do if the SUB array would be used for STEREO bass ???

A DBA is monaural, so you have to sum the bass of both stereo channels. But since most mixes contain only mono bass information and our ear is not good at locating bass this is no big problem.
post #77 of 321
Thread Starter 
Btw, my front looks like this now. Only the 3m wide screen is visible, nothing else.

post #78 of 321
"As you can see, the frequency response and the overall level stay nearly constant."

Yes; thanks for the measurements.

I realize that my statement is in error; I was overfocusing on the statement mentioning the front wall, completely forgetting that there are also drivers on the back wall.
post #79 of 321
Hi Follgott,

Quote:
Originally Posted by FoLLgoTT View Post

I measured my DBA and placed the microphone in the middle of the room's width and changed the distance to the front. The graph shows the distances 1m, 2m, 3m and 4m at 1/3 octave filtering. The room's length is 5.6m.
As you can see, the frequency response and the overall level stay nearly constant. Only in the upper bass where the driver grid is not dense enough differences occur. But these are also pretty small in comparison to single woofers.

Thanks for your measurements. I think it's indeed very interesting to see measurements at diffferent locations in the room.
Could you post the same graph, but without smoothing and maybe some waterfall plots at different locations?
This would allow for an evaluation of how far this concept can really eliminate the modes.

Thanks and besten Gruß
-Y
post #80 of 321
Thread Starter 
@Ydope
Sorry for the late answer. I'm very busy with my diploma thesis at the moment. When I get some time I will post more measurement plots.
post #81 of 321
Fascinating thread - thanks to Follgott for initiating it.

So, my room is tiny. As in small. I can stick a single driver in the center of the front and back walls and still be less than about 5' from any boundary. Given the c/2x equation, this seems to indicate that such a setup would satisfy the conditions up to ~100hz or a bit better.

It can't really be that easy, can it?

I have a couple of the original NHT1259's that I guess I'm just going to have to throw into an experiment. Like my project list wasn't already long enough.
post #82 of 321
As long as they are half way up the wall, I don't see why not. That would be a pretty easy experiment to try. Please reply with your impressions.
post #83 of 321
Quote:
Originally Posted by FoLLgoTT View Post

@Ydope
Sorry for the late answer. I'm very busy with my diploma thesis at the moment. When I get some time I will post more measurement plots.

Ah, thanks for the info. Good luck with the thesis!

Maybe you can tweak the topic a bit and include the DBA measurements.
;-)


Greets Y
post #84 of 321
I don't think my wife would like me hanging 4 18's on the wall...not her type of art
post #85 of 321
Hate to bump up a dead topic (not that it should be), but is it possible to do this with subs inward of the walls? Say a few subs 1 meter from the front wall and a few 1 meter from the rear wall?
post #86 of 321
Quote:
Originally Posted by MrHelpful View Post

Hate to bump up a dead topic (not that it should be), but is it possible to do this with subs inward of the walls? Say a few subs 1 meter from the front wall and a few 1 meter from the rear wall?

Maybe. I've run a good number of simulations. Performance (spacial uniformity of sound field) was relatively constant up to about 1.5 feet distance from the front and back walls. It then started to fall off, reaching a rate of about 1 dB standard deviation per foot at 2.5 feet.

This distance was from the acoustic centers of the subs. The acoustic center is a relatively new concept. It is the point at which spherical sound waves eminate from the sub. For an unported sub, this is a few inches in front of the driver. For a rear ported sub, it is somewhere within the sub box.

However, you can always face the subs toward the walls! Then the acoustic center of an unported sub can be put arbitrarily close to the wall. "But with the sub facing away from the room, the sound would get blocked by the sub," you might think. No. Subwoofer frequencies are truly omnidirectional. It makes no difference which direction the sub is facing. It has no directionality.

Regards,
Terry
post #87 of 321
Quote:
Originally Posted by Terry Montlick View Post

Maybe. I've run a good number of simulations. Performance (spacial uniformity of sound field) was relatively constant up to about 1.5 feet distance from the front and back walls. It then started to fall off, reaching a rate of about 1 dB standard deviation per foot at 2.5 feet.

This distance was from the acoustic centers of the subs. The acoustic center is a relatively new concept. It is the point at which spherical sound waves eminate from the sub. For an unported sub, this is a few inches in front of the driver. For a rear ported sub, it is somewhere within the sub box.

Regards,
Terry

Hi Terry,

What does "1dB std deviation/ft" translate to in average and maximum deviations? At first glance that sounds quite good in terms of real rooms.

As a minor correction/addition, a ported or other higher order subwoofer doesn't really have a different origin, but where the origin is constant for a sealed subwoofer, the origin shifts at low frequencies in a ported design. It is generally better to conceptualize it as a multi-way device (driver + port or PR) than as a lumped factor, as this better explains real world observations related to subwoofer orientation and placement.
post #88 of 321
Hi Mark,
Quote:
Originally Posted by Mark Seaton View Post

Hi Terry,

What does "1dB std deviation/ft" translate to in average and maximum deviations? At first glance that sounds quite good in terms of real rooms.

The 1 dB std deviation/ft is a guestimate based on the graph of composite spatial standard deviations (multiple experiments) vs. sub distance from wall, over the range of 0 to 3 feet distance. These spatial variations were determined by sampling at evenly spaced microphone positions in a plane at a fixed distanced from front and back walls. At 0 feet subwoofer wall distance, the spacial standard deviation was 1.3 dB. At 3 feet distance, it was 2.3 dB.

Yes, this is quite good in terms of real rooms. The microphone positions sampled included the corners of the plane! The performance from 0 feet to 3 feet changed from phenomenal to merely great.

Quote:
As a minor correction/addition, a ported or other higher order subwoofer doesn't really have a different origin, but where the origin is constant for a sealed subwoofer, the origin shifts at low frequencies in a ported design. It is generally better to conceptualize it as a multi-way device (driver + port or PR) than as a lumped factor, as this better explains real world observations related to subwoofer orientation and placement.

See John Vanderkooy's April 2006 AES UK presentation. That's where I got the info. Vanderkooy has a number of papers studying the acoustic center.

Regards,
Terry
post #89 of 321
Thread Starter 
Introduction
On the christmas days I had much time to take measurements of different configurations of my eight subwoofers.

The following configurations were measured:


DBA:
  • front array active
  • back array with inverted and time delayed signal

SBA:
  • only front array active

Multi sub:
  • both arrays are feeded with the same signal



Measurements with ARTA

Frequency response (not calibrated, without window, smoothing 1/24 octave)

SBA:


Multi sub:


DBA:




Burst Decay


SBA:


Multi sub:


DBA:




RT60


SBA:


Multi sub:


DBA:





Measurements with RoomEQWizard

RT60 (blue: SBA, red: DBA):


SBA:


DBA:




Btw, maybe someone is interested in translating the DBA article at Wikipedia from german to english.
post #90 of 321
Thanks for the updates, Nils! Google seems to be getting better at translating German to English. That wikipedia article is actually understandable.

http://translate.google.com/translat...hl=en&ie=UTF-8
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: DIY Speakers and Subs
AVS › AVS Forum › Audio › DIY Speakers and Subs › Double Bass Array (DBA) - The modern bass concept!