Sealed Sub for Acoustic Testing

[audiothings]

I build studios for my friends, and figuring out what needs to be done in the 20-125 Hz range is the toughest part. Typically, the sub, the measurement mic and the "trap" being tested are all in different tri-corners. This requires a sealed subwoofer, as we are primarily looking at decay times at various frequencies, and a ported sub will intrinsically have an extended decay time around the port's design frequency.

SO. What sealed sub can I build, which is extremely accurate in the 20-125 Hz range?

Thank you...

[Top Shelf]

Hi , what is the sensitivity of the monitors? Do you you have driver measurement software? Typically a Qtc tuning of 0.707 will give good results, not too boomy(1.1) not too bland(0.5).

[aackthpt]

Hey, good to see you here audiothings (I read GS a fair bit but I mostly lurk).

Are you looking specifically for a sub that has flat response in the frequency range you mentioned in 2pi space (ground plane measurement)? Without equalization? Because really since that application doesn't really require lots of volume (compared to home theater for example) it wouldn't necessarily take a lot of sub and it could be equalized into flatness without the room involved. Another approach would be to measure the 2pi response of your sub and use it as a correction curve of sorts to response the sub's influence from the measurement - same as equalization but from another angle.

The best source, but it is mostly focused on maximum ability rather than flat response (usually HT types equalize the subs) for sub measurements is Data-Bass . Josh Ricci has kind of taken over for Ilkka whose tests are at Subwoofer Tests - Archived at Home Theater Forum and Systems - HomeTheaterShack.com . There are also threads cataloging DIY sub projects at AVS forum and HT shack.

Do you have a weight or size limit in mind? Are you looking for a kit or a ground-up build? Are you looking for a self-powered sub? A guy from AVS forum has been working on getting some interesting flat-pack stuff going with Parts Express that might work for you.

Note that it would be possible (though it may or may not be practical) to use a ported sub tuned to 15 Hz or below. These aren't uncommon to build from Sonotube and if done cleverly could possibly weigh less than the sealed sub. I'm not saying it's a better option, merely that you may want to keep your options open.

[audiothings]

Aackthpt,

Thank you for your patient reply.

I have very little understanding of speaker building. And I know it will take me years to know enough. So I will depend on people who I think I can trust. As I understand, for my application, I should be looking for a low Fs (~20 Hz) and a high Qts (> ~ .40)... These are my main criterion at this time, till better advice prevails...

Are you looking specifically for a sub that has flat response in the frequency range you mentioned in 2pi space

As I understand, it is 1/8th space or 1/2 pi... but yes, flat frequency response, in the room corner, on the floor.

Do you have a weight or size limit in mind?

Anything which would fit in the back seat of a small hatchback. Actually, if it will fit in the boot, maybe I can use it in my car, when I'm not doing acoustic measurements This would definitely be an optional/added bonus, not a priority. Weight is not too much of a consideration. I usually have an assistant to do the ox work

Are you looking for a kit or a ground-up build?

Ground-up build. The big companies are represented here in India. Eminence is freely available and has great pricing. BMS, Beyma, JBL, P-Audio, 18 Sound, B&C, Faital Pro, Ciare, EV, Radian... most of these guys are available here. Stuff like Dayton Audio or the AE speakers would be a pain to import, and end up much too expensive.

Its easy for me to get across stuff that doesn't weigh too much, like miniDSP. However, transporting a 12" subwoofer is a problem.

If I can use the Eminence LAB12, (maybe with a a DSP implemented Linkwitz Transform?), it would be perfect. I can afford it, and it is available easily. But it has a Qts of .36, which is less than ideal for a reasonably small sealed enclosure?

A friend of mine bought a pair of (the now discontinued) Exodus Tempest-X2 subs, on my recommendation. They are lying unused, and I might be able to buy one off him. It would still be much more expensive than the LAB12, but it is a possibility.

Are you looking for a self-powered sub?

If it can't double as a sub for my car system, i'd prefer it to be self powered, so it'd be relatively easy to carry around. If I can use it in my car, I'd keep it passive so that it can be powered by car-appropriate amplification when in the car, and something that runs off 220V when in a room.

The best source, but it is mostly focused on maximum ability rather than flat response (usually HT types equalize the subs) for sub measurements is Data-Bass

Spent quite some time on the site. I found this to be quite interesting... over my budget right now, but I can wait...

Josh Ricci has kind of taken over for Ilkka whose tests are at Subwoofer Tests - Archived at Home Theater Forum and Systems - HomeTheaterShack.com . There are also threads cataloging DIY sub projects at AVS forum and HT shack.

Glanced at the link. So much in there... Will get into it when time permits...

Thanks again.

[cbdb]

For testing you need to shake the place (unless your looking for things that resonate, then power is all you need) so something small, if flat from 20 to 200hz will do. It dosnt even need to be low distortion, the test equipment will filter that out.

Actual a sub that has range from 30 to 200hz will still work becuase they still put out 20hz, just lower level, you just adjust your measurement (EQ) for that. And much smaller lighter cheaper.

[cbdb]

Small low, loud and clean is possible, but its not easy and not that light. Less than a cubic foot but 33lbs:MiniVee 8" Subwoofer

[aackthpt]

Originally Posted by audiothings

... low Fs (~20 Hz) and a high Qts (> ~ .40)...

These are parameters of the driver itself, when what you really need to pay attention to is the response of the finished subwoofer. But for help with driver selection, this article is a pretty good writeup.

As I understand, it is 1/8th space or 1/2 pi... but yes, flat frequency response, in the room corner, on the floor.

You will be using it [primarily] in 1/8th space (maybe less, depending on the size of room you are testing). But the sub can't really ever be tested to have "flat frequency response" in a room, because it's kind of the point to figure out IF the room has flat frequency response (in other words this statement doesn't seem to have a lot of meaning in this context). So the benchmark for sub testing, as you see on data-bass, is 2pi space. Now below the first room mode you will get "room gain" which is theoretically the inverse of sealed sub rolloff. Therefore if you build a sealed sub with a frequency response that begins to fall somewhere in the range of the first mode you have the best shot to measure flat all the way down. This, and the typical size, probably makes a sealed sub your best bet. Your problem is... you will be testing it in a different room every time, firstly, and secondly that what you are trying to do is test the room, not equalize the sub flat in that room! You can see that pretty quickly we get to the old "what is the reference" question which is why I make the point about the sub being "flat" _in which condition_ ... 2pi generally being the accepted reference in the industry, with the first (and main) advantage being that everyone can do the test equally.

Anything which would fit in the back seat of a small hatchback. Actually, if it will fit in the boot, maybe I can use it in my car, when I'm not doing acoustic measurements

I suggest measuring the size available and using that as a benchmark whether you design the sub, find one, or someone designs it for you. That Indian hatchback may be significantly different from my American one, you know!

Its easy for me to get across stuff that doesn't weigh too much, like miniDSP.

If I can use the Eminence LAB12, (maybe with a a DSP implemented Linkwitz Transform?), it would be perfect. I can afford it, and it is available easily. But it has a Qts of .36, which is less than ideal for a reasonably small sealed enclosure?

If you checked out the EBP above you can consider this yourself. I'd suggest you download WinISD Pro (alpha) to simulate with, as it's pretty simple even for a beginner, and the price is right (free). Also, you are on the right track with the LT circuit suggestion. There is an excel sheet for doing it on the miniDSP if you look in the forum there. But my understanding is that you can do similar with a Behringer DSP2496 or similar device.

Here are a few LT sub pages that might interest you, use drivers from lines you mentioned, and link to other useful information or have good design documentation: ESP sub, halfgaar (which takes you all the way from WinISD simulation to finished product, though it would be more efficient to do with a DSP!), Linkwitz Thor.

Spent quite some time on the site. I found this to be quite interesting... over my budget right now, but I can wait...

Yep, TC sounds is pretty much considered to be at the top of subwoofage. Here is the rig all the bassheads wish they could afford.

So in summary, yes go sealed. A 12" would probably work with the right driver, but smaller is possible. You need to probably take one of several approaches:

1. try to put the LF dropoff below 20Hz (initially and without EQ) so the whole thing is flat in 2pi (might require some higher-level equalization, but in any case you really should do a good 2pi test such as in a very large parking lot or in a field with wide tiled area laid down between mic and sub), then you would know for sure what the sub-first mode room gain was for a given room (which might aid you in choosing or sizing subwoofers in combination with customer's/friend's goals)
2. Build a sub that can be Linkwitz Transformed to "flat to 20" (whatever you mean by that) in 2pi, for the same purposes as above, also the sub may be able to be smaller? This sub might be better for the car, too, where you need a higher frequency LF cutoff (due to the very high first mode of the car and consequent vastly increased cabin gain)
3. Build a sub, again presumably smaller, that you would L/T differently for each room being tested to match the low corner to the lowest room mode. This because if you are not trying to get home theatre infra-bass, for acoustical testing purposes (in which you are mostly looking for SBIR/modal issues) you only need data down to the first mode (and to be conservative you could define the first mode by the room diagonal definition rather than between two straight walls for example). Sounds like a PITA to me, and not probably worth the effort.

There very well may be more approaches that would yield useful results. I claim to know FAR from everything in this arena.



Also, I would say to consider that the tri-corner type testing is mostly just to identify all of the modes (as a diagnostic test), not to test for quality of listening conditions as they might be in the finished studio. You could use the sub you build also, though, to test different subwoofer locations for the studio with the mic in the listening position rather than up in the tricorner like for the initial test. And don't forget that no matter what you do with the test sub you will probably need to make adjustments with the final speaker system including the sub in place - and I'm sure you realize those adjustments potentially include treatments and/or equalization depending on goals for the room, type of music or film mixed there, whether a plurality of listening positions is desired, and probably plenty I haven't thought of.


Hope you find this more-extended discussion useful.