AVS Forum banner

Multi-sub: adding DIY-SG flatpack 18"+15" sub to 4x15" IB line array

17K views 89 replies 17 participants last post by  Little Chris 
#1 ·
I'm joining the multi-sub approach, my IB sub does a great job but I can't get decent seat-seat flat respone in 1st row, the 2nd row is pretty decent seat-seat..

[edit] adding some good multi-sub related threads in post#1
Quote:
Originally Posted by mtbdudex  /t/1446659/multi-sub-help-diy-trian...s-x-34-face-x-32-tall-5-5-cu-ft#post_22740255

multi-sub article by "Markus Mehlau".

Some other multi-sub threads I had subscribed to, I know there are a few more:
http://www.avsforum.com/t/1388156/a-new-approach-to-multiple-subwoofers
http://www.avsforum.com/t/1386208/more-than-4-subs-is-completly-pointless
http://www.avsforum.com/t/1380013/my-multi-location-subwoofer-thread-inspired-by-welti-devantier-geddes

Option 1:

Build a "helper" sub using Dayton Audio TIT400C-4 15" Titanic Mk III Subwoofer in 3.5 cu ft sealed box

(this has tight fit into the corner due to my screen and LH Main location......)

[edit]

Option 1 has evolved into 2nd and 3rd sub approach, using an 18" driver for 2nd sub and 12" driver for 3rd sub

Here is graphic of that:



Option 2: (this thread discussion point)

Build a specific room designed Triangle sub w/ Dayton Audio RSS460HO-4 18" , amp, the sub triangle box will have 24" sides x 34" face x 32" tall; give or take 5.5 cu ft internal space after braces/etc.


Question:

If a triangle sub is possible and does not "break" the laws of boxed subs, then I'm doing option 2 and this thread can become a build thread, if not then I'll continue with option 1 and I'll get flat pack from Eric H (I posted in that thread a few days ago)


Basic triangle sub image is this:



Located just as shown here ................. This will give symmetric balance to the IB 4 x 15" line array on the RH side:
.


You can see my corner bass trap triangle support, this triangle corner sub would be located just under that.


My main goal is the first row has modal issues that can't be solved via all the corner bass traps I've installed, so I'd like to smooth them more by installing a just as (almost?) capable sub on the LH side as my IB sub on RH side.

(I can't do a stereo IB sub on LH side or I would have)

It will be sticking out into the room, while the IB baffle plate on the RH side is about 4" into the room, not purely symmetric.....

1st row seats 1-4, you can see/hear the seat-seat variations..............................2nd row seats 5-8, very consistent across the seats::
.



I'm looking forward to planning and "fixing" my 1st row seats modal issues.
 
See less See more
6
#2 ·
I was thinking that by symmetry, looking at seat 1 vs 4 and seat 2 vs 3, and combining their individual charts, the summed charts showed a flatter graph.....which dove me to consider 2nd sub in LH corner.



Here is seats 1 & 4 plots from 1+ year ago, those seats are "nearly" symmetrical .. Here are seats 2 & 3 plots, those seats are "nearly" symmetrical
.



The 2nd sub then should "fill in" the nulls from the 1st sub

So by visual overlay looks like it should work.

I'm sure not 100% symmetrical but quite close.
 
#3 ·
I vote for option 2 (triangle idea) I would test that placement out to see if it will in fact impact your 2nd row seating. If it were me I'd buy a decent commercial offering from a big box store, test the location, and return it before I built a sub that may not fix my modal issues. I would also try the store bought option nearfield to see what impact that has on your results. It's not impossible a small sub located near the 2nd row of seating could fix some variances (think Geddes sub theory).
 
#4 ·
I got up early this morning with intention of deep dive into REW and 2nd helper sub locations:

-moved my existing 12" Atlantic Technology 350 PBM sub from family room to HT


-moved the 24" iMac from craft room to the HT

-fired up REW v5 on my iMac, which I've used many-many times


Yep - this is why you don't upgrade OS on a older computer just for the sake of having the latest....


On some Mac's there was an issue with JAVA and Output, using Mandolane M3DMixer was a known work around that JohnM was aware of, but now there is a conflict with OSX 10.7.x it seems....just when I want to get down and dirty with REW!


There is a thread at the Shack I'm in trying to get it to work with OSX 10.7.5.
 
#5 ·
Can you have enough room to sneak the new HO 18 in the corner? Or maybe an SI 18? Just tacking on some more displacement to help it keep up with the IB.
 
#6 ·

Quote:
Originally Posted by WiSounds  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22734203


Can you have enough room to sneak the new HO 18 in the corner? Or maybe an SI 18? Just tacking on some more displacement to help it keep up with the IB.

I can fit an 18" driver as shown in a triangle approach subwoofer, LH side as shown in 1st post.

The LH main speaker has to go outboard of my 130" screen, and that limits the width/depth of a box sub to just 18" wide max.

I'm all for keeping up with the IB, the IB actually has the ULF below 20hz covered pretty well in all seat positions as the graphs show.
 
#7 ·
There is no guarantee you will get good results with symmetrical position for the second sub. You are likely to have better chances with the second sub away from the front wall, perhaps along the left wall closer to the lseating area.

It does not need to be a powerful or very extended subwoofer. Multisub is useless below 30Hz or so anyway. But it would greatly help if the sub has variable LPF, phase/delay and volume controls .


Good read on the subject
http://mehlau.net/audio/multisub_geddes/
 
#8 ·
Can you borrow a sub to move around the room and measure a few different positions before settling on a final location?
 
#10 ·
I need to read better. That's what I get for multitasking.
 
#11 ·
I appreciate the help and time from everyone, we are all busy.

I've got REW working again, new cal file/etc.

I've also had to learn my new Denon 4520CI and mapping HDMI to front panel while audio is Ext in so no EQ applied.

Family time now, so looks like another 5am measurement session Sunday morning

As sanity, seat 2 and seat 6 are pretty close to long ago measurements.....
 
#12 ·
Lets say getting up 4:30am to measure subs on a Sunday morning....


Started out with facing the 12" box sub along same triangle plane


Measured in all 8 seat positions box sub alone, IB sub alone, then both combined. Put good notes for each measurement.


Then, moved the box sub along LH wall 5ft, 6ft, 7ft from front wall and took measurements seat 6 and seat 1.


Later will post results .
 
#13 ·

Quote:
Originally Posted by mtbdudex  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22736874


Lets say getting up 4:30am to measure subs on a Sunday morning...Later will post results .
Woke up at 04:15 and am currently at work. Let's see those results!


OT. Those wood trimmed sound absorbers, did you use OC703 or cotton? I ask because I am days away from purchasing the cotton batts at Home Depot along with the wood trim and wanted to know what you thought of yours.
 
#14 ·

Quote:
Originally Posted by Looneybomber  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22736894

Quote:
Originally Posted by mtbdudex  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22736874


Lets say getting up 4:30am to measure subs on a Sunday morning...Later will post results .
Woke up at 04:15 and am currently at work. Let's see those results!


OT. Those wood trimmed sound absorbers, did you use OC703 or cotton? I ask because I am days away from purchasing the cotton batts at Home Depot along with the wood trim and wanted to know what you thought of yours.

Those are oc703, rapped in speaker grille cloth, 2" thick with 2" air gap positioned at first reflection points , forget using cotton will not absorb freq correctly, more here http://www.avsforum.com/t/1312693/diy-construction-methods-of-hang-able-acoustic-panels-not-fixed-frames#post_19947600



What do I think of acoustic treatments?

Best low cost upgrade by far, immediately noticeable. too many people chase speaker upgrades w/o addressing the room/speaker interactions.
 
#15 ·

Quote:
Originally Posted by Looneybomber  /t/1446659/diy-triangle-sub-w-dayto...4-face-x-32-tall-5-5-cu-ft/0_80#post_22736894


Woke up at 04:15 and am currently at work. Let's see those results!

OT. Those wood trimmed sound absorbers, did you use OC703 or cotton? I ask because I am days away from purchasing the cotton batts at Home Depot along with the wood trim and wanted to know what you thought of yours.

I used a combination of both.


1" thick linacoustic 48" up sides, and cotton batting from 48" to ceiling. Only exception was back 1/3 of theater behind surrounds, and back wall, I didn't put cotton on upper walls to keep from reducing surround effect.


Don't forget the front wall either. That made a huge difference for me when I covered entire front wall with linacoustic and cut out triangles from OC703 for corner bass traps.
 
#16 ·

Quote:
Originally Posted by bass addict  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22737630

Quote:
Originally Posted by Looneybomber  /t/1446659/diy-triangle-sub-w-dayto...4-face-x-32-tall-5-5-cu-ft/0_80#post_22736894


Woke up at 04:15 and am currently at work. Let's see those results!

OT. Those wood trimmed sound absorbers, did you use OC703 or cotton? I ask because I am days away from purchasing the cotton batts at Home Depot along with the wood trim and wanted to know what you thought of yours.

I used a combination of both.


1" thick linacoustic 48" up sides, and cotton batting from 48" to ceiling. Only exception was back 1/3 of theater behind surrounds, and back wall, I didn't put cotton on upper walls to keep from reducing surround effect.


Don't forget the front wall either. That made a huge difference for me when I covered entire front wall with linacoustic and cut out triangles from OC703 for corner bass traps.
Ok - let's not get too OT in this thread.....

1" is not thick enough to absorb the lower bass above 300hz, that is applying EQ by not truly absorbing down to 300hz, and wrapping it all around the whole HT can be debated in the Master Acoustics thread.

There is no need to absorb that much sonic energy, just at specific first reflection points, live sounding rooms are a good thing!


Use the mirror method to help guide you placement of your panels.

The ETC will guide you the real need if in doubt, and confirm if they properly located and thick enough.
 
#17 ·
Back to this thread....


Some charts to share:


I always start with 4" nearfield measurements, in this case the IB sub show how flat it is from the get go to ULF, the 12" box sub.....well it shows what looks like a HPF at 27hz?

I had both of these gain calibrated @ 75db seat #2 - about 11 ft away, the box sub it was put away from the corner to lessen the effect of corner on it.

I was kinda surprised how much "hotter" the box sub was at nearfield than the IB sub, which of course means the IB sub just carries itself so much better thru the room.



Here is overlay of the IB sub seats 1-4 with overlay of the IB sub seats 5-8, both these are pretty close to what I measured 1 year ago so I feel my hardware set-up and cal file is correct
..




Now, after I took all these measurements, and extracted charts, I realized that I did the 75db calibration yesterday, box sub via its front gain knob, while the IB sub the MV gain, and then it was home movie time for the family so I shutdown the set-up.

However, upon waking up this morning and firing up everything, I think I forgot to re-set the AVR to -27.5db, it goes to -40db at shut-down...duh on me


(the box sub of course gain is via the front knob)

I changed it during initial set-up back....or partly in....it was 5am.....

So I'll have to re-do a few to see if all my work this morning is good or needs re-doing.


anyway some observations:

I had the 12" box sub set-up like this; low pass 140hz, phase normal, pretty basic settings



It was actually facing the corner, along the 34" triangle face, to simulate same driver plane as outward corner box would, hopefully this was ok?



A "few" measurements later....and extracting charts....(for people who have not done this, this is the part that needs attention, putting good notes into each measurement group so later don't get confused)

I did each seat position like this, mic same position of course:

-12" box sub separately

-IB sub separately

-both subs running


Seat#1; you can never use the avg to join 2 subs measurements, , my early guess said should be able to but it never matched the combo sub measurement.



Seat#2 & Seat#3
.



Seat#4




From those 4 charts, hard to see;

-I find it hard to grasp why the box sub pulls down the IB sub consistently in it's roll-of below 27hz area......I'd love to have a real capable sub to test this out, one that goes at least 15hz before roll off.

-Next round of charts I'll not show the "avg plot", that confuses


Next post: seats 5-8
 
#18 ·
...continuing with seats 5-8


seat 5



seat 6 & seat 7
.




seat 8




Extracting just the combined charts for each seat 5-8:

I went from a consistent seat-seat response like this:



To this "mess":




My conclusion:

The box sub is "messing up" with the 2nd row.


Next up:

I put the 12" box sub in various positions along the LH wall 5ft, 6ft and 7ft from front wall, and took measurements......kinda revealed some stuff.
 
#19 ·

Quote:
Originally Posted by mtbdudex  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22738128


Seat#1; you can never use the avg to join 2 subs measurements, , my early guess said should be able to but it never matched the combo sub measurement.
Averaging won't work because it doesn't consider phase, but REW has a math feature that does. On the All SPL tab you can do A+B.


Not to belittle your efforts, but you could be drawing completely all sorts of short sighted conclusions if you don't consider relative phase and the effect that applying a delay to one of the subs vs. the other can have in the resulting frequency response.
 
#20 ·
^^^

+1


I'd add that you should not rule out any position until you tried all phase/LPF/volume variations there. RTA feature in REW is very helpful for this task.

I normally roll sub(s) around the room on a dolly cart, tweaking controls and watching the effect at the LP real time in REW RTA window.
 
#21 ·

Quote:
Originally Posted by Stereodude  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22738201

Quote:
Originally Posted by mtbdudex  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22738128


Seat#1; you can never use the avg to join 2 subs measurements, , my early guess said should be able to but it never matched the combo sub measurement.
Averaging won't work because it doesn't consider phase, but REW has a math feature that does. On the All SPL tab you can do A+B.


Not to belittle your efforts, but you could be drawing completely all sorts of short sighted conclusions if you don't consider relative phase and the effect that applying a delay to one of the subs vs. the other can have in the resulting frequency response.
Quote:
Originally Posted by zheka  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22738474


^^^

+1


I'd add that you should not rule out any position until you tried all phase/LPF/volume variations there. RTA feature in REW is very helpful for this task.

I normally roll sub(s) around the room on a dolly cart, tweaking controls and watching the effect at the LP real time in REW RTA window.

Guys, I'm a hobbyist so appreciate all help/insight.

Basic Q's:

1) Is this 12" box sub doing a decent job guiding me for 2nd sub placement, since it's got a HPF at 27hz or so?

2) I really would have liked the LH corner to give better results....that would be ideal, stereo subs behind the mains, hidden in the room, etc....what more can I do there to simulate better?


If there is a thread/info that shows me how to go about this please point me there and I'll read/learn/absorb and do it.


fwiw, I just went into REW and here is the A+B, which still does NOT come close to the real time combo measurement.

Gold = 12" box sub

Green = IB sub

Red = combo - both ran together

Purple = A+B



I did measurements along LH wall at 5', 6', and 7' locations from front wall.



Honestly, due to room layout I can put added subwoofer(s):

-on LH side corner thru 7' back, more then it would intrude too much into 1st row seat area

-anywhere along front wall

-Behind 2nd row seat either corner and anywhere along the back wall

Room layout:



I need some "quick" way to get a go/no go for each spot as I put the sub in perhaps many locations considering 0.5 ft increments ...

Or a gut feel guide on some best chance for success places (1/3 location, LH rear corner as it's opposite the IB sub...etc)
 
#22 ·

Quote:
Originally Posted by mtbdudex  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22739070



I need some "quick" way to get a go/no go for each spot as I put the sub in perhaps many locations considering 0.5 ft increments ...

Or a gut feel guide on some best chance for success places (1/3 location, LH rear corner as it's opposite the IB sub...etc)

Here is what I would do.


Get REW running in RTA mode and set the mike at the MLP.


Set the sub in the position you want to evaluate and start adjusting controls while watching the RTA screen for effect it makes at the MLP . I found volume and LPF settings most useful. My second sub runs quieter than the main and goes only upto 60Hz or so. Continues variable phase/time delay maybe very useful, but all I could do with my sub is switch between 0 and 180.


Find settings that produce the smoothest FR at the MLP then measure other seats.


Move to the next spot and repeat the steps.


At the end you should hopefully have one position that is markedly better than the others and the IB solo. You may tweak controls further in favor of better seat to seat FR which would probably mean losing some at the MLP. I am personally too selfish to do this.


If you have not worked with RTA before, here is something that may not be clear from the manual:

1. Use RTA mode, not "spectrum"

2. Proper signal to use is called "pink PN"


Good luck!


If you have not already, read that article on Geddes multisub I linked before
 
#23 ·
I like the corner build, but it might do a lot of "dancing" around if it is not bolted to the wall or restrained in some manner. A friend of mine tried this aproach and it was not useful until he fastened them to the floor and wall to eliminate the rocking of the enclosures. Just a thought.


Keep cranking,


Robert
 
#24 ·

Quote:
Originally Posted by mtbdudex  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22739070


fwiw, I just went into REW and here is the A+B, which still does NOT come close to the real time combo measurement.
I've found A+B predictions to give very accurate results to the actual measurement. So close the prediction and actual sit on top of each other and you basically can't tell there are two traces.


Does your receiver / pre-pro have 2 subwoofer outputs? Or how do you have these connected? Are you using a sub EQ with multiple outputs and delay capability (MiniDSP, QSC DSP-30, DCX2496)?


My first suggestion is to find a place where the sub has an opposing response to your IB. IE: peaks and valleys in the opposite places. That way when you blend them you'll get a flatter net response before equalization.
 
#25 ·
I'd imagine the fact that the subs are not identical and one of them is a four driver IB may complicate the A+B calculations.


I have to try the feature one day. I am surprised it can give accurate predictions for in-room source summing at all given how much the response is affected by the room itself.
 
#26 ·
Due to Christmas preparations and travel, this multi-sub analysis will take a break and resume Thursday Dec-27th @ full speed.
Quote:
Originally Posted by zheka  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22734312


There is no guarantee you will get good results with symmetrical position for the second sub. You are likely to have better chances with the second sub away from the front wall, perhaps along the left wall closer to the lseating area.

It does not need to be a powerful or very extended subwoofer. Multisub is useless below 30Hz or so anyway. But it would greatly help if the sub has variable LPF, phase/delay and volume controls .


Good read on the subject
http://mehlau.net/audio/multisub_geddes/

zheka - I have to apologize, the article you posted earlier is spot on what I needed for multi-sub analysis! I have read it, and will re-read it and apply it.

I've read the multi-sub concept approach of course before, but never saw a good tutorial how to implement, that should be added to sticky here for others..

Thx for the Christmas gift

Quote:
Originally Posted by zheka  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22739234

Quote:
Originally Posted by mtbdudex  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22739070



I need some "quick" way to get a go/no go for each spot as I put the sub in perhaps many locations considering 0.5 ft increments ...

Or a gut feel guide on some best chance for success places (1/3 location, LH rear corner as it's opposite the IB sub...etc)

Here is what I would do.


Get REW running in RTA mode and set the mike at the MLP.


Set the sub in the position you want to evaluate and start adjusting controls while watching the RTA screen for effect it makes at the MLP . I found volume and LPF settings most useful. My second sub runs quieter than the main and goes only upto 60Hz or so. Continues variable phase/time delay maybe very useful, but all I could do with my sub is switch between 0 and 180.


Find settings that produce the smoothest FR at the MLP then measure other seats.


Move to the next spot and repeat the steps.


At the end you should hopefully have one position that is markedly better than the others and the IB solo. You may tweak controls further in favor of better seat to seat FR which would probably mean losing some at the MLP. I am personally too selfish to do this.


If you have not worked with RTA before, here is something that may not be clear from the manual:

1. Use RTA mode, not "spectrum"

2. Proper signal to use is called "pink PN"


Good luck!


If you have not already, read that article on Geddes multisub I linked before

Good advice, I'll have this post on my 130" screen as I'm doing this Thur Dec-27.
Quote:
Originally Posted by robertcharles  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22739339


I like the corner build, but it might do a lot of "dancing" around if it is not bolted to the wall or restrained in some manner. A friend of mine tried this aproach and it was not useful until he fastened them to the floor and wall to eliminate the rocking of the enclosures. Just a thought.


Keep cranking,

Robert
I also desire the 2nd front LH corner for the hidden approach......possible if I go down that road then I'd be forced into the 4 corner sub approach for flat freq response, instead of the 3 sub approach as laid out in the article zheka linked?
Quote:
Originally Posted by Stereodude  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22739598

Quote:
Originally Posted by mtbdudex  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22739070


fwiw, I just went into REW and here is the A+B, which still does NOT come close to the real time combo measurement.
I've found A+B predictions to give very accurate results to the actual measurement. So close the prediction and actual sit on top of each other and you basically can't tell there are two traces.


Does your receiver / pre-pro have 2 subwoofer outputs? I have the new Denon 4520CI with true independent subwoofer pre-outs

Or how do you have these connected? For daily use the IB sub goes thru a FBQ2496 Feedback Destroyer in bypass mode (long story), then into the AVR, then into the EP2500.

For these REW measurements I bypass the AVR EQ by going thru the Ext In sub input, the 12" box sub is also bypassed via direct cable to it's plate amp, both come from the Mobile Pre directly split by Y cable.


Are you using a sub EQ with multiple outputs and delay capability (MiniDSP, QSC DSP-30, DCX2496)? For these measurements, no. I have a FBQ2496 Feedback Destroyer but not using it.

My first suggestion is to find a place where the sub has an opposing response to your IB. IE: peaks and valleys in the opposite places. That way when you blend them you'll get a flatter net response before equalization.

Stereodude - Response to your Q's above in RED bold, I have the new Denon 4520CI with true independent subwoofer pre-outs.

I also have for EQ the FBQ2496 Feedback Destroyer, prior with my 4308CI I ran some filters on that, but then I just let Audyssey handle the bass
and as "ok"

Your suggestion - find peaks/valleys in opposite location....YES, but the $ question is where is that?

My room really is non-symmetric from LF/ULF viewpoint:

LH/back walls true outside are concrete foundations walls, RH/Back walls are 2 x 6 sill plate with double offset stud construction, plus there is no "door", I have a 6' opening back RH side as semi-multi purpose room for game night.

I'm thinking the LH rear corner may be a good place to "try" as opposite...but I also thought the LH front corner would excite modes the RH side did not....
..

Quote:
Originally Posted by zheka  /t/1446659/diy-triangle-sub-w-dayto...s-x-34-face-x-32-tall-5-5-cu-ft#post_22739958


I'd imagine the fact that the subs are not identical and one of them is a four driver IB may complicate the A+B calculations.


I have to try the feature one day. I am surprised it can give accurate predictions for in-room source summing at all given how much the response is affected by the room itself.
Possible the harmonics of the fundamental of each sub in room response may cause the A + B REW summation not to work?

From basic wave theory it should work......


As clean-up to this thread, I'll re-size some of the charts I posted prior as they kinda clutter the thread....
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top