Acoustics‎ > ‎

QRD-1

Newport's QRD Construction

QRDs can be purchased or built in a variety of sizes and, like traps, there is no single solution other than understanding that the impact on room acoustics is a function of surface area. A 2-ft x 2-ft panel can have some localized effect but it will have little impact on a room, even a very small studio like this one, simply because it effects a very small percentage of total room surface area. After treating walls, corners and ceiling with traps, the wall behind the primary listening position provided about 20 sq-ft of treatable surface area. This formed the upper bound on QRD panel size.

Dividing this available wall space in half allowed two (2) QRD diffusion panels to be constructed from 1/8-in Lauan hardwood ply and 1-in thick EPS foam sheets using the design outlined in the previous Diffusion page. These two panels provide nearly 20 sq-ft of modulated wall space diffusion, weigh <30 lbs each with material costs at about $75 per panel. These panels provide an extremely light weight, custom diffusion as a very inexpensive acoustic treatment -- that is, IF you discount the labor as just part of the design/build/music fun. However, have no illusions. Crafting these diffusers took significant design, planning and construction time.


Precision

The web is over-loaded with claims about QRD construction requirements -- many of them downright silly. Here are some observations to put those claims in perspective.
  • Saws with laser gauges - Nice but unnecessary. Just use a good table saw with a good blade and think ahead.
  • Well depth within 1/32-in - That will work, but is unnecessary. Even at an upper frequency cutoff of 5 kHz, 1/8-wavelength is still nearly 1/2-inch and about the same at 1/32-wavelength at 1 kHz. Build carefully but don't sweat the precision.
  • Well bottoms must be totally sealed - Yes, this works, but why the worry? The issue arises from having a gap between the well bottom and the side -- but ONLY if there is a hollow cavity below the well bottom to form a Helmholtz Resonator. Such resonators are tuned (by volume and dimensions) and will re-radiate energy back at that single frequency. In the design shown here, that cavity is filled with a solid foam block, glued in place. Any tiny gap cannot form a resonator and, if it did, that frequency would be way above the design range of the QRD design. Make the wells tight as reasonable and coat the wells and cracks with polyurethane but don't fret about cracks.
  • Well widths must be very precise - Yes, a good thing, but what matters is the uniformity of the well positions -- not precise uniformity of the well widths. Again, Lauan warps which makes deeper wells more difficult but small variations in well widths is not an issue.
  • Well bottoms must be super flat - No, and in fact one study showed that adding a slight slope to the well bottom actually improved the diffusion. Sloped well bottoms are not used primarilly because they are so complicated to build with uniformity.
  • Surfaces must be smooth - Again, this is fine, but just how smooth is smooth? The only references, with measurements, come from Cox and D'Antonio where the comparison is between 'smooth' materials, like wood, and 'rough' materials like concrete blocks. Don't worry about the Lauan surface being 'rough' -- it's not; it is quite 'smooth' enough.
  • Must use heavy MDS materials for rigidity - Sure, these will work, but they are not necessary. MDS is very heavy with a hard, smooth surface which does work. However, hanging a 5-section panel like the design here, made of MDS, would be extremely difficult both to build and to mount. All that is needed is a rigid material -- one that will not respond to the room sounds and flex in sync. Cardboard (regular), flexible plastic, etc are not rigid enough but high-density cardboard can work -- if you can get the material at a reasonable price. Lauan ply is the cheapest, lightest, most rigid and most readily available material. It is rigid enough but it does have a bit of warp but construction technique can compensate for much of the warp.
  • QRD's are based on mathematics which requires mm-grade precision - No; though QRDs are based on the math of phase-gratings and the original designs tested by Schroeder were measured at microwave frequencies. However, this is acoustics and the effects are subjective as the end result is human perception, likes and dislikes. Yes, the GHz theory holds true at kHz frequencies but the mechanical precision scales as well. Sure, precision QRDs are the goal, but the cost of that last 10% is generally too high and you'll never hear the difference.
Remember Occam's Razor - "all things being equal, the simpler is more likely than the complex". For room acoustics, this means that it is significantly more likely that simple things like musicians, instruments, chairs, windows, doors, gear etc. will remain the biggest acoustic effect rather than complexities of QRD fin spacing, well-depth variation, small cracks between surfaces, etc. Diffusers have limited bandwidth, limited working angle and provide more scattering than diffusion as the sound source drops below the design frequency. Build the diffuser well; just don't sweat the tiny stuff.

Working with EPS Foam
  • Cutting EPS foam, even with plastic and foil backing, is messy -- very, very messy.
  • Do ALL the cutting of EPS in one single session -- then clean up once and you're done.
  • Cutting anything less than 4-in width EPS immediately gets difficult and requires a jig to hold the material
  • While EPS cuts easily, it also "bucks" easily with a table saw with considerable force -- extreme caution is advised!
  • Gluing to the plastic or foil backing of EPS works, but ultimately the glue can separate from the backing
  • PL300 is a bit stronger bond to the plastic, but then the plastic can de-laminate
  • White Glue, Carpenter's Glue (yellow) or PL300 tube adhesive bond equally well to EPS and do not dissolve the foam
  • All glues still require clamping during the cure, but Carpenter's Glue is the best, easiest, fastest, inexpensive solution.
Working with Lauan Ply

Lauan ply is warped when you buy it, as you work with it and so just adapt. Lauan hardwood ply is sold in 1/8-in thick 4x8 sheets so it's rather "floppy" to handle and cut material is generally warped even though it is stacked flat at the hardware store. However, the surface is hard, it looks reasonably nice, it's laminated so has sufficient rigidity and it's only 12-lbs per sheet.

This QRD design is constructed almost entirely from Lauan ply to keep the weight down but Lauan is more difficult to work with during assembly than stiffer, heavier and solid materials. The key points to remember are 1) the practical precision requirements as listed above and 2) one must constantly be aware and compensate for material warping. However, by paying attention, and careful gluing, this structure will be sufficiently square and true with minimal defects -- at least any defects which have any significant affect on acoustic performance.

Some suggestions:
  • Cut the Lauan ply 'with the grain' so that the grain is lengthwise with the fin
  • Cut ALL pieces of a given width using a single setup -- it's obvious, but worth remembering
  • With a little effort, the well-bottoms can be cut very precisely and repeatably with a table saw
  • Lauan hardwood ply is only S1S (Good Surface-One-Side) and the back-side is often much darker
  • One option is a) stain the lighter side, b) coat both sides with polyurethane and c) alternate the Fin sides to 'average' the color
  • A significant percentage of the Lauan fins will not be flat, so select the better ones for the deeper wells which have less support
  • QRD assembly is a bit easier if the final cut pieces are hand-graded for flatness. Use the less-warped material when gluing to shorter foam pieces (greater well depths).
  • Sand ALL edges - caution, fresh-cut Lauan ply, even very clean cuts, are full of splinters
  • Sand both surfaces on the Fins -- sanding the back side will also lighten the darker color for a better match to a stained front
  • Lauan ply is warped -- this cannot be repeated enough so just work with it.

 
 
EPS Foam Glue Test #1
White Plastic Side (White, Carpenter's, PL300)
Foil Side (White, Carpenter's, PL300)
EPS Foam Glue Test #1
  Plastic removed (White, Carpenter's, PL300)
Foil Removed (White, Carpenter's, PL300)
 
 
Table Saw Mitre Guage Jig
(3/16-in strip to hold foam down when cut)
Cutting EPS is VERY messy
( a shop-vac or old vacuum cleaner helps )
 
 
Stack of cut EPS foam QRD well-spacers
Two stacks of cut QRD Fin blanks
(the paint cans didn't help with the warping)
 
 
Table Saw Jig
(to cut the well-bottoms)
Carpenter's Glue Set Time
(LL->LR:CW - 30, 60, 90, 120 min)