For the past few years, many have become used to having virtual meetings in their homes. Spaces like kitchen tables, couches, spare bedrooms, and hammocks in the yard have all become “offices”. As you can imagine, many of these spaces aren’t well known for their acoustic qualities. [Zac] built a sound diffusion art piece out of scrap pieces of wood to help his office sound better when recording.
Reverb is caused by sound bouncing off hard, flat surfaces like drywall. These reflections are picked up by the microphone and lead to a noticeable drop in perceived sound quality. There are generally two ways to kill reverb in a space: diffusion and absorption. Diffusion is the technique that [Zac] is going for, with thousands of faces at different angles and locations, it breaks up the harsh reflections into millions of tiny reflections. Absorption is usually accomplished with foam and other typically soft substances.
[Zac] happened to have a large pile of offcuts and extra material from past projects of various wood species, making it easy to make a visually interesting piece. He used a table saw to rip them to a consistent width and a drum sander reduced them all to the same depth. Next, the long sticks were cut with a miter saw into 5 different lengths, leaving him with thousands of little pieces of wood. The hard part began when he had to glue several thousand pieces to a plywood backer board with CA glue. Sanding, finishing with poly, and a french cleat made the three pieces ready to hang on the wall.
Overall, the effect is stunning. While we’d love more hard data on the improvement, it certainly does sound better anecdotally. If you’re interested in more woodworking, take a look into making an inlay without a CNC. Video after the break.
CA??? who uses CA with wood?
Anyone that builds model boats or planes for sure
Woodworkers. Very common for smaller parts, tacking bits together (while wood glue sets), or as a glue for parts which aren’t loadbearing or where you don’t want to / can’t easily clamp them.
In this case, 100s of small bits of wood, not loadbearing, and difficult and very slow to clamp.
“There are generally two ways to kill reverb in a space: diffusion and absorption.”
My understanding is that while absorption is definitely used to “kill” reverb, diffusion is more about eliminating undesired resonances and standing waves, so the reverb sounds nicer (less like plain echo). This is why some concert halls use large diffusion panels very much like these.
Lots of uses with wood. I often use it to repair veneer (furniture).
Well, I do. It’s a wonderful adhesive and a very nice surface finish.
Having a miter saw, I expected angular pieces to scatter the sound waves.
It is a nice artful piece.
I expected the same. I wonder if he didn’t do it for a reason or if he didn’t think about it due to some preconception.
Phase. The same principle applies in optics for spatial light modulators. Locally changing the phase (usually done in a discrete manner just like this) causes the overall wavefront to change..
I suspect some optimization could be done to make this approach much more effective, since the height of each block here appears to be random..
This has nothing to do with ‘diffusion’. But then again, like this fellow, not many people understand the physics involved regarding acoustic diffusers and just enjoy the aesthetics of something that resembles one.
Phase changes will bend the wavefront, so the more sharp the edges of the wells are the higher its performance will be (but in his case there is no diffusion performance, maybe a small amount of scattering but that is not diffusion).
If the wells are not square it just means a different lowest effective freq for each axis (vertical vs horizontal), but only when it would be effective :D
For true diffusion each reflected angle needs to be unique (not random!). Which is why we use primes in the design and calculations of Primitive Root Diffusers (the type of diffuser this guy tries to fake). If you need many wells (for a large wall) then you will need a high prime to base your design off to prevent repetition of angles, mine is 2971 to fill my back wall.
Depth and size of the wells are precisely chosen based on speed of sound and the resulting wavelengths to meet acoustic criteria that were defined by measuring the acoustics of the problematic room. The well size this guy (had to) chose is useless.
Do the panels this guy made provide ideal diffusion? As you point out, no, because they’re not based on the mathematics behind diffusion. Are they useless? Hardly. They are not flat, so they redirect sound, predominantly high frequencies. That is, by definition, diffusion. A bookshelf can do much the same thing, even though it’s not your (quite impressive, I’m sure) high-prime-based solution. I have heard the difference, and it’s far from useless.
Once I also thought they were the same, I could not have been more wrong, they are related though. You are really talking about scattering, by definition :) https://www.physicsforums.com/threads/scatter-vs-diffusion.110261/
For sound diffusion I just have a wall of spare parts and junk just sitting in the shelves all in irregular shapes boxes or no box at all – Some times it pays to be a disorganized mess.
Definitely a good-looking approach, and I’m sure it’s diffusing sound very well. But it is a lot of effort for the amount of diffusion.
Back in the day, we used egg flats ;-). You can often get good results with some clever placement of furnishings. Bookcases with artfully random placement of book sizes are effective. You can hang decorative carpets or tapestries on the walls. Grandma’s quilt. Etc
+1 for egg flats. We’ve used these to soundproof an amateur recording studio, with excellent results and very low cost. They don’t look half as good though…
So, which came first; the soundproofing or the studio?
B^)
A CLASSIC!!!
Egg flats are definitely useful but they tend to help with high frequency absorption rather than diffusion.
Both are often necessary in creating a pleasant-sounding space, unless you’re going for totally dead.
If you want to do this right then look at a few research papers on the subject, just randomising the block lengths is naive.
In fact those diffusers should be wells (a collection of special reverb-boxes) instead of free-standing blocks (just disperses waves). If you make free standing boxes, you can as well just make them random, there will be no special diffuser effect anyway.
There is perfect diffusion, and then there is enough diffusion. Just because a solution doesn’t conform 100% to the mathematics in a scientific paper means the effort was wasted. The bookshelf behind my recording position is not perfect, but it’s enough. Perfect is expensive, and typically unnecessary.
This will not diffuse, it might scatter a little. Then again it is way too shallow. If you need inspiration for a real solution that has been properly calculated adhering to physics look at my diy stab at Primitive Root Diffuser design: https://www.electroacousticlabs.com/stories/
As for insiration, it has the same level as this article. But some more “calculation” would be nice. Like, there is exactly 0 calculation explained on your blog. Nice results though, I liked how you can make it much more easily by cutting whole rows of polystyrene.
Searching on google shows enough examples, just search for “Primitive Root Difussers”.
Thanks Krzysztof! The big take away is, I am not saying I built a plane when I just built a 1:1 scale model plane without an engine. Maybe I’m the only one, but I get inspired by proper implementations of scientific principles. But you have a point, I am just too busy with a completely new design. I will keep your advice in mind when writing up the new project.
No problem, but from what I’ve seen in patents and wikipedia, just pillars is not a proper PRD, it should be wells, it should look like a big array of boxes of different depth, not like pillars. Otherwise, you have much more scattering from sides of those pillars.
I’ve read your BBC paper just after my comment, so I stand corrected, please disregard my comment.
The BBC begs to differ: http://downloads.bbc.co.uk/rd/pubs/reports/1995-01.pdf
“The background and theoretical basis of the BBC
design is given in Ref. 6. The numerical basis of that
particular design resulted not in a uniform reflection
pattern but in a selective cancellation of the reflected
energy in the specular direction, at some frequencies.”
I know that paper by heart, you did not read it. REF 6
“The background and theoretical basis of the BBC
design is given in Ref. 6. The numerical basis of that
particular design resulted not in a uniform reflection
pattern but in a selective cancellation of the reflected
energy in the specular direction, at some frequencies.”
I actually know that paper by heart, but you did not read it it seems :) please check ref 6. Let me help you: the pattern is not random.
That is an awesome diffuser!
I’m still at the “put bookshelves on the walls” phase…
Thank you Elliot, bookshelves are perfect scatterers and was on my list of options also, I love the looks of them, it is the first step towards true diffusion. But once you hear a true wide band diffuser your brain will make you wonder how did that wall disappear when I can still see it.
I know the phenomenon very well. There was this lecture hall in university, and if you’d walk along the back row, next to the sound-absorbing wall, it was almost disorienting.
Fun fun!
The way a friend described walls similar to that to me, in his sound studio, about 20 years ago was think about a concave fresnel lens works, now imagine making that exact structure out of wood – it would probably do an reasonably Ok job at focusing sound towards to a point or a line. Now that you have that image in you head now imagine the effect if it was made up of wooden blocks with random lengths and orienteered at sloping angles. If three truly random numbers are used for the length of the wooden blocks (within a narrow range, it is a form of fresnel after all) and exact angle (also within a specific range, you do not want any that could accidentality stab someone if they fell against the wall) and rotation (zero to three hundred and sixty degrees) at the front of each block it “should” do a very good job at dispersing waves in random directions.
I should probably add that the one I was standing in front of was created using three random numbers per block, and it was a very odd experience. I knew that there was a wall at about arms length directly behind me, but my brain from the audio cues in that physical space was telling me that there was nothing behind me at all.
There’s reverberation, and there’s discrete echoes. You can measure reverb with a mic and pulse generator/speaker, and measure the time that the reverb dies away (to -30 dB as I recall, in seconds). There’s a value that sounds nice, and a value that makes speech unintelligible. Unfortunately I forgot what they are… maybe something like a second for speech, couple seconds for average music, and several seconds for Gregorian chant in a cathedral.
One of the best absorbers is fabric. Tapestries look good on the wall and soak up a lot. I know for a fact that some NPR news reports have been recorded in a clothes closet during pandemic time.
And you don’t want discrete echoes unless you’re a bat searching for bugs. Hard parallel walls give you that slap-back sound and also create standing waves at certain frequencies. Irregular hard walls can give you smooth reverb with no resonances.
“Spaces like … hammocks in the yard have all become “offices”.”
The future has truly arived.
Back in high school we were asked to bring in cardboard egg cartons, which we did by the hundreds. These were fixed from floor to ceiling across the back wall of the music room, to kill the reverb.
I recall going on a school trip to the Sydney Opera House and being told the thin wood panelling of the main concert hall walls had a purpose other than just appearance, and that was they had the property of reflecting high frequency sounds back into the hall whilst absorbing the low frequency sounds. I thought that was very clever.
yes. Also, the position and spacing of those panels from the wall, and whats behind the panel, can be chosen to help eat up problem resonances at lower frequencies (mentioned below).
Whole foam mattresses eat up bass as well. Foamed plastic egg containers, why. Unfortunately only the solid wood is fire safe.
I offered a bunch of foam shipping pads (with the egg crate type of cut in them) to an acquaintance for his home studio.
But he didn’t want them outgassing in his house.
Rockwool is an excellent sound absorber. I’ve placed 4 panels of 1.2m by 2.53m and 0.1m thick with rockwool of 80mm thick in 2 strips of 600mm wide in my living room and it got rid of the reverb completely. I’ve put it in a wood frame with thin plywood back and upholstered with canvas. Doesn’t look half bad either. Only some first order reflections from windows,TV and ceiling remains but that doesn’t bother me.
Even one such panel has a noticeable effect. I’ve put 2 of them in a corner and when standing there it’s absolute silence. Complex orchestral pieces really come to life now as staccato pieces don’t become a muddy mess anymore allowing you to hear the individual notes, and electronic music with sudden breaks sounds great!
Many studios have been treated using 2″-thick rigid fiberglass panels, framed with 2x2s, and fireproof architectural cloth stapled over that. It’s easy to make these up in a shop, then install in a room.
Nice project. It is perhaps more a scattering panel than a diffuser. But it looks nice and if covered much more wall would probably have worked well.
Have thought about building these things as, well, acoustics and Hi-Fi are fun and a hobby that the wife doesn’t disapprove of too much. Also, designing acoustic chambers was at one point my job.
The thing is, they need to be big to be effective. Or at least more big is more effective. So for that reason, my one is somewhat hidden in plane sight. When an extension was being built, rather than the wall behind where the Hi-Fi would go being blockwork and covered with plasterboard, had it made of bricks. The bricks match the exterior of the house and are a supporting wall, so nothing fake here. They’re quite nice too as far as bricks go, being heavily textured and a bit rustic. But the raking of the mortar is what does the trick. It’s about 8 to 10 mm back from the front face of the bricks and square edged rather than rounded (no need to shed rain indoors I hope). This is to scatter sound – not diffuse. And it works very well. It’s also big, being a wall and everything.
Yeah exactly, I agree it just ‘looks like’ a diffuser. But it will not work (our brains need a little bit more science and effort to be fooled).
Acoustic treatment of rooms is a science by it self. These old-school diffusers works well at diffusing echoes if you cover most of the wall. A big bookshelf with random books works really well too.
When it comes to damping the echoes the most common error is putting up to shallow dampeners. a drape for example is just going to swallow all high freq and do nothing with the most problematic areas lower down. a proper studio is frequency measured and the treatment is built from that to have a linear damping of the whole register.
Something that i find work surprisingly well for most rooms is to use BAD-panels. (just google them) they are panels with a specific hole pattern that works by phase cancellation/diffusing. and on the backside there is damping. What they do is diffuse higher frequencies and dampen lower frequencies. i’ve build my own and there are pattern guides online for the panels.
As an aside:
Many woodworkers claim they don’t have “scraps”, they have small pieces for small projects, the pieces that are too small for small projects are fuel for the shop stove.
B^)
https://hackaday.com/2021/08/21/keep-scraps-around
Just sayin’.
Some people need to recognize that just because they’ve spent a lot of time reading about, or even actually working with, acoustic diffusion, doesn’t mean they can discount someone else’s efforts because they don’t meet whatever standards they hold for their own work. This solution diffuses sound. Not ideally, for sure, and likely over smaller frequency window than required, but scattering sound is diffusing sound, even if you wouldn’t be comfortable selling panels like this as “acoustic diffusers”. Which I probably wouldn’t.
And before you accuse me of being ignorant, I’ve been a practicing acoustician for 15 years. I know whereof I speak. But I’m also a pragmatist, and not a purist. If someone’s paying for actual acoustic diffusion, they’ll get it. But if they want something useful, and not ridiculously expensive, I know how to do that, too. And I won’t get all sanctimonious about it.