Experienced Opinions Requested : Sub Design Choices

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi All,

I'm looking for some opinions on a choice between 2 designs for my home theater sub. Due to spousal restrictions I'm confined to placing the 60" TV, the sub and the front speakers into an alcove 10 feet wide, 7 feet high and 2 feet deep. This should be plenty of room. The sub should be able to produce 106.5 dB in room and down to at least 18 Hz. Now I've selected and purchased 2 Creative Sound SDX-15 woofers and 2 Acoustic Elegance 18" 2500 gram passive radiators. My first thought was to go isobaric to reduce box size. This goes nice and flat to about 18 Hz (see below). I then thought about room gain and tried a normal passive radiator design and the output drops like a 0.5 Q sealed box but has a bit more output. This would allow for placing the woofers further apart and may help with room modes. The Right and Left speakers will be able to get down to maybe 28 Hz.

I will be using the soon to be released Outlaw Audio pre/pro that has room correction abilties so it can boost or cut the lows is required.

My request is that I get some opinions on what the pros and cons for the designs are. Any and all opinions appreciated, even if I don't agree!:D
 

Attachments

  • PR Excursion CSS 2 SDX15 Isobaric 2 PR1825.gif
    PR Excursion CSS 2 SDX15 Isobaric 2 PR1825.gif
    42.6 KB · Views: 313
  • PR Impedance CSS 2 SDX15 Isobaric 2 PR1825.gif
    PR Impedance CSS 2 SDX15 Isobaric 2 PR1825.gif
    43.8 KB · Views: 310
  • PR Response CSS 2 SDX15 Isobaric 2 PR1825.gif
    PR Response CSS 2 SDX15 Isobaric 2 PR1825.gif
    47.1 KB · Views: 304
  • PR Step Response CSS 2 SDX15 Isobaric 2 PR1825.gif
    PR Step Response CSS 2 SDX15 Isobaric 2 PR1825.gif
    39.3 KB · Views: 301
  • PR Excursion CSS 2 SDX15 2 PR1825.gif
    PR Excursion CSS 2 SDX15 2 PR1825.gif
    38 KB · Views: 305
  • PR Impedance CSS 2 SDX15 2 PR1825.gif
    PR Impedance CSS 2 SDX15 2 PR1825.gif
    42.3 KB · Views: 29
  • PR Response CSS 2 SDX15 2 PR1825.gif
    PR Response CSS 2 SDX15 2 PR1825.gif
    45.2 KB · Views: 31
  • PR Step Response CSS 2 SDX15 2 PR1825.gif
    PR Step Response CSS 2 SDX15 2 PR1825.gif
    38.1 KB · Views: 31
Personally, I think isobaric is a waste of a driver as it doesn't improve low-end. It just drops the top-end. It's a nice idea when the driver aren't expensive, though. I like your choice of components.

What's the longest dimension of your room? Your space might have some room gain, so you might not need your speaker to go to 18Hz to get there. If these sims aren't taking into account the 1/4 space boundary effect I'm assuming you'll have also, you'll see a 6 dB jump in output compared to the sim's free space assumption (that I'm assuming it has).

If this was me, I'd choose the 330 liter alignment.
 
If these sims aren't taking into account the 1/4 space boundary effect I'm assuming you'll have also, you'll see a 6 dB jump in output compared to the sim's free space assumption (that I'm assuming it has).

The sims are free space. I've seen articles stating either 3 dB or 6 dB per octave gain under about 80 Hz depending on the size and shape of the room and the location of the listener. Since I'm not going to do the calcs on room gain and find all the nodes etc, I'm leaning towards having 2 subs spaced about 8 feet apart in the alcove. More sources should result in a more chaotic distribution of those nodes. At least that is my guess.
 
Looks to me like you won't get much room/boundary gain at all. With the space opening wide to the right and far back, I don't think much pressure containment will happen. The 45 degree walls on the left (that I'll assume are curtained windows) might make some LF focusing, but probably not a strong antinode. When you close your eyes and listen, the room might have a left heavy tone in the LF region. The parallel back wall to front wall will make a resonance at around 100 Hz, but due to the opening on the right side, I doubt you'll have any nodes to be concerned about. If 100 Hz seems empty in the center of the room, you might consider some helmholtz resonators on the back wall tuned for 100. Sonotubes might work well as their round shape can also act as a specular diffuser if wall mounted.
 
Looks to me like you won't get much room/boundary gain at all. With the space opening wide to the right and far back, I don't think much pressure containment will happen. The 45 degree walls on the left (that I'll assume are curtained windows) might make some LF focusing, but probably not a strong antinode. When you close your eyes and listen, the room might have a left heavy tone in the LF region. The parallel back wall to front wall will make a resonance at around 100 Hz, but due to the opening on the right side, I doubt you'll have any nodes to be concerned about. If 100 Hz seems empty in the center of the room, you might consider some helmholtz resonators on the back wall tuned for 100. Sonotubes might work well as their round shape can also act as a specular diffuser if wall mounted.

Thanks for the info. This is actually a basement and I have a thing for 45 degree corners. The right bottom opens into a small "office" pod. A 12 foot by 7 foot room with 2 more 45 degree corners and filled completely with a wrap around desk and upper cabinets. The opening on the upper right is a landing/foyer of sorts with doors for 2 bedrooms and the stairs giong to the main floor.

Does this all mean I should try for a flatter response from the sub?
 
The good news is with that very irregular space you should have pretty randomized modes other than the floor to celling axial modes. So limited placement options may matter less. As for room gain, little way to know.

Thanks for the feedback. The ceiling is just short of 8 feet. The joist space is 14" which has been filled with a combination of fiberglass, a floating layer of drywall and rock wool insulation. The surface consists of drywall with square and rectangular cutouts (28"x28" and 23"x28") that will be covered by cloth covered picture frames. The floor is a laminate on a OSB subfloor which is spaced up about 3/4" over concrete. Those are my uneducated attempts at reducing sound transmission and reflection.
 
That's a pretty huge space, so I don't see why you would want to go isobarik at all. Much better to get the two sub drivers next to each other along the viewing axis. That creates a kind of virtual acoustic pressure wall between the two.

I'm not sure what you have in mind by a "flatter" response, but something with a gentler rolloff is probably good. Don't worry about just f3, look at f6 and f10 etc.

As for "More sources should result in a more chaotic distribution of those nodes."-nah. I don't think so. The nodes are just from geometry at the LISTENING position. The sub positioning just excites some modes or not. Multiple sources may randomly not-excite resonances at the listening position and happen to sound better, but that's luck not acoustics. Better to load the drivers next to each other, adjust the listening position, and apply time-domain EQ (is that what Outlaw is bringing?)
 
I'm not sure what you have in mind by a "flatter" response, but something with a gentler rolloff is probably good. Don't worry about just f3, look at f6 and f10 etc.

By flatter I meant f3 being as low as possible. If I were to design to f6 or f10, would I be counting on room gain or equalization to get the output up at these frequencies?

As for "More sources should result in a more chaotic distribution of those nodes."-nah. I don't think so. The nodes are just from geometry at the LISTENING position. The sub positioning just excites some modes or not. Multiple sources may randomly not-excite resonances at the listening position and happen to sound better, but that's luck not acoustics. Better to load the drivers next to each other, adjust the listening position, and apply time-domain EQ (is that what Outlaw is bringing?)

Thanks for the advice. The to be released Outlaw pre/pro incorporates Trinnov software and I believe it does time-domain EQ plus more.
 
By flatter I meant f3 being as low as possible. If I were to design to f6 or f10, would I be counting on room gain or equalization to get the output up at these frequencies?.

Yes, I suppose. But keep in mind f3 is just an arbitrary quantity that Richard Small chose for mathematical convenience kind sorta.

In your case, I'd model the "room gain" as 3 distances: woofer to floor, woofer to ceiling, and another "wall" at 1/2 the woofer to woofer (presuming they're symmetrical around the viewing axis). Your other boundaries are farther away, and will affect only progressively lower frequencies. Hopefully you have or can get a software that can model your box design + boundaries. Keep in mind to simulate also the position of the passive radiators. Their position may be more important, since a lot of the bass from a sub can actually come from the port/radiator.0


Thanks for the advice. The to be released Outlaw pre/pro incorporates Trinnov software and I believe it does time-domain EQ plus more.

Sure, your bill is in the mail :-D
Seriously, the Trinnov bit looks interesting, I'll have to look into that. Makes me wonder if those Audyssey receivers can store multi-setups, so I can listen off-axis on the couch, but still maintaining "perfect sound forever"
 
Hopefully you have or can get a software that can model your box design + boundaries.
I'll be learning how to use SoundEasy which I believe isn't all that easy.


Makes me wonder if those Audyssey receivers can store multi-setups, so I can listen off-axis on the couch, but still maintaining "perfect sound forever"
Not sure about multi-setups but you can set the test microphone at multiple locations and it calculates a best overall response/EQ curve.
 
I'll be learning how to use SoundEasy which I believe isn't all that easy.

I was going to recommend LEAP instead, then I looked at the pricing. Wow! For pros only, I guess, though it's great software.

For what you're doing, something simpler would probably work fine. Nothing that I'm aware of is going to model your odd room, so forget that. You just need a simple package to model the 3 boundaries and look at how to juggle the spacing to get a smooth lift. Then a simple speaker package to look just below say 100 Hz, and try and design a low frequency rolloff complementary to the "room" boost (remembering that one of the "walls" is the imaginary one between the two subs).

That's not to talk you out of SoundEasy, just that unless you really want to spend a lot of time measuring and refining, simpler packages might suit you fine. But I'm not too familiar with what's out there currently, so that's a whole other thread :eek:
 
I think you will find that guesswork and simulation all goes out the window when you actually look at real measurements and see +/- 20 db peaks and dips! Anyone fancy playing darts blind folded?! That's how I view guesses about room modes and room gain. It's not hard or expensive to find out. Just download REW for free and combine it with an SPL meter calibrated and a mic preamp. All easy and cheap.

Then you will know what you are dealing with. You measure your sub near field (as close as possible to the cone) then in the listening positions. You'll find out what your sub is doing and what your room is doing. Then you put your sub in the listening position and put the mic all over and measure all the curves. Quick and easy. It may be that another location that you could get approved by the boss might actually work really well. If you have the drivers I'd go with two subs and try to optimise their locations.

It generally works this way. If one sub has a dip and the other doesn't, the dip won't show up when combined. If you sum a peak and a dip, they won't average, the peak will dominate the response. So you arrange your subs so that most of the dips are eliminated, then a little eq can pull down the peaks and sometimes give you quite a bit of extra headroom.

I'd use Win ISD pro for simulations - you should factor in all filters used. Add in a low pass and things look very different. Also do you have a rumble filter? Otherwise you are throwing away a lot of your potential output and wasting a lot of excursion. I'm assuming that with a pair of 15" high excursion drivers you want to get quite a bit of extension and depth. Why throw it away? The right rumble filter will ensure that the cone excursion below tuning will never exceed excursion above. All it takes is a movie with insane 5 Hz output and you'll send the cone to xmech. You either live with the risk or turn them way down to avoid it. I'd just rather put in a rumble filter and know the below tuning excursion will never creep up and get out of control. Go ahead and call me a control freak.
 
frugal-phile™
Joined 2001
Paid Member
You biggest problem with modes in a room like that are going to be floor to ceiling. Something like 70-80 Hz in your case.

Toole has shown that F3 is meaningless as far as bass is concerned. I use F10 as a metric.

4 drivers, i'd build 4 low Q sealed boxes and stuff one in each corner of the alcove (given thats where they have to go) such that a pair are at the floor and a pair are at the ceiling. Each should have a seperate amplifier. A pair of towers the height of the room, the depth of the alcove and as skinny as necessary would be one solution.

dave
 
Just download REW for free and combine it with an SPL meter calibrated and a mic preamp. All easy and cheap.

Thanks for all the advice. I'll look into REW.



Also do you have a rumble filter? ..... The right rumble filter will ensure that the cone excursion below tuning will never exceed excursion above.

I am waiting until Outlaw actually releases their new pre/pro to see if it has a low "high pass" filter for the subs. If they don't, depending on the design response curve and the power of the amp, I may add a rumble filter.
 
You biggest problem with modes in a room like that are going to be floor to ceiling. Something like 70-80 Hz in your case.

dave

The ceiling will have 36"x36" panels (picture frames covered in speaker cloth) to let the audio energy into 14 inches of rock wool insulation. I'm hoping that tames a bit of the problem.



Toole has shown that F3 is meaningless as far as bass is concerned. I use F10 as a metric.

dave

Good to know. I'll incorporate that into my planning.


4 drivers, i'd build 4 low Q sealed boxes and stuff one in each corner of the alcove (given thats where they have to go) such that a pair are at the floor and a pair are at the ceiling. Each should have a seperate amplifier. A pair of towers the height of the room, the depth of the alcove and as skinny as necessary would be one solution.

dave

Actually that is fairly similar to what I am now planning after getting feedback here and doing lots more reading. I hope to get the woofers of my R/L speakers playing down to 25 Hz (F3) which is in sub territory. These will occupy the bottom right and left of the alcove. The subs (I've decided to make 2) will occupy the top right and left in the alcove. After that I'll be counting on EQ to fix any issues.
 
The ceiling will have 36"x36" panels (picture frames covered in speaker cloth) to let the audio energy into 14 inches of rock wool insulation. I'm hoping that tames a bit of the problem.

I agree, it probably won't make a tremendous change actually. The very low frequencies can "penetrate" the rock wool too easily. Bass traps tend to be more flapulent, to coin a term. Mass loaded rubber and stuff like that. Usually I see bass traps in corners.

So if you want to do the ceiling as decor, that's fine, but don't kill yourself trying to fix the bass like that. Whatever the effect is would just be to blunt the sharpness of the modes a little bit.

Hey, won't the cloth sag, by the way? Or do you mean like screens, not frames?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.