Wasting my time with homemade diffusers?

Posted: Fri Apr 21, 2006 5:33 pm

Anyone else seen this:

http://www.bobgolds.com/DifuserKgveteran/home.htm

What do you all think? Certainly a hell of a lot cheaper than the RPG stuff. Of course RPG claims to use sonically engineered depths and whatnot, but is there really going to be that much of a difference? FTR, I do NOT have the money for skylines. So it's these or nothing (planning to build out absorption and bass traps with 703 as well).

Thanks in advance.

Posted: Fri Apr 21, 2006 10:18 pm

hi

i'm not an engineer but i am also looking out for DIY traps and diffusers.

RPG is made of hi density open-cell foam where as styrofoam is not in that category. It might offer some hi freq diffusion but I dun think it'll work that well. Also it's a fire harzard to have styrofoam sticking on walls!

To save costs, I suggest substituting foam with MDF (glue them together to form the correct thickness, and procced as per link); or cheap local wood that are readily sawn into strips and cutting them up and glueing it onto a think plywood or MDF backing.

Not only will you have the right density to trap and reflect the offending freq, it is also more durable considering the effort gonna be put into building.

Tools required: hack saw and wood glue. Should be cheap.

The only thing to worry is the weight, but i recommend making each piece smaller (somewhat like a piece of tile or photo frame) and you can go decorating your entire studio!

Cheers,

nehpyh

Posted: Sat Apr 22, 2006 12:48 am

ok that makes sense. thanks.

by MDF are you referring to medium density fiberboard? What about using open cell foam instead of closed cell foam for the diffusers?

Posted: Sat Apr 22, 2006 1:05 am

well I did some more research and here's what I came up with for the blue styrofoam:

Styrofoam®, the blue extruded [expanded] polystyrene CFC free rigid foam has particularly good thermal insulation, low water absorption, high compressive strength and is flame retardant

here's the description of the RPG skyline:

The Skyline® is fabricated from flame- retardant, high- density expanded polystyrene foam

So maybe they are basically the same material? And I suppose this puts the flammability argument to rest. The only hump left seems to be the mathematically calculated depth of each "tower". Any thoughts?

Posted: Sat Apr 22, 2006 1:44 am

Or you could think of using polys, which are just bend panels.
They are easier to make, giving a smooth response.

Have a look here:
http://pcfarina.eng.unipr.it/Public/Papers/150-ACTA2000.PDF

And without saying that the data you find from RPG, Dr. Peter D’Antonio and Dr. Trevor J Cox should be anything else than objective science, both are as well Prof. as having direct/indirect commercial interests in RPG. And people are people.

Therefore it can be interesting to see some studies or data from another university not allied with RPG as well.

Posted: Sat Apr 22, 2006 6:51 am

PCM wrote:

Anyone else seen this:

http://www.bobgolds.com/DifuserKgveteran/home.htm

What do you all think? Certainly a hell of a lot cheaper than the RPG stuff. Of course RPG claims to use sonically engineered depths and whatnot, but is there really going to be that much of a difference? FTR, I do NOT have the money for skylines. So it's these or nothing (planning to build out absorption and bass traps with 703 as well).

Thanks in advance.

PCM,

first, a few questions:

1. How large is your room.

2. What tests have you performed in your room?

2. What leads you to believe that you'll gain anything from a skyline?

Sincerely,

Rod

Posted: Sat Apr 22, 2006 9:21 am

Hi Rod I haven't gotten in and measured yet as I just signed the lease. Regardless, it looks to be on the small side and fairly cubish with high ceilings as well (all dimensions approx 15x12 feet, with ceilings at about 12 feet). I think most of my problems are going to be with bass, and when I calculated the room modes, it looked pretty ugly, with drastic issues in low freq, a bad bump in mids, and a bad bump in highs too.

The bass is going to be tough to tame, so thinking of building out some serious corner absorption with 703 or rockwool stacked corner traps, as well as a few panels in front of the mix position and was thinking diffusion for the wall behind the mix position. Maybe even some 703 on the ceiling similar to David Frenchs cloud absorption design.

Now obviously this is all speculation, as...

a) I don't really know that much about acoustics (enough to get by but not on your level)

b) I don't even have the exact room measurements yet and will most likely post a drawing as soon as I'm ablt to get in there.

The problem is that the apartment is a rental, and I move around a lot so I don't want to do anything heavy or permanent. I'v built out 703 panels before and they are fairly lightweight and don't require too much hassle to install. That's why the skyline diffusers seem attractive as they are cheap and lightweight.

I work in a very similar size room right now... it's about 15x17x12 and the bass/mid response is terrible, but I've learned the repsonse well enough to get pretty good mixes. I'd like to not have to spend a month or so learning my room and still guessing.

Thanks

EDIT: it has thick wall to wall carpet

Posted: Sat Apr 22, 2006 9:27 am

Room Dimensions: Length=15 ft, Width=12 ft, Height=11 ft
Room Ratio: 1 : 1.09 : 1.36
R. Walker BBC 1996:
1.1w / h < l / h < ((4.5w / h) - 4): Fail
l < 3h & w < 3h: Pass
no integer multiple within 5%: Pass
Nearest Known Ratio:
"2) L. W. Sepmeyer: 1965" 1 : 1.14 : 1.39
RT60 (ITU Control Room Recommended): 206 ms
Absorbtion to achieve RT60: 470 sabins
Volume: 1980 ft^3
Surface Area Total: 954 ft^2
Surface Area Floor: 180 ft^2
Surface Area Ceiling+Floor: 360 ft^2
Surface Area Front Wall: 132 ft^2
Surface Area Front and Rear Wall: 264 ft^2
Surface Area Left Wall: 165 ft^2
Surface Area Left and Right Wall: 330 ft^2
Surface Area 4 Walls: 594 ft^2
Surface Area 4 Walls + floor: 774 ft^2
(sabins - front wall - carpet) / Left+Right+Rear wall: 34 %
(sabins - front wall) / Left+Right+Rear wall: 73 %
Schroeder Fc: 114hz
Frequency Regions:
- No modal boost: 1hz to 37hz
- Room Modes dominate: 37hz to 114hz
- Diffraction and Diffusion dominate: 114hz to 456hz
- Specular reflections and ray accoustics prevail: 456hz to 20000hz
Count (37.6-204hz) : Axials=12, Tangentials=47, Obliques=60

Posted: Sat Apr 22, 2006 11:50 am

PCM,

I don't reccomend the use of either skyline or well type diffusors if they are within 10' of the listener (as they will be in your case if you're seated in accordance with the 38% rule).

I do not have any experience with the polys Eric is reccomending - so you may well want to read thru the thread he's linking to.

Sincerely,

Rod

Posted: Sat Apr 22, 2006 1:18 pm

Hi Rod, thanks for your reply.

Unfortunately, due to this room doubling as an office for my graphic design business, I don't think I will be able to apply the 38% rule to my mix position (desk will have to be closer to the wall).

What sort of setup would you recommend for this situation? I believe that 3 of the 4 corners are obstruction free (1 has a closet door). There is nothing on the ceiling and the floor is wall to wall carpet. Only one window and it's tiny. good solid contruction and no one beneath me.

I think in that case my mix position will be about 10' from the back wall. Could you elaborate on when you would recommend diffusion and why it is not applicable in this instance?

I assume then that absoprtion and traps would be your recommendation?

Forgive my basic knowledge, but I assumed that the rear wall diffusion would break up mid to high frequencies to diffuse troublesome modes and then the corner absorption and 703 panels would absorb the scattered and remaining low-mid-high modes.

Thanks

Posted: Sat Apr 22, 2006 5:39 pm

PCM wrote:

Forgive my basic knowledge, but I assumed that the rear wall diffusion would break up mid to high frequencies to diffuse troublesome modes and then the corner absorption and 703 panels would absorb the scattered and remaining low-mid-high modes.

Modal issues are low frequency in nature - not mid or high frequency - thus you need to understand that diffusors are not designed to eleiminate your modal issues.

The 10' rule is my own rule of thumb - especially in small rooms - closer than 10' (to me) tends to smear stereo imaging, so I will not use those in rooms I treat.

In larger rooms - utilizing an LEDE approach - so a very dead front end - diffusion on the rear wall does help to mainain a lively room feel.

you say your listening position will be 33% instead of 38% = due to your desk location, but I don't see the corrolation.

My room is a combo tracking / control room.

Thus - for layout purposes (best room use) my desk is tight to the wall - but that has nothing to do with my listening position during mixdown

I make my adjustments - and then slide back a little more than 4' to my sweet spot to listen.

In a room your size I just use absorption to handle everything.

Rod

Posted: Sun Apr 23, 2006 7:54 pm

Of those experiences you've had, Rod, (or anybody else) how well do you remember specifically what it sounded like with Skyline-style designs? I can easily imagine the problem with 1D diffusors like the common QRD 734, but at the last AES in NYC, I asked Peter D'Antonio what he recommended if you'd like the benefits of diffusion but just don't have 10 feet behind you. He suggested using 2D diffusion since that would do substantially more to break up the mid and HF waves and scatter more of the energy away from the direct path back to the listener, thereby gaining you several extra feet as the waves bounce off additional room boundaries. So I'm very curious as to what it would actually sound like with less than 10 feet between the back of your head and the rear wall with Skylines back there. Anybody?

Lee

Posted: Tue May 02, 2006 11:42 am

rod and everyone thanks for your help. The foam is so cheap I'm going to give it a try as i have a friend that said he'll buy them off me if they don't work out. Placing my order for 703 today so I'll report on the success or failure of the diffusers in a month or so when I'm set up in my new place.

Posted: Mon Aug 13, 2007 7:35 am

eric_desart wrote:

Or you could think of using polys, which are just bend panels.
They are easier to make, giving a smooth response.

Have a look here:
http://pcfarina.eng.unipr.it/Public/Papers/150-ACTA2000.PDF

Hello Eric,

Very old message I now.

I looked at M. Farina's paper & I have a few questions about it :

He's comparing panel with different sizes : flat panel 0.51m2, Schroeder 0.51m2, curved panel 1.8 m2 (page 5)

He doesn't tell how curved and how thick the curved panel is

From the tests, the curved panel seems to have the best scattering properties

the impulse response (Fig. 12 on page 19) seems not as interesting as the Schroeder panel

Should we understand curved panel are more scattering than diffusive devices ?

Posted: Tue Aug 14, 2007 7:10 pm

msikio wrote:

eric_desart wrote:

Or you could think of using polys, which are just bend panels.
They are easier to make, giving a smooth response.
Have a look here:
http://pcfarina.eng.unipr.it/Public/Papers/150-ACTA2000.PDF

1) He's comparing panel with different sizes : flat panel 0.51m2, Schroeder 0.51m2, curved panel 1.8 m2 (page 5)

2) He doesn't tell how curved and how thick the curved panel is.

3) From the tests, the curved panel seems to have the best scattering properties.

4) the impulse response (Fig. 12 on page 19) seems not as interesting as the Schroeder panel.

5) Should we understand curved panel are more scattering than diffusive devices ?

msikio
I accidentally saw this.

Just gambling a bit:

1) Since such bend panels also have edge diffraction/diffusion I have the feeling he made the panel longer to limit this edge effect on the top ends.
Note that the way this is measured (study the extensive document), does not really relate to the surface area of the sample (bit simplified). You send sound to the sample and measures the reflected energy on discrete points. Having a larger panel does not alter the principle.
Without knowing for sure since not documented, I have the feeling that he wanted to isolate the effect of bending from other phenomena (edge absorption and diffraction)

2) I see he refers to plywood without telling anything further. He calls it hemicylindric, at one spot in the PDF, which could refer to 1/2 circle, but at most other spots he refers to a curved panel.
Hence it's not clear for me as well.

3) That it gives the smoothest is to be expected. But I referred to the document for what it is. Anyhow Angelo Farina is a top acoustician and his Univ is specialized. I rate his findings equal to P. D'Antonio's.
And he investigates something in function of comparing different methods to rate diffusion/scattering. He doesn't seem to have whatever commercial relationship to whichever type.
Whether this is best or not I leave open. What I do know is that I have a lot of experience with polys, always with succes (also using the edge diffraction by hanging them free), and that as far as you go back in history labs are executed with polys, including the own reverberation room of RPG (has additional reasons) itself.

4) I find this a wrong interpretation of these graphs.
It's bad that the axises are hardly readable but the Y axis gives a level (amplitude of some kind), and the x-axis a time scale (of some kind).
Shown is that the flat panel gives a very high level (look at the Y-axis) and the bend panel a MUCH lower level (also confirmed in text).

What do you expect from a bend diffuser?
Theoretically/simplified: sound arrives, is spread open, hence the energy distributes itself over a larger hemicylindrical surface, thereby diminishing the returned/reflected energy on any discrete spot with an inverse ratio.
Well that's exactly what that graph shows for the bend panel.

The fact that this is a short pulse is not related to lack of scattering or polar distribution of energy.
One measures the returned energy (on that graph) from the surface at ONE single point in the normal, which after being diffused has a significant lower level (compared to flat panel). The rest of the energy is distributed in other directions.
The Schroeder shows a lobbed time delay at the same point, but in total a seemingly higher total energy at that same point (confirmed by lower diffusion number). I assume this is frequency related and has to do with that lobbed pattern (and phase shift).
Hence whether that's better or not, is not so clear to me. It shows some coloration. There is too little data to make final conclusion about these graphs and these scales are hardly readable.
But clear is this is a time delay effect for that Schroeder diffuser, which means that another sound is returned/reflected than what you sent to the diffuser. And since the total reflected energy seems to equal or exceed the bend panel curve, I'm not sure how to link the notions 'better' or 'worse' to this comparison.
Also Farina doesn't do this.
It should be interesting to see a spectral analysis of this graph as well.

5) I don't do explicit statements about this. RPGs have proven their value over and over. I think there is a rightful place for both.
What I did notice in other groups if people understandably ask for a comparison between both (one looks and is simple to make, the other looks learned but is difficult to make), that some "Experts" explicit reply that the Shroeder type wins with glance. The problem is that I know a couple of them, and that they mainly base this on the fact that RPGs look special and more pro-like, not on really knowing or understanding.

That's what I 100% respect in Rod's reply. He simply states that he doesn't have experience with polys, hence that the questioner, if interested can look in that document as well. That's a scientific attitude.

Warm regards
Eric

This topic sponsored by: