What a beautiful room with some great gear in it.

Mctwins solution looks great, provided those diffusers are designed for close up use. If it's feasible, I'd suggest you move the listening position to at least 2 metres from the back wall. Right now you're sitting within the minimum listening range of most diffusers, so rather than getting the benefits of diffusion you may hear artefacts. If you don't want to move your listening position, you could get a hybrid absorber-diffuser that absorbs at low frequencies and only diffuses sound above about 1000 Hz.

Typically, you can take the minimum listening distance to be about 3 times the longest wavelength diffused.

I normally suggest 1D diffusers on the back wall, but if you're sitting that close you'll want limit the amount of sound energy that's scattered back towards you, so an efficient 2D diffuser might cause less artefacts. Another approach (if you insist on diffusion and you want to keep your listening postion where it is) might be to mount some broadband absorption with strips of high frequency diffusion (e.g., RPG flutterfree) over top of it.

Thanks

The Wing diffusor behind me is not a diffusor, per say, like ordinary diffusors but instead they are broadband time delay lines which are optimizes in the time domain.

Here is a measurement where the mic is 1meter from the V-wing.

https://www.gearslutz.com/board/7186699-post71.html

So, if I would have more space around me, it would look like this behind me.

Those diffusors looks great, but I would call them something else other than broadband time delay lines. I expect time domain diffusion (or "temporal dispersion") is their function.

Your graphs show a strong early reflection less than 2ms after the direct sound, but I'm assuming that reflection is from the floor because sound travels 0.688 meters (or about 2 feet) in 2ms.

If you're 1m away from the V-wing I'd expect a delay of about 6 ms between the direct sound and diffused sound, and it looks like your graphs shows that. So I personally think "time delay lines" is misleading because they don't delay the rear wall reflection. I'd call those wing diffusors or scatterers (depending on their broadband diffusion performance), or if you want to be really geeky you could call them something like "temporal scatterers".

Just my thoughts.

Ok, I trying to decide what to do, I have 2.1meters in the back of my well treated composing room and the original design had a diffuser there. So, I am tempted to build the array of 5 A1LF. Maybe I could fractalize it like Schaap did. But I am tempted to try and build a B2frac.

Has anybody built one of those? Is it worth the hassle? The main problem I think is making the small fractal units with 8.57 width. (I live in argentina, and we have 9mm or 12mm MDF, to build the base structure, other woods would be to expensive).

So bottom line: how is the b2frac performance compared to a A1LF (with and without fractals)....maybe is unknown because nobody has built one

Ok, I trying to decide what to do, I have 2.1meters in the back of my well treated composing room and the original design had a diffuser there. So, I am tempted to build the array of 5 A1LF. Maybe I could fractalize it like Schaap did. But I am tempted to try and build a B2frac.

Has anybody built one of those? Is it worth the hassle? The main problem I think is making the small fractal units with 8.57 width. (I live in argentina, and we have 9mm or 12mm MDF, to build the base structure, other woods would be to expensive).

So bottom line: how is the b2frac performance compared to a A1LF (with and without fractals)....maybe is unknown because nobody has built one

On it's own the B2frac module has the best performance, BUT the array of A1LF diffusers can be arranged to perform better than the B2 arrays IF you use one of the profiled modulations given on the last pages of the blueprints document.

The fractal designs add high frequency diffusion performance and I encourage you to try them out if you have the means. If you're tempted to build a fractal design, I recommend you try making A1frac diffuser modules (the specs are not yet posted, but all the information is in Chapter 8 of the thesis). Then, I recommend mounting 5 or 7 modules according to the profiled modulations given at the end of A1LF blueprints.

If I had the time to build a fractal design, I would build 7 A1frac modules and mount them using the fractal modulation labelled "Profiled Modulation 2". The resulting design would have high frequency diffusers (the little fractal cells), nested inside A1LF diffuser modules, which are nested inside a big profiled array (7 modules mounted with Profiled Modulation 2).

Effectively, a third-order fractal diffuser .

P.S. To my knowledge, no one has done this yet ;-).

so, the A1frac is a A1LF with "little A1LFs" on top of each steps? (0 11.4 14.25 8.55 14.25 11.4 0)

I have space for 5 modules, so the sequence would be:
Build 5 A1LF
Build the fractal modules
Mount them in a profile modulation 1

several questions arise:
1. Should I fill all the cavities with fiberglass so they don´t resonate? (the profile and the ones inside each module?)

2. if I use 12mm MDF (the imperial model you tested) do others dimensions have to change, or each step can remain at 60mm width, and the fractal heights too?

I hope it´s not to many questions, but it would be a great project to tackle!

You got it. A1frac is a A1LF with "little A1LFs" on top of each step. To answer your questions:

1) It's not absolutely necessary, but I would do this. If you were feeling adventurous you could try perforating this diffuser with widely spaced holes (using this online Helmholtz Absorber Calculator) to turn it into a combination diffuser / low frequency absorber. In this case, having it stuffed with fiberglass would widen the bandwidth of absorption.

2) Normally I would say keep all the ratios the same, but I've tested this particular case and it gets very good performance if you use 12mm MDF while keeping the 60mm width the same. I've not tested any variations with the fractal cells, but I think if you scale up the fractal heights by a factor of 1.2 it probably won't hurt, and may in fact provide better high frequency diffusion. It's up to you though.

Hope this helps!

Ok, due to wood size (and mill size) restrictions here, I think the best solution would be similar to Schaap´s one:

for the fractals: 40mm slat (5x8mm + 2 at the base) 18mm depp so we would have (rounded):
0-14-11-18-14.
Those numbers are close to the original numbers by a factor of 1.2

The problem is that the base slats of the A1LF would be 56mm, is that ok?
Are my numbers correct according to the theory?

Ok, due to wood size (and mill size) restrictions here, I think the best solution would be similar to Schaap´s one:

for the fractals: 40mm slat (5x8mm + 2 at the base) 18mm depp so we would have (rounded):
0-14-11-18-14.
Those numbers are close to the original numbers by a factor of 1.2

The problem is that the base slats of the A1LF would be 56mm, is that ok?
Are my numbers correct according to the theory?

It's okay if the base slats are 56 mm wide. Alternatively, if you have 9mm material for the fractals, you could make the base slats 7x9mm = 63mm wide.

I calculated the fractal depths if they're scaled by 1.2, and I got something different than you:

1.2*[0, 11.4, 14.25, 8.55, 14.25, 11.4 0]mm = [0, 13.68, 17.1, 10.26, 17.1, 13.68, 0]mm

So rounding, I would make the fractal depths [0, 14, 17, 10, 17, 14, 0] mm.

I got the same numbers for 1.2*, but I can´t get 17mm mdf here.

So I did it with 18mm as a base, with 3.6mm (18mm/5) as step, I got:
0 - 14.4 - 18 - 10.8 - 18 - 14.4 - 0

I think it would be ok, I am trying to find 8.5mm or 8mm cutter for the router. 9mm wood make the five modules too big for the wall space.

Ah, I see now. Thanks for clearing that up. I think your approach should work well. It's useful knowing what sizes of mdf are available for people, so it's great that you're bringing up these details.

Keep us updated on how it goes, and if you take any photos of the build progress or completed build I'm sure everyone in this thread would love to see them!

hey. great work with diffuser blueprints. i couldn't find vector blueprint for cutting fractal diffusers from XPS so i did one myself. first one is like the original 42cm wide and the other is rescaled to 50cm to fit my absorbing panels. from one sheet of XPS (125x60x10cm) you can make 1.5 m2 of fractal diffusers 50cm wide.

i plan stack and glue those 10cm pieces together. what do you think about this idea? i will try it out next few days.

https://www.gearslutz.com/board/attac...s-cutting-.zip

hey. great work with diffuser blueprints. i couldn't find vector blueprint for cutting fractal diffusers from XPS so i did one myself. first one is like the original 42cm wide and the other is rescaled to 50cm to fit my absorbing panels. from one sheet of XPS (125x60x10cm) you can make 1.5 m2 of fractal diffusers 50cm wide.

i plan stack and glue those 10cm pieces together. what do you think about this idea? i will try it out next few days.

https://www.gearslutz.com/board/attac...s-cutting-.zip

Thinking of something similar myself now.
Out of interest in Australia the common MDf sizes are 6mm, 9mm, 12mm, 16mm and 18mm.

Thanks so much for posting these vector blueprints! I like how efficiently you've utilized the space on each sheet.

For some reason I could only open the 50cm wide .cdr file (I got an error when trying to open the 42cm wide one).

Yes, I think stacking and glueing these together would work.

Now, a real challenge would be to build these out of something other than expanded polystyrene. I think it's nasty stuff but it just so happens to be one of the easiest materials for implementing this complex Stepfractal design. That's something I should have considered before posting the designs for such a complex diffuser shape.

Anyway, it's my responsibility to suggest an alternative to people, so I'm going to start suggesting A1-Frac "The Leanfractal" as a simpler and more elegant alternative to B2-Frac. A1-Frac actually has better diffusion performance than B2-Frac... IF you mount the A1-Frac modules using one of the Profiled Modulation schemes given in the A1-LF fabrication drawings.

You can see an image of an A1-Fractal diffuser module on the homepage here. The base shape can be built using the A1-LF blueprints, and the fractal cells can be built in a variety of different ways (Schaap build them using a milling machine and has posted some fractal diffuser photos like these throughout this thread). I've attached his photos to this post for easy viewing.

Thanks for posting these files and please let us know how the project goes!

Tim

Photos below are of the fractal diffuser (A1-Frac / The Leanfractal) built by Schaap.

Thanks for this info.

Through testing we've found that material close to 12 mm thick (e.g. 12.3 mm, 12.7 mm) is a great substitute for 10 mm material as the basic building unit. For example, simulations with 12.7 mm step increments performed slightly better than 10 mm step increments, likely because the diffuser is deeper when 12.7 mm material is used.

by mistake one .cdr file was saved in corel 15 and the other one in corel 9. if you have older version of corel this might be the reason why you can't open it. anyway, yesterday i visited company which will cut XPS for me. it's pretty cheap (in Poland around 17 EUR for one XPS sheet with cutting which gives us around 1.5 m2 of diffuser). unfortunately the cutter needs bleeding around 2cm from each side so i will need to find some other way to efficiently utilize the space. i will post new files when they are ready.

What combination of diffusers you would recomend for my space? i have around 120 cm on each side wall behind my listening position and in my back i have 308 cm of windows. my listening position is 150 cm from each side wall and back wall. more info about my room here: small studio problems

i was thinking about 2x60cm diffuser on each side wall and 5x60cm diffusers on back wall. should the side wall diffusers be on the same depth? or should i randomly mount in on different depths?

Yes I think this is a good approach. Also, see my comment in your thread (talking about acoustic treatments in your room). I bring up the minimum listening distance for these diffusers to ensure a good RFZ (reflection free zone) in your room.

It's up to you how you mount the side wall diffusers. It would probably be better to mount them at slightly different depths, but since there are only two modules I doubt the difference in diffusion performance would be that significant. Check out how John White mounted them in his studio:

Tim,

My dipole speaker designs (magnetic planar, electrostatic and open baffle) benefit from a front wall that is diffuse so I decided to build an array of QRD's to mount there. But after getting my butt kicked by building four 11-well units and still needing more units to get the entire wall covered I decided go with your brilliant diffuser design/solution to finish the job. Where each QRD build took me about five or six hours, not counting time spent calculating and obtaining materials, I built three of your units in under two hours. HELL YES! You can also see in the attached picture I've placed three 6" thick panels of SafeNSound on the front wall to reside behind the diffusion devices (as was suggested in another thread here in GS).

So, the reason for my post here is two-fold....one, to show off another build of your leanfuser and two, to ask if you think I'll be courting sonic disaster by combining the QRD's with the stepped diffusers. Also, you can see the center stepped diffuser has all vertical values doubled. I used 1/2" (12.7 mm) birch plywood so the total height for for the center device is 5 1/2" including the base plate. I thought I'd try it this way, with the center device having deeper wells/taller steps, after reading your comments about avoiding periodicity. Everything is screwed together, not glued, so I could easily take the unit apart and create a third and fourth unit of the same height as the outboard units (though I can only fit three units between the QRD arrays). It also occurred to me that perhaps I should consider replacing the QRD's with more stepped diffusers as that would result in the entire front wall having the same type of diffuser design. Of course, this would require more build time which I'm hoping to avoid as free time for me has suddenly become very scarce. Thoughts?

Thanks,
Michael

Hi Michael,

Nice work! I agree... three units in under two hours = hell yes .

Don't beat yourself up about building the QRDs. They look great and they will work well.

That's a great way to orient the 3 modules (with the center one having extra depth).

The only issue with mounting it between the QRDs is that the QRDs are so much deeper, and because of this they will block some of the scattering by the outer two Leanfusers. It would be better if the Leanfusers were mounted a few inches further off the wall so that they are not so shadowed by the QRDs.

Another better option (if you have space) might be to put the 4 QRDs in the middle, and 2 Leanfusers on each side.

Thanks for sharing your build!
Tim

Thanks for your input and compliments, Tim......and also thanks for sharing the Leanfuser design, BTW. I failed to include that in my last post and don't want to seem dismissive of your generosity.

I considered moving the QRD's to the center but that would create two issues for me. One, it would require I move the equipment rack further out into the room which can hinder sound stage resolution and two, I wouldn't have room for a pair of Leanfuser modules on either side of the QRD array. Now I would have room if I did a three unit QRD array in the center, but then I'd be left with one QRD and no good place to put a single unit (and no desire to build another now that I've been spoiled by the Leanfuser!). I totally see your point about the inner QRD's creating interference with the outer Leanfuser modules, but figured having them there is at least better than bare wall. I suppose I could do just two Leanfuser modules in the center rather than three and create two more Leanfuser modules from the leftover center module and place one on each sidewall. Do you think that would be a better arrangement wrt diffusion performance from the central units?

- Michael

Hey Michael thanks for your kind words! To me, simply seeing people build these is more than enough gratitude.

I think you're good leaving the setup as-is, but another possibility that might look interesting (or weird) is to stagger them:

QRD-Leanfuser-QRD-Leanfuser-QRD-Leanfuser-QRD

Considering how both the QRDs and the Leanfusers were designed, neither of these diffusers are designed to be mounted like this. But I hypothesize that the overall result would yield decent diffusion... possibly more uniform than if you mounted the three Leanfusers in the middle (bear in mind that this is just an intuitive guess).

Either way, the choice is up to you!

I've gone through the thread looking.... also checked Tim's website... but I can't find the imperial version. It was mentioned early in this thread that one existed (or was in works... that was 10 pages of reading). Anyone know where I can find the imperial blueprint?

I'm guessing I could just use 1/2" for everything and make it proportional but this type of thing isn't my strong suit and I'd rather check it against a sketch.

For example, if I use 1/2" board would the width of each well be 3" to maintain proportions?

.

Anybody here ever thought of gluing together stacks of FoamCore material to build these things?

.

[QUOTE=spiderman;9440451]I've gone through the thread looking.... also checked Tim's website... but I can't find the imperial version. It was mentioned early in this thread that one existed (or was in works... that was 10 pages of reading). Anyone know where I can find the imperial blueprint?

I'm guessing I could just use 1/2" for everything and make it proportional but this type of thing isn't my strong suit and I'd rather check it against a sketch.

For example, if I use 1/2" board would the width of each well be 3" to maintain proportions?[/QUOTE]

2 3/8"

[QUOTE=tinnitusintx;9440747]You can keep the width the same and use 1/2 board. The well depths and widths both effect performance. Narrower widths extend the HF range. Deeper wells extend LF diffusion range, but also the minimum listening distance. Anyone feel free to corect me if I am explaining this incorrectly.

You're right that the imperial version is not well documented.

Luckily, it's simple:

Just use 1/2" material, and but keep the well widths the same (6cm = 2 6/16"). This design performs slightly better than the original, even though the proportions are different.

Optionally, you can scale the proportions up and use 3" wide wells, but I think the performance will be better if you keep the well width at 6cm.

Hope this helps,

Tim