Treatment of a small boxy room - REW file attached

[Greyhound]

Here's a slightly more refined version of the overall soffit geometry. I think this is getting close - next step will be the internal construction and integration with other treatment elements. The dimensions are in inches here.

[Greyhound]

Hello? Where did 2022 go!?

Argh - life got in the way of plans, as it will do... As I finish up some other building projects I am hoping to resume work on this home studio.

The only silver lining is that all the side projects have only added to my building capabilities so hopefully this will pay off in regards to the studio.

Hope all are well!

GH

[Greyhound]

What to do with 500 square feet of duct liner!?

I'm about to resume this home studio project, and still have around 500 sf (46 m2) of 1" thick Manson Acousti-Liner sitting around in 4'x8' sheets.

As mentioned earlier - the material is listed on Bob Gold's Absorber Page, and is considered somewhat similar to Owens Corning 703 of similar thickness. It has a smooth fabric facing on one side, that is reasonably permeable to airflow although I don't have any quantitative values for that (and it's not listed on the manufacturer's TDS).

I sourced it in quantity a couple of years ago, planning to incorporate it into an even older design which is now a bit outdated (the liner would have made an attractive "facing" layer for treatment elements backed with pink fluff or mineral wool in the original design). Based on more recent feedback, including from the original designer, I am now rethinking the treatment approach (unfortunately the original designer isn't available at the moment).

At the risk of letting the "tail wag the dog", I thought I'd ask if anyone had thoughts on how such a quantity of material might be incorporated into a modern studio design.

I realize this is a hugely open-ended question that depends on the site conditions and requirements - suffice it to say I'm working with the quintessentially bad 12'x12'x7' room described earlier. The entire surface area of wall plus ceiling is less than 500 sf. I have major issues from the predictable modes at 47Hz and 81Hz, along with all the other familiar small boxy room stuff.

I'm just getting back at this after a long interruption, and am looking (based on earlier recommendations in this thread) at pressure-based absorbers (Helmholz, panel, VPR) possibly incorporated into a Boggy-diffuser/MyRoom type of design. I realize that the three types of resonant/pressure absorbers I mentioned are tricky to design, but I have worked on similar applications in the RF domain and it seems like an interesting challenge.

One crazy idea was to cut the 16 sheets (each 4'x8') into 64 big triangles that could then be stacked into a huge superchunk (5' long or more with gaps). This could be mounted along the "soffit" zone at the intersection of the back wall and ceiling. It would extend downward to within three feet of the floor, which would be acceptable space-wise.

I realize that the latter approach is probably a hugely inefficient use of material compared with what can be achieved with pressure-based absorbers, but it shows how desperate I'm getting!

[gullfo]

i would cut them into 2' x 4' and stack 4 for a 4" thick absorber (ceiling, front and back walls), and 2 for a 2" thick absorber (wall first reflection). if they're stiff enough, just wrap in some cloth. for the corners - i might use 1 layer for the front span but back fill with the pink stuff.

[Greyhound]

Source of the post i would cut them into 2' x 4' and stack 4 for a 4" thick absorber (ceiling, front and back walls), and 2 for a 2" thick absorber (wall first reflection). if they're stiff enough, just wrap in some cloth. for the corners - i might use 1 layer for the front span but back fill with the pink stuff.

Thanks for the voice of reason and common sense! It's hard to go wrong with 2'x4', and these things are stiff enough that they could be reconfigured in all sorts of ways once they're cut into that format. It's actually a nice material to work with, so long as the mystical acoustic parameters work out. There's virtually no shedding of fibres or crumbling once they're cut, so very clean to more around into different configurations.

Will try this next week, and post some pre/post measurements. I'm also still planning on doing the speaker flush mounts (unless the prevailing wisdom on that has shifted, which I doubt).

[Greyhound]

If nothing else, I could cover my entire ceiling to four inches, but of course much more efficient to distribute as suggested.

[RJHollins]

just following ...

I used the KNAUF rigid fiberglass panels throughout my treatment.
They came 2'x4'x2"

When I built the 'Chunky Corner' Bass Traps, I had to cut into 2'x2' triangles.

Don;t know if you know .... but I found that an electric knife [like we'd use cutting up a turkey] worked great cutting this rigid
fiberglass. Just thought I'd pass that along.

best !

[Greyhound]

Source of the post just following ...

I used the KNAUF rigid fiberglass panels throughout my treatment.
They came 2'x4'x2"

When I built the 'Chunky Corner' Bass Traps, I had to cut into 2'x2' triangles.

Don;t know if you know .... but I found that an electric knife [like we'd use cutting up a turkey] worked great cutting this rigid
fiberglass. Just thought I'd pass that along.

best !

Thanks for the tip, and glad to hear that a similar material was useful in your treatment!

[Greyhound]

So I did manage to redistribute the duct liner that's been leaning behind my sofa for the last year, and I believe it's made a significant improvement.

As a reference, here is the waterfall for the bare room:

Last year I leaned 16 inches of Manson Akousti-Liner (in 1" sheets) against the back wall, as shown in my initial posts of this thread. The result was this waterfall:

.

This week I built a frame so I could put four sheets (4" thickness total) as a makeshift cloud over the listening position:

To make the cloud, I attached 2x4 lumber across the studs on the side walls, with two 12' spans separated by four feet. The insulation sheets were just lifted in between and I screwed furring strips along the bottom. It seems very secure, and avoids having to anchor anything to my ceiling (which is three layers of 5/8" drywall on resilient channel - it's already heavy, and anchoring is non-trivial due to the thickness). The insulation sheets are not touching the ceiling.

This also removed 4" of insulation from my "pile" against the back wall:

I happened to have some 3.5" slabs of mineral wool left over from another project (not acoustic - people actually do use this stuff for insulation ). For a rough treatment of the first reflection points on the side walls, I wrapped two slabs (total thickness 7") in some breathable landscaping fabric for each side, as well as a single slab located approximately where the front reflection points would be (at least for the tweeters, and as well as my current monitor placement would allow):

I will continue experimenting with the vertical placement of these (and maybe even a more aesthetic mounting system ) - the one on the right is constrained by a window:

The resulting waterfalls after this treatment are shown below. It's also worth noting that I added a subwoofer (Focal Sub6) in the last couple of weeks, but in the waterfall below it's disabled (for the sake of comparison with earlier tests). I also can't be sure that the REW output levels were exactly the same here, although I did the 80dBC calibration as prescribed by Stuart in all cases.

The new treatment (beyond just the back wall) appears to have reduced the time-domain ringing across the spectrum, although the most severe low-frequency modal peaks and dips remain.

Somebody will no doubt notice the large speakers sitting (disconnected) in the back corners of the room. These are a pair of old Kef 104/2 speakers (of "American Psycho" fame) that I have nowhere else to put. I was toying with the idea of using the enclosures to try and make some kind of resonant absorber, but will probably just go with a slatted Helmholtz made from MDF. I suppose the enclosures might be impacting the room acoustics - would be interesting to measure this. I've read some articles on how people put RLC networks along the speaker terminals to try and tune the resonant absorption, but as I recall Stuart was not a big believer in this.

PS the three additional attachments below show waterfalls acquired with the subwoofer, at various stages in the recent treatment. If nothing else, the side and front treatment (using my leftover mineral wool on the sides and front) seemed to have had an effect in the high frequencies.

[Greyhound]

In regards to the sub, I have not really started optimizing the placement and/or settings (which include overall level, crossover frequency, and continuously variable phase from 0-180deg).

If anyone has pointers on this, I'd be grateful. Keeping in mind, I still plan to take a stab at soffit mounting the two main monitors.

[Greyhound]

I work a lot with Matlab in my "day job" and have spent a few evenings trying to code acoustics equations from various sources into Matlab. I just realized that Cox and D'Antonio already provides an appendix with code in Matlab for the most important concepts and figures in the book!

This is freakin' amazing, as it takes out a lot of guesswork as well as angst about sketchy online Helmholtz calculators and the like. I realize this is still just theory, which is not a substitute for experience in real-world application of room treatment, but at least it provides some kind of solid foundation for experimenting.

Hopefully I'll be able to adapt their code to my room, and post some example results with testing.

[Greyhound]

Well the code seems to work - now to adapt the parameters to my situation...

This is Figure 6.11 from Cox & D'Antonio, showing effect of open area percentage in a Helmholtz absorber.

[gullfo]

yes, not only does the Q narrow but the frequency drops. which is why when trying to build LF HH you want to use the narrow coverage to target the problem frequency(s). if you need broadband then you need a pressure trap - like a VPR (steel on foam) or large panel (plywood on sealed container) or limp membrane (MLV). and of course you can stack these - an absorption front on a panel trap is fairly common (keeping the absorption off the panel ) and sometimes more layers of things - slotted panel for scattering over absorption over panel or HH to address several aspects at once. depends on depth available and what needs to be targeted. "broadband" slot resonator on angles in the front of the room with several chambers to improve efficiency are very effective.

[Greyhound]

Source of the post yes, not only does the Q narrow but the frequency drops. which is why when trying to build LF HH you want to use the narrow coverage to target the problem frequency(s). if you need broadband then you need a pressure trap - like a VPR (steel on foam) or large panel (plywood on sealed container) or limp membrane (MLV). and of course you can stack these - an absorption front on a panel trap is fairly common (keeping the absorption off the panel ) and sometimes more layers of things - slotted panel for scattering over absorption over panel or HH to address several aspects at once. depends on depth available and what needs to be targeted. "broadband" slot resonator on angles in the front of the room with several chambers to improve efficiency are very effective.

Thanks Glen,

I think something like this could work for my 47Hz mode:

This uses the C&D'A code to simulate a HH resonator with a 5-1/2" gap, filled at the front with 3.5" of mineral wool. The hole pattern would be 4" spacing on-centre, nominally with a 1/4" drill bit, in 3/4" MDF. In practice I'd titrate up starting with smaller bit sizes.

Something like this could be mocked up using a couple of sheets of MDF with 2x6 lumber forming the peripheral frame. A sufficient air seal for a prototype could probably be achieved using foam strips in all the joints, compressed when assembled with screws.

Not sure what the best way of checking the resonant frequency would be - I've seen this done with a measurement mic around one of the holes during a frequency sweep. Perhaps also rapping the unit with a small hammer, or generating an impulse response by firing off a Ramset charge (powder actuated fastener tool) nearby?

If it looks good, I'd build a couple of 2'x6' units for my back corners, and maybe two more 2'x4' units for the front corners if that seems reasonable.

I want to try and target the major LF modes with something like this first, before getting into the broadband stuff. I'm wading through Fuchs's book and patents on VPRs, but these are nowhere near as clear as Cox & D'Antonio (plus they provide code in a language I already work with!)

[Greyhound]

Playing around with the simulation code a bit more, it's easy to see how Helmholtz resonators have earned their reputation for being tricky to design.

Both the resonant frequency and bandwidth are sensitive to the resistivity and depth of absorbent material placed in the cavity. I naively expected this to be so for the bandwidth, but hadn't realized that it could also have a significant effect on the centre frequency.

Here are a couple of simulations for my hypothetical HH trap with a cavity depth of 5.5", with 1/4" diameter holes spaced at 4" intervals.

This first one simulates a 3.5" thickness of absorber in the cavity, with two different flow resistivity values (20,000 as is assumed by C&D'A for mineral wool, and a hypothetical higher value of 40,000):

Of course doubling the mineral wool resistivity is going to show an exaggerated effect, but it does demonstrate that the resonant frequency can shift with this parameter (in addition to the expected change in bandwidth/Q).

Another simulation shows the effect of different thicknesses of mineral wool within the 5.5" cavity:

Here too we can see a shift in frequency as well as bandwidth, although the differences are small between 3.5" and 5.5" mineral wool thickness. Underdamping (no absorber) seems like it will be a bad thing, as the HH could be expected to "ring" following an excitation so I'd be inclined to fill the thing with mineral wool.

Both of these results suggest that it could be a mistake to tune an "empty" HH resonator prior to adjusting the damping by adding absorber. It would seem necessary to make a principled decision on the type and amount of absorber to use, then tune the resonator (with absorber in place) by gradually increasing the hole sizes. Mind you, the difference in resonant frequency shown here (no absorber vs full depth of absorber) is only a few Hz and may not be of much practical significance so far as tuning is concerned (you still need absorber to avoid ringing).

The other question is how much room coverage area is required before this actually helps with a 47Hz mode. Doubtless the coverage needs to be large, which will introduce other room issues that would need to be addressed (ie having a 4'x8' wall of perforated MDF may not be ideal for mids and highs, even if it does tame a single low frequency mode). Hence the need for layering that Glen described.