C.R. building. Can you help to analyze these files?

[Glenn Kuras]

Kind of hard to read but if you can get +/-7 db or so then your doing pretty well. Also frequency response is great to look at but decay time is what you should focus on fixing.

[gpiccolini]

thanks. I get an error and the FM file didn't make it to my previous post, so here it is.
Regarding fixing decay time: How can I achieve it?. Absorber, resonant membrane (steel), limp membrane?. I have a closed window hole almost in a corner (back, left) which is 120x80x23cm and it's almost a limp mass... I just need to put the limp mass, but the most difficult part is to calculate the limp mass surface to achieve the required frequency.

[syncussion]

Hi gpiccolini,

For reasonably effecient bass absorption your 60kg/m3 may be too dense.
It seems to me such a density rockwool has a gass flow resistance of at least 20.000 (and possibly quite a bit more).
If you put this into the porous absorption calculator: Porous Absorber Calculator you'll find that it mostly reflects bass in the areas you have problems.
Furthermore remember that bass is not directional like mids and highs, especially below 100hz. So you may well get destructive reflections in the bass from places you're not expecting / where there is no direct reflexion path for mids/highs.

A possible solution could be much less dense and much thicker rockwool / fibreglass over large areas.

[gpiccolini]

Ok, I'll think about "fixing" the ceiling, but the Porous Absorber Calculator gives pretty good LF absorption numbers, so I'm not sure what are You pointing at exactly. I have a 4x4m ceiling with 10cm Rockwool with veil and between 80cm and 100cm of air gap (the roof is sloped. solid bricks). Maybe add a double layer of lighter wool with paper on both sides as first layer...
I'm thinking to put an EPA on the middle of the front wall, just below the ceiling. About 2x1,20m. Not sure if it'll work in that position (G.E., please), and also have a space for a limp mass bass absorber near a corner. I'm not sure if these devices can help to shave my LF reverb time.

[syncussion]

Ah sorry you're right. I just checked and thought that 60kg/m3 rockwool at 10cm did worse in the bass.
Still.. it may not do enough, and with less dense material, for instance 10kg/m3 fibreglass with GFR 5000 or even less you can get about twice as much absorption in the bass (with more thickness).

Just throwing around thought here, but perhaps your speakers are somewhat high and there's a floor reflection cancellation as well?

[DanDan]

Somethings not right in Fuzz I reckon. I see a Magnitude of 1.2 or 'SPL' of 120dB.
Normally, averaging the FR of two speakers or mic locations is easy. However Decay/Waterfall stuff needs to be Time Domain Combined. For that to work you need to use Loopback Correction when recording the measuring sweeps.

It would be simpler to use single speaker measurements. If you want the best visibility of such low modes, place the speaker on the floor in the front corner.
The mic can be at the opposite back corner if you wish but listening position is fine too, just keep away from the room centre where these primary modes null. I include Height and Width in that recommendation. Go off centre.

Given that your peak levels in measurement are 20-30dB above the usual, it is possible you are seeing far too much of the modes also. You may have little or no problem to solve.

In any case, try playing a Sine Wave at exactly the lowest modes frequencies. The Generator in REW is excellent at this, as is Signalsuite.
Sweep slowly and carefully until the whole room starts to rattle and roll. Full Resonance. Listen, and measure levels if you have an SLM at the possible spots for Bass Traps. Trap where it is loudest. Don't forget to check up near the ceiling and at the floor. Soffit Traps are big and efficient.

The VPR design would be my recommendation. 2mm steel, both back corners 1Metre wide floor to ceiling.

Eq would very likely do a reasonable job also. The don't call it Room Eq Wizard for nothing!

DD

[gpiccolini]

sorry, it seems I normalized the graph and posted it normalized by mistake. I just took a look at the original file (which contains some other measures at different distances and was too big to post here). The original peaks around -3db. I can post it again if that's necessary.
It's quite difficult to put VPRs in the back corners... almost not possible due necessary size. I can put a 1,80x1,00m in the front wall, or possibly a couple of smaller ones (1,30x1,00) in the front corners, also at ceiling height and always wider than taller. Also I'm concerned with weight and was under the impression that 1mm works lower. Anyway: Can it work? I never saw EPA/VPR waterfalls in that position. Possibly for a reason
That said, I still plan to cover the front wall with wool. Trying to decide if regular glass wool with paper (sandwiched, paper on both sides) and 5cm air gap can work better than the same Rockwool I have in the ceiling plus air gap (5cm). ATM the front wall is still solid brick and the window.

[DanDan]

The best info we have on VPR's shows them working well in corners. All LF traps are about 3 times as effective in corners. Also, I have not seen evidence of 40Hz action. To have any chance I reckon 2 or 2.5mm steel, as big as possible, in corners, is essential. If you can't do that, forget VPR.

I can't see how a normalised graph would show levels 20% above Full Scale. It should Normalise to FS.

I would stick with single speaker measures and throw away any not credible measured results. Until I see a measure that looks valid I can only continue to guess that the level of your 40Hz and nearby stuff is very low.

It would be easy and cheap to try Eq.

What is your reasoning in covering the front wall? How deep? Why the paper?

DD

[gpiccolini]

I'm trying to achieve a sort of LEDE room, Hence the absorbent front wall (and front half of side walls, half done ATM). I have those Rockwool 60kg/m3 2" panels and have access to glass wool rolls, which I don't like but are lighter, in two forms: with or without paper. I was under the impression that the paper helps to retain some HF and absorbs more LF, and I can sandwich it to keep the wool trapped inside.
I will take a few pictures and post later
This is the original file without normalizing. I have several others at different distances and in the back corner.

[gpiccolini]

some pictures.

[DanDan]

Looks lovely. Good work. I don't think you have the space to get serious about 40Hz. With advice from experts on Helmholtz you might be able to do a shallow tuned effort, but beyond my experience and I have seen too many failures in DIY Helm. Similarly not enough evidence to suggest smaller VPR flat on the wall might work, I don't think they will. MLV is a popular talking point at the moment but there is word of mouth that they work.
What I do know, from my own experience, is that Eq will improve your situation enough to make the small problem you have become totally tolerable.
DD

[gpiccolini]

I think You're right about 40Hz. Front right corner will get a Superchunk with a 1m rack in the middle (still have absorption in walls and door); maybe that will ease things a little more. What really bothers me is the 80hz null at listening point. Maybe the remaining absorption will help... Not sure if should I go with the light glass wool with FSK, the 60kg/m3 Rockwool panels or a combination of both. All in the front wall, of course.
I'm really thankful for your help and this great forum.

[DanDan]

De nada. 60-80 Hz nulls are common at the listening point.
Your ears are typically at half width and half height. Double problem.
Sometimes you can move the woofer to half height where it will not drive this height mode. Even if it means the speaker is upside down.
VPR's, Helmholtz, Slat Resonators, should all work fine at the side reflection points. Two to four 1mm large VPR's with IsoBond on BOTH sides.
Any of these side traps and even the cloud can be angled to divert any reflection this is not fully absorbed. I think it is Gullfo who uses Hard angled clouds, in an attempt to break up the H modes. Heavy though.

DD

[gpiccolini]

Hello again.
It's getting better but I still have some swings I'd like to work between 70/210hz. I think a device tuned around 120hz should ease things, but I'd like to hear opinions about which type and how to locate the places to put them.
Work done since previous measurement includes a big super chunk in the R front corner, speakers mounted a little higher and closer to the wall (almost touching) with Rockwool behind them and a new listening point @ 1,40m from front wall (old one was @ 1,70)
These measurements were done firing both speakers.

[gpiccolini]

Hello,
I changed my listening point 10cm closer to the front wall and it sounds better to me. Here are the new analysis. I'd like to have less swings in the 60-200hz range; maybe a device centered around 120hz would help to that and to reduce the long decay in that range. There are lots of 120hz close to the 5th wall, the 45 degrees one that cuts the front left corner.
Another thing that came to my mind is that I have a big air gap between the 10cm Rockwool ceiling and the sloped roof, between 40/80cm aprox. Maybe some devices can be located there to work with the 35/40hz fundamental which can be the root of all.
Both are beyond my knowledge... :D
any help?

[syncussion]

Are your speakers isolated and on (sand filled) stands or on your desk?
Anything in contact with your speakers that can vibrate / has a cavity?

[gpiccolini]

Originally Posted by syncussion

Are your speakers isolated and on (sand filled) stands or on your desk?
Anything in contact with your speakers that can vibrate / has a cavity?

Hello,
Speakers are on shelfs attached to the wall (solid brick). The shelfs are not a sheet of wood but built with small pieces like a web. All the contact points have a dense rubber and the speakers themselves lay over thick rubber pieces.
something that can vibrate is the acoustic window in between the speakers. It still needs to be adjusted.
What did you see to ask this?

[DanDan]

Unless you are doing a new build or extremely extensive treatment, your graphs will never look 'pretty'. Even at the highest end, we rarely see response graphs for world class facilities. So focus on the improvements as the treatment progresses. Obviously you have a 100Hz resonance. IMO the best way to identify the cause and the solution is to play a 100Hz Sine wave in the room. Tune carefully for maximum resonance. DAW plug-ins usually jumping in steps so you may have to type the frequency in. For smooth tuning, SignalSuite is good, as is the Generator in REW. Tune slowly and carefully. Move throughout the room listening and perhaps observing level on a SLM or mic with a level meter attached. You will probably find areas of very high SPL. Corners are likely but it can be a whole wall or ceiling, or floor. A surface. Obviously LF treatment is most useful at these hot spots.

DD

[syncussion]

Originally Posted by gpiccolini

Hello,
Speakers are on shelfs attached to the wall (solid brick). The shelfs are not a sheet of wood but built with small pieces like a web. All the contact points have a dense rubber and the speakers themselves lay over thick rubber pieces.
something that can vibrate is the acoustic window in between the speakers. It still needs to be adjusted.
What did you see to ask this?

Just to make sure that the 100 hertz thing isn't something vibrating, it looks so sharp / specific frequency. (and isolating the speakers is good overall anyhow)
But it could very well be the window indeed!
Maybe measure again while you lean against the window dampening it with your hands? Then you'll probably know.

[syncussion]

Oh and by looking at your pics you may be able to get another improvement (not related to the 100hertz thing) by attaching some thick absorption to the back of your desk and making it stick out above your desk so you can still see/hear the speakers directly but the sound has no direct path to bounce off your desk back to your ears. (or is your desk now too close to the speakers to do this?)
This is what I'm doing, matters quite a bit.