Another project I’m thinking about, at least to get an idea of the cost involved, to see if it could be done on the cheap side.
I have multiple questions about spl, matrix and driver constrains but let's start slowly with maybe the most important one: the room, is it suitable?
Based on the wiki page and Nils Öllerer papers DBA can work well enough in many configuration even with limited number of drivers.
But the sims are all based on a deeper than wider room, 6x5m, I use mine in the other way and can't change that.
Wider means more subs I get that, but is there also a minimum length needed for the wave to become plane enough?
Plus the room is semi L-shaped, the leg of the L is pretty small but still there and it already messed up my previous setup.
Never had good bass, even with multi-subs, until I tried H-frames in the center.
Basic room dimensions:
Main space is 8x5m, 2.6m high (front wall on top here)
Top right area is 2x1.6m, same height
Bottom left for stairs, only 90cm open to a 2x3x5m high volume behind then opened to second floor.
Maybe already too much details so not showing doors, windows, chimney etc
All walls are granite, ceiling and floor in concrete.
Based on this I have few options:
A - simple 4.4 grid, clearly not optimum but I wonder if it could still deliver a good compromise as I'm targeting a 50hz XO.
Maybe it could already significant, like the dipoles.
Now I wonder if a minimal 2.2 could work, as the 2.6m height would still be lower than the horizontal d
B - better for the intended XO, 4 more drivers (ouch), but still unknown about the top right opening.
I use this space right now but could also maybe fill it a bit with fiberglass, trying an hybrid SBA/DBA but in lateral.
Also the front central bottom sub will have to be placed a bit lower but it should not be a big deal.
B' - Deal with the opening with 2 more subs, that will require 1 more dsp channel (unless by luck just delay and not inverting it could work).
C - in case 65hz is not good enough... insane, but DBA is insane
D - No I'm not even showing how to deal with the stairs, I will have to live with the mess from there 😀
So,
In all cases I could also adjust a bit the rear array, with maybe less driver like 4.2 or 6.4, these solutions seem to work well enough in the papers.
And then depending of all this will come the selection of the driver, another story...
I have multiple questions about spl, matrix and driver constrains but let's start slowly with maybe the most important one: the room, is it suitable?
Based on the wiki page and Nils Öllerer papers DBA can work well enough in many configuration even with limited number of drivers.
But the sims are all based on a deeper than wider room, 6x5m, I use mine in the other way and can't change that.
Wider means more subs I get that, but is there also a minimum length needed for the wave to become plane enough?
Plus the room is semi L-shaped, the leg of the L is pretty small but still there and it already messed up my previous setup.
Never had good bass, even with multi-subs, until I tried H-frames in the center.
Basic room dimensions:
Main space is 8x5m, 2.6m high (front wall on top here)
Top right area is 2x1.6m, same height
Bottom left for stairs, only 90cm open to a 2x3x5m high volume behind then opened to second floor.
Maybe already too much details so not showing doors, windows, chimney etc
All walls are granite, ceiling and floor in concrete.
Based on this I have few options:
A - simple 4.4 grid, clearly not optimum but I wonder if it could still deliver a good compromise as I'm targeting a 50hz XO.
Maybe it could already significant, like the dipoles.
Now I wonder if a minimal 2.2 could work, as the 2.6m height would still be lower than the horizontal d
B - better for the intended XO, 4 more drivers (ouch), but still unknown about the top right opening.
I use this space right now but could also maybe fill it a bit with fiberglass, trying an hybrid SBA/DBA but in lateral.
Also the front central bottom sub will have to be placed a bit lower but it should not be a big deal.
B' - Deal with the opening with 2 more subs, that will require 1 more dsp channel (unless by luck just delay and not inverting it could work).
C - in case 65hz is not good enough... insane, but DBA is insane
D - No I'm not even showing how to deal with the stairs, I will have to live with the mess from there 😀
So,
- A or B, is A a good starting point to experience the DBA effect or should I already target B?
- then B or B'? Does anyone have experience with a similar "small" L shape?
In all cases I could also adjust a bit the rear array, with maybe less driver like 4.2 or 6.4, these solutions seem to work well enough in the papers.
And then depending of all this will come the selection of the driver, another story...
Last edited:
The whole DBA concept relies on a "plane wave", going the wide way you have pretty much doubled the amount of drivers needed for that to be achieved.
That said, doubling the drivers means you can go for about half the driver displacement for the same SPL, so the costs don't have to double...
I'd think 12 drivers would be "close enough", if you aren't going to bother with the left side (stairs?) no need to bother with the smaller right side...
StigErik's DBA measurements are certainly impressive:
https://www.audiosciencereview.com/forum/index.php?threads/double-bass-array-dba.51919/
He used 24 Seas L26ROY drivers, ~$12,000....
That said, doubling the drivers means you can go for about half the driver displacement for the same SPL, so the costs don't have to double...
I'd think 12 drivers would be "close enough", if you aren't going to bother with the left side (stairs?) no need to bother with the smaller right side...
StigErik's DBA measurements are certainly impressive:
https://www.audiosciencereview.com/forum/index.php?threads/double-bass-array-dba.51919/
He used 24 Seas L26ROY drivers, ~$12,000....
Good points about the global Vd! Then maybe I can find the sweet spot for the driver size/cost and numbers.
Because sure I won't go the StigErik’s route!
But if I could get half of his smooth response it would already be great.
I was playing a bit more with amroc, trying to build a room that would match my REW measurements.
It's not possible to have variable height (or I didn't find it) so I played with the stairs opening and the volume of the stair space behind.
This one seemed to work, extending the volume but still at 260cm, and reducing the opening width to match the Sd of the « non door », if it makes sense.
Anyway:
Rew modes, no Eq:
The simulation, the closest I could find with few matches: 23, 30, 37 and 47Hz
With the multiple shapes I tried it appeared the stair are not the issue really as the modes build up behind and don't interfere much with the main room.
But the right side space has more impact.
So probably treating it will help.
Ok I think I will do a plan for 12 drivers, 6.6, it should cover the 20-65hz range and maybe careful placement of the mains could help for 75hz.
With the option to add a big basstrap in the right side.
Because sure I won't go the StigErik’s route!
But if I could get half of his smooth response it would already be great.
I was playing a bit more with amroc, trying to build a room that would match my REW measurements.
It's not possible to have variable height (or I didn't find it) so I played with the stairs opening and the volume of the stair space behind.
This one seemed to work, extending the volume but still at 260cm, and reducing the opening width to match the Sd of the « non door », if it makes sense.
Anyway:
Rew modes, no Eq:
- a few below 20hz
- 23hz
- 30hz
- 37hz
- 47hz
- 55hz and some smaller ones around 70hz
The simulation, the closest I could find with few matches: 23, 30, 37 and 47Hz
With the multiple shapes I tried it appeared the stair are not the issue really as the modes build up behind and don't interfere much with the main room.
But the right side space has more impact.
So probably treating it will help.
Ok I think I will do a plan for 12 drivers, 6.6, it should cover the 20-65hz range and maybe careful placement of the mains could help for 75hz.
With the option to add a big basstrap in the right side.
Ok then come the choice of the drivers...
So let's try a path like that:
So I picked 2x12" as a reference, a baseline, so a total Vd of nearly 2 liters.
I found an answer in the big DBA thread on AVS, there is room gain in dba/sba, confirmed by measurements.
So I would go with sealed, much easier but that drives the driver selection.
The plan is 50 L. net per box, shallow like 60x60x20 cm outside.
Then I compared everything I could find in winisd, mostly from the TLHP library as it's very easy with their filters, speakerlite box was useful too.
Even if I have limited trust in all these I believe it should be close enough to real world results.
So tried a lot from SBacoustics, some ciare, sica, visaton, daytons + some PAs etc
But to match the total Vd with 6 front drivers means a Vd per driver near 0.3 Liter, not that easy for non expensive ones.
That's either an 8" with 15mm Xmax or a 10" with 10mm Xmax, with low FS, low Vas, reputable and reasonable price, I couldn't count them on my hand...
So price was the driving factor and I took the other way around, starting based on price.
I ended up with 10", either the SB26STWGC or the SLS-P830668.
Here are some sims maximising excursion with eq, 8x sls / 6x sb26 / 2x um12:
8x sls + boxes would mean around 800eur, 6x sb26 + boxes around 700eur (double it for both walls).
I think will try with the SLS as it would give a better spread.
So let's try a path like that:
- define SPL goal
- evaluate room gain
- select sealed or ported
- select driver per box size
So I picked 2x12" as a reference, a baseline, so a total Vd of nearly 2 liters.
I found an answer in the big DBA thread on AVS, there is room gain in dba/sba, confirmed by measurements.
So I would go with sealed, much easier but that drives the driver selection.
The plan is 50 L. net per box, shallow like 60x60x20 cm outside.
Then I compared everything I could find in winisd, mostly from the TLHP library as it's very easy with their filters, speakerlite box was useful too.
Even if I have limited trust in all these I believe it should be close enough to real world results.
So tried a lot from SBacoustics, some ciare, sica, visaton, daytons + some PAs etc
But to match the total Vd with 6 front drivers means a Vd per driver near 0.3 Liter, not that easy for non expensive ones.
That's either an 8" with 15mm Xmax or a 10" with 10mm Xmax, with low FS, low Vas, reputable and reasonable price, I couldn't count them on my hand...
So price was the driving factor and I took the other way around, starting based on price.
I ended up with 10", either the SB26STWGC or the SLS-P830668.
Here are some sims maximising excursion with eq, 8x sls / 6x sb26 / 2x um12:
8x sls + boxes would mean around 800eur, 6x sb26 + boxes around 700eur (double it for both walls).
I think will try with the SLS as it would give a better spread.