OK, so here's this idea I have, and I wonder if you people think it will work.
The idea is based upon the oh-so-familiar slat resonator. But on the slats you mount long pieces of wood that have been cut length wise to a triangular (or quarter-circle) shape, as seen from the short end. This will make the resonators front diffusive AND it will make the slots much deeper, which should lower the resonant frequency of the resonator.
Damn, it's difficult to explain these things in a foreign language - I hope you get the idea.
Here's a pic, seen from above (assuming the slats are mounted vertically). The light brown area is the mineral wool behind the slats, the dark brown squares are the slats, and the triangles are the diffusive pieces of wood mounted to the slats.
Hey - like all Swedes your English is excellent - wish I had a second language :):)
I'm not sure what effect the angled fronts would have on the action of the resonators themselves as the slat width is part of the formula.
Looks like a question for Barefoot again :):)
cheers
john
Yeah I understand that Sen - but I think he meant the whole thing to act as a diffusor as well as a slot resonator - correct??
Look it's going to act as both but to what degree I don't know. :)
cheers
john
Right, the idea for the added pieces of wood was that they'd serve as diffusors but also lower the resonant frequency of the resonator. According to the Excel file on the SAE site, this frequency is heavily affected by the depth of the slots. I guess you could also mount standard 2x4's, angled 90 degrees to the the front slats.
I guess nobody will know until we build one. But it seemed like an OK idea.
Cheers
/Henrik
Henrik - in Guruland http://johnlsayers.com/Studio/index.htm I designed slots similar to what you are suggesting. I used 6" x 2" with the 2" being the front face thus giving a 6" slot depth. They are on the side walls of the drum booth to give high end diffusion and low-mid absorption.
I've attached a drawing
cheers
John
Right, that's precisely the same principle I'm suggesting. So do you think the deeper slots do lower the resonant frequency of the Helmholz resonator?
Cheers
/Henrik
Henrik, the way you've drawn your idea I would think that you'd have the equivalent of several concave "focused" reflectors, which could really raise havoc if aimed back to the mix position in a control room, or at a mic in a booth - I'd be more inclined to do what John did, with the 2x lumber on edge for deeper slots. The variable depths should act somewhat like a QRD as John mentioned... Steve
Henrik, the way you've drawn your idea I would think that you'd have the equivalent of several concave "focused" reflectors, which could really raise havoc if aimed back to the mix position in a control room, or at a mic in a booth - I'd be more inclined to do what John did, with the 2x lumber on edge for deeper slots. The variable depths should act somewhat like a QRD as John mentioned... Steve
Knightfly, thanks for this. I see what you mean by the "focused" reflectors the way I did my drawing. I suppose John's idea is better. Do you think the angles would work OK if you had wider slats, I mean like double the width compared to what I drew? Either way, it's absolutely nothing I'd place behind the monitoring position in a control room.
What is a QRD?
And do the deeper slots lower the resonant frequency according to the Excel file on the SAE site (The Helmholz calculator)? If so, John's approach would be a very efficient way of getting the resonator to absorb really low frequencies.
Cheers
/Henrik
Sorry, a QRD stands for Quadratic Residue Diffusor - looks sorta like John's pic, except the wells are sealed (it's ONLY a diffusor) and each well's depth is calculated from a prime number sequence - the well width is usually about half the wavelength of the highest frequency (shallowest) well. You've seen them in nearly every picture of every pro studio you've ever looked at -
Here's a fairly good description, but not very understandable pix -
http://www.suite101.com/mypage.cfm/unpluggedsound/8336
Here's a little better picture
http://www.acousticdiffusers.co.uk/diffusers/qrd734.htm
Diffusion is primarily only in a plane that is perpendicular to the wells - if the wells run vertically, diffusion is horizontal. That's why when you see these, usually there's a bank of them with some running horizontal but most running vertical... Steve
Sorry, a QRD stands for Quadratic Residue Diffusor - looks sorta like John's pic, except the wells are sealed (it's ONLY a diffusor) and each well's depth is calculated from a prime number sequence - the well width is usually about half the wavelength of the highest frequency (shallowest) well. You've seen them in nearly every picture of every pro studio you've ever looked at -
Here's a fairly good description, but not very understandable pix -
http://www.suite101.com/mypage.cfm/unpluggedsound/8336
Here's a little better picture
http://www.acousticdiffusers.co.uk/diffusers/qrd734.htm
Diffusion is primarily only in a plane that is perpendicular to the wells - if the wells run vertically, diffusion is horizontal. That's why when you see these, usually there's a bank of them with some running horizontal but most running vertical.
No matter how wide the elements in your slot idea, they would still be concave, so no. That's why when you see poly-cylindrical diffusors, they are bowed OUT into the room, not the other way 'round. You're looking for DIFFUSION, not CONCENTRATION -
Yes, deeper slots lowers the frequency... Steve
No matter how wide the elements in your slot idea, they would still be concave, so no. That's why when you see poly-cylindrical diffusors, they are bowed OUT into the room, not the other way 'round. You're looking for DIFFUSION, not CONCENTRATION -
I understand that, but I suppose even polycylindrical diffusors bowed out into the room should probably be placed at a certain distance from each other, right? I mean, if you place them side by side, you will get a certain amout of concentration in the area between each pair of diffusors.
If you look at my drawing, the wood parts attached to the slats might as well have been quarter-circles looking from the short end, instead of triangles. A pair of these quarter-circles would make a polycylindrical diffusor (with a slot in the middle).
See what I mean?
Cheers
/Henrik
Henrik,
Your idea will definitely work with respect to lowering the resonance frequency. If the wavelength of your target frequency has a similar size as the triangles (which it almost certainly will not) then the angled slat fronts will cause a very minor tuning deviation - but this really isn't a concern at all.
My only concern is that this design will probably not be a very good diffusor. The even periodicity and depth may actually cause reflection problems at wavelengths having integer relationships to the structure.
John's Guruland design looks outstanding. It would serve as both a diffusor and a true broadband resonator. As I've mentioned in other threads I don't think that varying the slat width, slot width, or cavity depth results in a true broadband absorber. Varying the slot depth, on the other hand, does.
John,
It's not clear from your drawing if there is a formal acoustic principle behind your choice of slat depths (thichnesses). If not, your design could be taken even further by making the slat depths correspond to those of a Primitive Root Diffusor (sort of a QDR without the thin partitions). The adjacent slats don't need to be the same depth for the slots of the resonator to work properly. Great design!
Thomas
Yeah - that's what I was sortta aiming at Thomas. I was trying to kill two birds with one stone as they say. It's definitely up for further development. I also put one of these on the rear wall of the control room.
cheers
john
Thats a pretty cool idea. I calculated the sequence in excel using 17 as the prime root. The sequence is 16 long before it repeats.
2x6 lumber is 5-5/8" - so this would be the sequence in mm (124 = no cut)
0, 8, 31, 70, 124, 62, 16, 116, 101, 101, 116, 16, 62, 124, 70, 31, 8
John - can you add .xls as a permitted upload file extension?
Below I've attached a generic design for a 10 slat primitive root sequence (prime root = 11). The depth D of each slat in the sequence is simply calculated by multiplying the unit width W by the sequence multiplier m then adding in a minimum depth d. The slot width s should be relatively small compared to the unit width. A slot width of a 10th to an 8th of W is probably a good choice. No need to vary s since the varying slot depths already make this a broadband absorber! :D
Thomas
Neat :)
You are right I did use a quadratic residue sequence - residue of n squared divided by the prime root. Then normalized the series so the max = 5 5/8" (to use 2x6 lumber).
Your diagram is fantastic. Using 2x6 lumber - the high point would be 40 inches - would the design still function by normalizing the series so 10 = 5-5/8" but keep the width at 2" ?
You & John should write a book - you'd make a fortune with both of your skills combined.
Andrew,
Notice I updated my post. Actually, I was the one who was mistaken. I thought primitive root diffusors were the only ones that didn't require partitions. However, I did some further reading and found that ideally primitive root diffusors should also use partitions. Removing the partitions is just a compromise and can be done with either quadratic residue or primitive root diffusors. The diffusive characteristics are slightly degraded, but they are much easier to build without the partitions.
Still, I guess I'll stay with primitive root sequences for the time being. I find it's better to stick with as many known quantities as possible when developing a new design. PRDs are the only ones I actually know I've seen without partitions. Maybe they handle it better...?
Anyhow, my bad. :oops:
Thanks! :D
Thomas
Hey people,
those quadratic residue Helmholz resonators look way cool. This was of course the effect I had in mind, only much better achieved than my idea.
Love this forum. Thanks guys.
Cheers
/Henrik
And, um, I'm not REALLY a math dummy - I understand math, but I've never really been taught anything that you'd call the least bit advanced. So can anyone explain a little more at length what operations you actually do to get that quadratic residue secuence out of a figure?
Cheers
/Henrik
Henrick,
I think this is the easiest way to describe it.
First of all we pick a prime number. The bigger the prime number the longer the sequence, the greater the aspect ratio of D/W, and the wider the bandwidth of the diffusor. I think 11 is a good compromise.
p = 11
For a quadratic residue:
n = 0 to p-1 = 0,1,2,3,4,5,6,7,8,9,10
and the sequence is given by:
QR = n^2 OR the remainder of (n^2)/p if n^2 > p
so
n^2 = 0,1,4,9,16,25,36....
We can only use the first four values of n^2 because they are the only ones less than 11. The following values, however, are greater than 11. So we have to divide them by 11 and find the remainder.
16/11 = 1 remainder 5, 25/11 = 2 remainder 3, 36/11 = 3 remainder 3, .....
Therefore:
QR = 0,1,4,9,5,3,3,5,9,4,1,0
Similarly for a primitive root:
n = 1 to p-1 = 1,2,3,4,5,6,7,8,9,10
and the sequence is given by:
PR = 2^n OR the remainder of (n^2)/p if 2^n > pPR = 2,4,8,5,10,9,7,3,6,1
Now remember, QR and PR are the multipliers to determine the well depths of a partitioned diffusor. With this non-partitioned DSR design we're interested in the slat depth not the well depth in front of it.
So our multiplier sequence is given by
m = QR(max) - QR
or
m = PR(max) - PR
Make sense? :)
Thomas
Thomas
one question on the calc in the diagram. Is it important to tie W to m in calculating D? If the aim is to use say 2x6 lumber. W =approx 2, so the largest D would be 2x9 = 18 - which would be impractical.
With the W*m +d I don't follow the minimum value being d - isn't that d=0?
How about
D = (m/9) * d
where d = the max length of D allowed (limited by lumber size selected)
It could well be that I am missing something.
Hello Thomas. Wow, this is the first time I've visited your forum. Cool. I'm sorry about not getting back to you about Stylmark. I've been offline for three months while looking for a new home. I'm now in Oregon. Just wanted to say this thread is toooooooo good.
I will try to use this to solve the extrusion problem. I'll post my plan section soon and see if it is any way near what your saying here. First I have to unpack the Stylmark stuff and detail the extrusions. That will take a while. In the meanwhile, I will check the whole site out. Gotta go say hi to
John and Steve. :D
First though, I want to reply to this:
one question on the calc in the diagram. Is it important to tie W to m in calculating D? If the aim is to use say 2x6 lumber. W
=approx 2, so the largest D would be 2x9 = 18 - which would be impractical.
What would be impractical? Just cut it out of 2 layeres of 3/4" laminated ply or MDF, if I'm understanding what your saying correctly. And if I am, then using solid lumber for diffusers is exactly the problem I'm solving with Stylmark. The width of wells and partitions is adjustable, if the extrusions are detailed correctly. And they are hollow. Plus, they look fantastic. I'll post a pic of my prototype as soon as I get a one built. That is onece it is approved here. Ha! Also, using laminated ply, you can use different thickness, AND prefinish each piece as required prior to assembly,with veneers, laminates, paint, or better yet, CHEMETALS. That is what I will be doing in conjunction with the Stylmark.
fitZ :P
Hey Fitz, good to see you here!:)
I'll check to make sure I have the design methodology correct. However, tying the depth to the unit width just makes intuitive sense to me, considering the diffusion is two-dimensional. How can a two-dimensional effect be independent of one of the dimensions?
Thomas
How can a two-dimensional effect be independent of one of the dimensions?
I haven't got a clue thomas. :shock::P No, really, I don't understand if that is a question or statement. You kind of lost me. Maybe it's just me. Thick skull syndrome.
fitZ
PS, Thank you for the welcome Thomas.