Tiny All-In-One room

#46

Hello Stuart,
Thank you for your detailed clarification. Regarding the images I begun with some settings in REW (I don’t remember the reason why ..) and for reference with the older posts I kept them. Now that the room almost finishes I think I will switch to the defaults.
Here is the mdat:

dio Ministudio 73º Front limp mass overhauled.mdat: (7.03 MiB) Downloaded 747 times

dio Ministudio 73º Front limp mass overhauled.mdat: (7.03 MiB) Downloaded 747 times

But keep in mind that I have made progress from that “snapshot” and the room is now more complete. Anyway this mdat corresponds to what you see in the photos.
Thanks
Dio

#47

one of the nice things about a peak (e.g. 42hz) you can use EQ to drop the level there. dips (nulls) cannot be eq'd correctly but peaks you have the option to reduce the energy there and keep the overall response even.

#48

Here is the mdat:

Excellent! I just downloaded it and took a look at it: and overall, things are not as bad as they seem from the graph images posted earlier!

First, the 41 Hz peak:

... it really isn't totally terrible, and as Glenn already pointed out, it can be treated with EQ: A parametric filter applied at around 41 Hz with a cut of maybe 6 dB would probably do the trick. Adjust the bandwidth for the best result (but don't make it too wide! There's a phase shift at around 33 Hz, that you should stay away from). Another filter at around 280 Hz, -4dB and fairly broad would help even out that low-mid peak (there's no phase issues here, so you should be able to make it fairly wide: maybe half an octave or so.).

Next: There's a strange dip on all measurements at around 4.7 kHz:

That seems to be either your speakers themselves, or something in the signal chain. That is very unlikely to be an acosutic issue with the room, or the mic placement, since it is the same for all three measurements (speakers individually and also together). It also is not the calibration file: removing that does not change that region at all. It might be the measurement mic, but also not so likely. My guess is speakers, or something in the signal chain (EQ?).

One more: There's a rather large reflection arriving at your mix position from "somewhere", at around 4.4 ms after the direct sound:

That implies a specular reflection that is taking a path that is 152 cm (about 5') longer than the direct path. You can use the "string trick" to find out what surface is causing that, so you can fix it. That's a fairly serious reflection, since it is quite big (about -15dB) and quite early (4.4 milliseconds), so it deserves treating.

On that same graph you can see that the overall decay times are rather low for your size room: T20 is around 80 ms, and T30 is around 130 ms. I would shoot for something more like 160 ms for that size room, maybe even more if you can. In other words, there is too much surface area of insulation exposed directly to the room, so it needs some type of diffusive/reflective surfaces on top of it. That won't affect the low end absorption (which you need!), but it will improve the mids and highs. You just need to be careful that you don't cause any new specular reflections back to the mix position.

Those are the "big" issues that I see in your data: everything else looks quite reasonable, and your room should be pretty good, once you solve those.

Finally, reinforcing what Glenn said: Do not try to boost the frequency "dips" with EQ! That can create more problems than it solves, since dips are almost always phase related, and applying EQ where there is a phase shift, is a recipe for disaster. Only use cuts, keep the bandwidth adjusted to only deal with one issue, and don't over-do it with excessively large cuts. Just enough to "take the edge off" the peaks.

- Stuart -

#49

Thank you for the analysis Stuart, the anomalies on the speakers response is something I do have noticed but wait for the room to be in a good acoustic shape to address them. These speakers are custom made and an old design (and probably a weak link)I already had before I start this studio construction journey. Currently they are tri-amped with a dsp processor driving the amps, but being fitted in such a small and untreated room it was futile to try to “tune” them.That high mid dip could be a cross point. Now that I am almost there it is their time… this also includes eqing out that 42 hz resonance Glenn so patiently is pointing every time..
The 4,4ms reflection has to be the cavity of the window on the right of the mixing position, that cavity is already addressed (to isolate the window, mostly..), just have to take the measurements to verify..
and then we will see what to do with the short decay!
Thanks,
Dio

#50

Hi Dio,

Could you tell us which steps you took to design the limp mass absorber?

Thanks!

Eric

#51

Hi Eric and sorry for the late response.
While I was toying with the acoustic modelling simulator I noticed that if you have about 25cm of empty space, a 5kg/m2 membrane and some insulation then you can target 42-43hz. The space i had between my soffits was 23cm deep, 5kg/m2 MLV is easy to find, I had plenty of attic insulation so I thought it would be shame not to at least try...
The MLV is just stapled without any tension to the soffits and a good amount of caulc sealed the whole thing.
That's the "math" i used:

: Limp mass cal;c

One thing that needs attention:
I used the outer wall as a rigid back for the absorber and this old stone wall gets cold during winter and as a consequence the air inside the absorber gets also cold.The cold air occupies less space and kind of "sucks" the membrane inside (try to see in photo, it is not that visible).

: MLV Sucked

I guess that this changers somewhat the absorber's tuning.. perhaps I should make "pressure equalisation holes" ? Will that matter anyway?
Cheers
Dio

#52

Thank you!

Did you put the insulation against the MLV? Or is there an air gap between layers?

#53

There is a plastic mesh holding the insulation in place (shown in picture in post #44) and a very small gap so that there mlv can vibrate freely. The insulation is on the mlv side, not on the wall.
Dio

#54

From everything you have said, it seems like that should be working properly as a membrane trap. You built it right, and I'm assuming you also did the math right and got the depth correct. One issue is possibly the cold: the speed of sound varies with air temperature, so it could be slightly de-tuned, but the difference is very small, so I doubt that's much of a problem.

But the pressure change "sucking in" the membrane could definitely be an issue. Maybe drilling a very, very small hole through the frame would help equalize the pressure without affecting the device. But it needs to be VERY small. The thinnest drill-bit you have, less than 1mm if possible. And make the hole long: drill at an angle through the frame, so the hole is as long and thin as you can make it.

Have you checked if the membrane is working correctly? One simple method is to run very slow frequency sweeps across that frequency range (say, 30 Hz to 70 Hz) using REW, and gently touch the surface of the membrane with your fingertips to see if you can feel it vibrating at the tuned frequency. But here too, your touch must be very gentle, so that you don't change the tuning or over-damp the membrane. There should be a noticeable vibration at the frequency it is tuned too.

Are you planning more acoustic devices for your room?

- Stuart -

#55

also, if the MLV is stretched tight so it has a bass drum like resonance, then that is the pitch it operates at. it needs to be really loose, almost like its going to fall off :-)

also using roofer nails with the big orange disks are the best way to attach as its less likely to tear over time.

#56

Hello, I realised that I left this thread with no updates for some time, so here is one!
After making the diaphragmatic absorber (with its not so good performance…) between the speakers, it was the time to finish the plan I had (see post #15) with the angled side slot resonators (wings). The idea was to try to make this tiny room not feeling dead and unpleasant, without further compromising its already difficult acoustics.
So following J Sayers’s paradigm and my own calculations I used 2cm thick wood slats of alternating widths of 5cm, 10cm and 7,5cm and alternating gaps of 8, 12 and 5 mm respectively and cavity varying from 5 to 28 cm acting as a “ramp” from the first reflection absorbers to the end of the room.
Building the left hand wing was easy, just made the frame, bolted on the wall, sealed, put the insulation, the black cloth and the slats and that’s it!

But on the right side there is a window I wanted to somehow keep, so I built an extension in order to put a 1 cm laminated glass on the inside and make an isolated and angled window merged into the slot resonator. Not an easy job for an amateur like me… but with all the mistakes I always make I finally didi it.

One more problem was the heater radiator which I wanted to keep, but its metallic construction sings along even with spoken voice…. not much but noticeable. So I made a smaller wedge resonator, packed it on the radiator with foam between and that was the end of its days as a singer..

And then I made some measurements:

mdat:

dio Ministudio 73º both Slat Wings.mdat: (12.54 MiB) Downloaded 664 times

dio Ministudio 73º both Slat Wings.mdat: (12.54 MiB) Downloaded 664 times

Now the room is not only functional but also beautiful!
One more super nice thing with this double window is that it has very good isolation from the outer world. Will take measurements on that.
Update following soon

Cheers,
Dio

#57

Since I have sort of finished the room, it has been time to eq the speakers. I have a Void Acoustics Digidrive+ and a tri-amped setup, so a lot of flexibility.
After a couple of tries and a ton of eqing I came to this state, with is a big improvement to the previous one.

: EQ v3 Waterfall

: EQ v3 Spectrogram

: EQ v3 RT60

: EQ v3 IR

: EQ v3 FR

.... continues...

dio

#58

But this 40hz resonance was still there strong and proud, so I made the decision to revisit the MLV absorbers.

I drilled two “pressure equalisation holes” (one each) sealed them and re-sealed some parts that the caulk seemed to have cracked.
And then I took some measurements.
And after a few eq refinements finally I was happy!

The 40hz spike is now tamed for good, even without the notch eq it has been drastically damped, I guess now the mlv absorbers do what they were supposed to do…
I finally tried the “House curve” eq that REW has, and kept it, flat is too treble and results to dull and bass-heavy mixes..