Metlako: A Small, Affordable Two-Way Unity Waveguide

Here's a recording of Metlako III

Obviously this is a work in progress, so there's room for improvement, but I dig it

Here's the original song, you can compare and contrast the original recording versus what's reproduced by the horn. In particular, note how good the percussion sounds. (Unity horns are really good at percussion.)

To compare and contrast, I strongly suggest using headphones or a full-range single driver speaker. Conventional speakers screw up the phase in the crossover region and will obscure what Unity horns do so well.
  • Like
Reactions: 1 users
It's 2am here so I'll need to post the STL files and the model tomorrow, but here's an animated gif that shows:

1) the distortion of the "stock" SB26ADC tweeter

2) the distortion of the SB26ADC tweeter with a metamaterial absorber that I designed and 3D printed

3) same as above but with polyfill in the absorber

4) same as #2, but with rockwool in the absorber

I think the results are interesting. With the last metamaterial absorber that I made, documented in the thread "3D printed metamaterials", I didn't use any stuffing. But today I found that stuffing reduces distortion even further than with the metamaterial absorber alone. In all, the metamaterial absorber with rockwool stuffing reduced harmonic distortion by about 6dB below 1khz. More importantly, it flattens out the impedance curve quite a bit, which makes it a lot easier to use a simple passive xover on the tweeter. Basically you can avoid using a zobel filter. Which is especially nice, now that inductors cost about ten bucks these days.

You'll probably notice that the overall frequency response is pretty rough. This is because I measured the tweeter with no baffle. I wouldn't go that route if I was focused on evaluating the frequency response. But since all I was doing was measuring distortion, and I was doing an "apples to apples" comparison (all variations were measured with the same voltage, distance, etc) I think the results tell me what I need to know.

Which is basically: metamaterial absorbers work. Realistically we should probably be using them in ALL sealed boxes, particularly tweeters.

It's a pretty simple print too. It took less than two hours and the cost of the filament is probably about three bucks.


  • sb26adc.gif
    393.4 KB · Views: 69
According to "Sound on Sound:"

"Second-order harmonic distortion comes from non-linearities in the system — for example, the 'air springs' on either side of the bass driver being different (the air on the inside of the cabinet is stiffer than the open air outside, and more so in a sealed cabinet than a vented one) — and it sounds nice. It brings 'warmth' and 'fullness' to the sound."

That would seem to indicate that the reduction in distortion we're seeing is due to the absorbtion of the tweeter's back wave. Taken to the extreme, you'd absorb 100%, similar to what B&W strives for in their "Nautilus" speakers.

Some might argue that one could get the same results by simply using a very large back chamber for the tweeter. But there are additional advantages to the metamaterial absorber:

1) It's pretty darn small. Mine isn't remotely as small as the one in the Kef speakers, but it's still small enough to fit in the palm of your hand. (I'll post pics soon and the STL file.)

2) I believe the reduction in impedance is a pretty big deal. It makes the crossover a heck of a lot easier (and cheaper.) And another cause of distortion is variances in inductance. Since the impedance peak is reduced substantially by the metamaterial absorber, it stands to reason that inductance related distortion should also be reduced. In order to test this out, I'd likely need to test at much higher power levels (to push the excursion up) and I wasn't too keen on doing that because I'm not eager to destroy my tweeter.
Moar pics.

I included a Behringer 6.5" two-way for comparison's sake.

3D printing allows you the opportunity to pack the drivers together to a ridiculous degree.


  • IMG_20230524_172000.jpg
    349.4 KB · Views: 99
  • IMG_20230524_172005.jpg
    384.5 KB · Views: 95
  • IMG_20230524_172017.jpg
    242 KB · Views: 93
  • IMG_20230524_172041.jpg
    450.4 KB · Views: 100
  • IMG_20230524_172047.jpg
    474.8 KB · Views: 102
  • Like
Reactions: 1 users
Firstly, I'd like to say well done as usual Patrick, you always inspire me to try more with my own projects and see what possibilities lie ahead.

I wanted to talk about the metamaterial absorber and why it's reducing harmonic distortion... It more than anything just seems novel to me. I'm hypothesising that it's acting somewhat as a "horn" (not a horn, but an additional transducer I guess?) and helping with energy transfer, thus loading the dome more (less excursion) - it reminds me in many ways of an aperiodic enclosure in the way it reduces the Q value of the resonance peak, both designs seem to transfer acoustic energy into a resistive load (releasing heat).
However as noted by your "apples to apples" comparison the SPL also drops for a given input voltage, so I'm wondering if at the same given SPL (not voltage) the distortion will be at all improved with the metamaterial absorber (in particular at resonance), as it stands to reason that the second harmonic should drop somewhat, but what about odd order harmonics which are not cause by asymmetry?

Anyway, not to to discourage you at all, I really am just curious as to the outcome, it's really cool to be able to discuss these kinds of things.
Attached are some pics and the 3D models for the waveguide.

If you just want to get to the print, grab the file titled "Metlako III STL"

If you want to make any changes to the original model, grab the zip file with the 123D file inside. 123D hasn't been updated in years, but it's still free and it's my fav.

The finished speaker has a roundover that's fundamental to it's design. That file is attached and named "roundover may21"
With the roundover, note that you must print four, one for each quadrant of the waveguide.


  • roundover
    93 KB · Views: 19
  • Metlako III 123D
    3.3 MB · Views: 20
  • 2023-06-04 02_17_43-MeshLab 2022.02 - [Project_1].jpg
    2023-06-04 02_17_43-MeshLab 2022.02 - [Project_1].jpg
    51.1 KB · Views: 45
  • 2023-06-04 02_15_54-MeshLab 2022.02 - [Project_1].jpg
    2023-06-04 02_15_54-MeshLab 2022.02 - [Project_1].jpg
    32.3 KB · Views: 45
  • 2023-06-04 02_13_41-MeshLab 2022.02 - [Project_1].png
    2023-06-04 02_13_41-MeshLab 2022.02 - [Project_1].png
    49 KB · Views: 43
  • Like
Reactions: 1 users