OK to use "non-final" subwoofer to determine approx. placement?

[NegativeEntropy]

Hey all,

Mixed use theater room will be done ~September. Drywall is about complete.

I have an older sub that goes to 20Hz pretty flat. I will be buying new subs, but quantity and model and even sealed vs ported is undecided.

Can I use REW with my existing sub in this complete but empty room (no furnitire, no carpet) to determine how the room behaves and therefore likely future sub locations?

Basically trying to determine how many good locations I have, use REW to sum them up, and then that will help me figure out how many subs I need (and output each) for a decent combination of SPL and proper response. Rather than guessing, buying more sub than I need, or coming up short.

Seems to me like it should work decently. Not perfectly, as the detailed FR will change with sub model, but the "room effect" should not?

Thanks!

 

[Hexspa]

I imagine it will be fine. Isn’t placement in a given room more important than the specifics of a subwoofer, assuming it’s of reasonable quality? What I assume the biggest difference will be is amount of distortion, bandwidth, and whether the physical dimensions of the sub make you put it somewhere slightly different.

For my current room, I placed the subwoofer and listening position before anything else. After furniture and acoustic treatment, the SPL response was similar to what I started with.

Eventually, I’d like to upgrade to a bigger sub but by then you’ll have your answer.

 

[droid2000]

I'm not super knowledgeable on this topic, but it seems like summing a single sub response at multiple locations won't take into account how multiple subs will constructively and destructively interfere with each other.

If multiple sub interference is negligible to the overall response, then your method should work (I think). But if interference plays a large part in the overall response (as it does in my setup), then you might not get an accurate estimate of the number of subs you need and where to place them.

Seems like the safest method might be to buy them one at a time and then measure the overall response. Although that would be time consuming if you end up needing a lot of subs.

I wonder if REW room simulator could help here?

 

[Chrispy]

I'd think if subs have similar response should work well enough, altho may not be ideal.

 

[NegativeEntropy]

I'm not super knowledgeable on this topic, but it seems like summing a single sub response at multiple locations won't take into account how multiple subs will constructively and destructively interfere with each other.

If multiple sub interference is negligible to the overall response, then your method should work (I think). But if interference plays a large part in the overall response (as it does in my setup), then you might not get an accurate estimate of the number of subs you need and where to place them.

Seems like the safest method might be to buy them one at a time and then measure the overall response. Although that would be time consuming if you end up needing a lot of subs.

I wonder if REW room simulator could help here?

The room is not rectangular, so simulations would be worse than typical. Since I have the real room (well, soon will), I want to use it

My understanding is that estimating combined responses is pretty straightforward, and the newish "alignment tool " (edit:cool) makes it quite straightforward:

 

[droid2000]

The room is not rectangular, so simulations would be worse than typical. Since I have the real room (well, soon will), I want to use it

My understanding is that estimating combined responses is pretty straightforward, and the newish "alignment cool" makes it quite straightforward:

That's amazing. I didn't know about that alignment tool. Thanks for the info (video).

 

[Dal1as]

Not a bad idea. You have all summer to test. Do it right.

Learn your room response and where all the nulls are especially.

The trick with multi sub in non standard rooms is to make sure the subs help one another and mask the faults of the other ones.

Will there at least be doors and outside walls?

Any schematics?

 

[Dal1as]

The room is not rectangular, so simulations would be worse than typical. Since I have the real room (well, soon will), I want to use it

My understanding is that estimating combined responses is pretty straightforward, and the newish "alignment cool" makes it quite straightforward:

I'm about to do a quick layout change in my theater and am going to try that tool (along side doing it manually) and see how it goes.

One thing to remember is the delay/distance setting in the AVRs is different than what you're doing with subs. Also why it's helpful to start with the farthest subs 1st.

He mentions that but doesn't explain it fully and it's helpful to know.

 

[NegativeEntropy]

Not a bad idea. You have all summer to test. Do it right.

Learn your room response and where all the nulls are especially.

The trick with multi sub in non standard rooms is to make sure the subs help one another and mask the faults of the other ones.

Will there at least be doors and outside walls?

Any schematics?

Room will be fully sealed (HVAC via joist mufflers). 2 interior doors (1 "door" is a set of french doors), 1 outside window, 1 patio door. This is a walkout basement. Double drywall, green glue, wall decoupling, acoustic sealant along all edges, etc. Concrete floor.

MLP is about 10' from screen wall, or about 38-40% of room depth away.

This drawing is approximate. The "lower left" has a vertical and horizontal soffit for return air joist mufflers and as sketched this shows the inside dimensions of that, but in reality it's open under the soffit, so the full step is closer to 2', not 4. That corner is a potential sub location.

My plan is to basically test the sub everywhere I could potentially have one (limited only by foot traffic patterns, fortunately - spouse approved speakers on stands so there will be room for at least sealed subs), or at least the classic positions of corner, 1/4 and center of each wall. I'm even considering having 2 (smaller) nearfield subs behind or beside the MLP. I know this is not the classic spot for smooth bass, but it would allow lower overall room SPL while still having good SPL at MLP - that way I bother the rest of the house less when others are trying to sleep. Mains and surround are capable so I could cross over the nearfield subs at 70 or even 60 Hz to help with localization.

Could then have a couple of combinations of minidsp + dirac for different listening conditions (e.g. every sub for regular listening, only 2 "optimal" subs for music, etc.). Still working on ideas. Also discovered Crowson tactile transducers, so those are on the table to help make up for potentially lower "night time" listening levels. As usual, too many ideas, not enough time.