Help with HF horn combing/lobing

So I prototyped a single Minimalist MEH (Minmeh) using Faital HF108/2x4" and passively filtered the response quite flat 160hz-12khz. Sounded A+ until I swayed my head laterally -- just a couple inches -- then it became noticeably dull/flat -- and a couple inches further -- lively sound came back -- and so on and on. Classic comb-filtering?

This was with a single speaker at 3m. To explore further, I disconnected the mids and listened from ~0.5m keeping one ear constant-30°-angled toward the throat for maximum acuity, as I adjusted head position laterally. Well, obvious combing or multi-petal lobing on test tones 4-12khz; 10khz sounded like pulse-pulse-pulse as I shifted. Then I flipped my body/neck/head 90° pretending the horn was narrow rather than wide, and moved accordingly (with some difficulty). Same result, qualitatively. Again, this is a single horn with just the HF108 playing.

I searched but could not find this issue mentioned. Help, please.

Minmeh HornFaital described here:


 
Last edited:
So I prototyped a single Minimalist MEH (Minmeh) using Faital HF108/2x4" and passively filtered the response quite flat 160hz-12khz. Sounded A+ until I swayed my head laterally -- just a couple inches -- then it became noticeably dull/flat -- and a couple inches further -- lively sound came back -- and so on and on. Classic comb-filtering?
If this is what your MEH horn looks like:
Screen Shot 2025-05-04 at 11.23.58 AM.png

Then the HF driver's output will diffract across the large flat square surface around it, the diffraction will reflect off the horn walls, the multiple arrival times from those reflections combined with the direct path from the HF will cause "classic comb-filtering" peaks and dips, constructive and destructive interference.

Art
 
  • Thank You
Reactions: wchang
Very cool for palying around but still faaaar from optimal.

Thanks for the response -- learning its specific shortcomigs and potential improvements is my goal.

Assuredly non-optimal. With the mids pre-existing in-box the drivers could not fit any closer etc. "Optimal" in terms of max SPL or bottom extension in-horn weren't needed. Dispersion/directivity limited by cardboard etc. Actually this experiment exceeded my expectations in these regards given the (lack of) size. Very-very-HF atop my current hearing range was surely above normal, in the sense that it sounded like a harpsichord used to sound, when I was 2.5 decades younger tapping keys during concert intermissions etc. Perhaps that's not good? The top "demerit" in my book is likely going to be soundstage depth when completed in stereo (passively filtered). I don't know how "phase perfect" across bandwidth and angular coverage a 2-way MEH can be, even with DSP -- or for that matter, just a HF-only horn. I'm interested in this, but it now seems to me a veeery high bar to attain for a horn.
 
I don't know how "phase perfect" across bandwidth and angular coverage a 2-way MEH can be, even with DSP -
With FIR DSP, the phase (and frequency) response can be made perfectly flat.
DSP won't have any effect on angular coverage other than making time alignment independent from physical location, and minimizing the overlap region where physical location of the horn entrance affects the coverage.
- or for that matter, just a HF-only horn. I'm interested in this, but it now seems to me a veeery high bar to attain for a horn.
You have set the bar veeery low- if you simply make your conical horn's throat match the HF driver's exit diameter, it's off axis response would improve tremendously.
As an example, the polar response of a 3.5" TC9 "full range driver" is consistent up to ~4kHz, the wavelength of the horn throat diameter, all the way out to 30 degrees off axis, the cardboard horn wall angle:
3.5%22 TC9FD-18-08.png

A simple first order passive filter could flatten it's HF response above 1.6kHz.

Conical horns have some limitations as far as polar response, for instance the Danley SH-50 MEH is a bit rough, but maintains a fairly constant 50 degree pattern from 400Hz to past 10kHz:
SH 50Polars.png

The B&C ME 64 horn using a co-ax HF driver directivity is more constant:
Screen Shot 2025-05-05 at 1.54.00 PM.png

Anyway, having set the bar so low, improvements will be quite obvious.

Art