Opus 12 is the largest ATH horn designed to date — the result of turning horn reproduction into a true passion.

The main goal was to create a large horn that would support the most powerful compression drivers down to their low-frequency limits, covering most of the acoustic range with just one source. Reaching that goal alone would not be so difficult if one gave up on directivity and could live with strong beaming, but it becomes significantly more challenging when an only slowly narrowing directivity is another (in fact, opposing) requirement. Based on the development of the TwinRay waveguides, Opus 12 represents the current culmination of these efforts. As a result, this new big horn achieves virtually perfectly linear SPDI between 500 Hz and at least 12 kHz, rising only around 5 dB across this whole range.

Large-format compression drivers can be truly captivating — they render even the most demanding signals effortlessly, with dynamics to spare and with very low distortion. The biggest challange is the size of the throat itself, especially when it is 2 inches. Such a device will do (more or less) what it wants above ~7 kHz and there's no easy way around that. One viable approach is simply tailoring the horn to what the driver wants to do, and only make sure that the response is still gradual and smooth, without abrupt changes, diffractions and resonances. What this means, in a nutshell, is that the horn itself must be at least as directional as the source itself. And this means that it must be big to retain the required qualities. A positive effect is that also the acoustic loading is increased, which is given be the narrow coverage angle.

Three size versions are available as of now: Op.12.50 for a 2" throat, Op12.36 for a 1.4" throat and (a down-scaled) Op.12.25 for a 1" throat.

The large versions, with their 790 × 520 mm (31.1 × 20.5 in.) mouths, are obviously intended for larger-than-normal listening spaces. This mouth size corresponds well to a dual 15-inch woofer configuration, which was the basic design concept. A group of smaller woofers in a grid having a comparable overall size is also a perfectly viable option. Further details of the bass cabinet (closed, ported, open, etc.—all are possible) are left entirely up to the builder.

The Acoustic Performance

As always, the trusted ABEC/AKABAK from R&D Team was used as a BEM solver for all the optimization work. Simulations were performed "only" up to 12 kHz, as it is quite difficult to solve a full-3D mesh of such a big device in a reasonable time. The performance of the real-world compressions drivers will determine the behaviour above ~12 kHz anyway.

First set of measurement data with Celestion Axi2050 can be seen on diyAudio: https://www.diyaudio.com/.../post-8308374

The following figures show the simulated polar responses of the Op.12.50 for the horizontal, vertical and diagonal orbits, with the -6 dB contour lines marked black (the images can be manually opened in a new window to see the full detail). No additional smoothing of the data has been applied, so there may still be a bit of numerical noise apparent.

Performance of the Op.12.36 is basically the same as of the Op.12.50. The Op.12.25, as a scaled-down version of the 12.36, has everything shifted 1.4x higher in frequency.

The overall acoustic response, including the total radiated sound power, is extremely smooth and linear across the whole band. This ensures that both the direct sound and the reverberant sound field in the room remain spectrally neutral. The performances of all the three versions, expressed in terms of the ANSI/CEA-2034-A standard are shown in the next figures. The downward slope of the total radiated power curve (approx. -10 dB from 200 Hz to 10 kHz) means that the direct sound can be kept spectrally flat, which corresponds to a typical target for optimized room-correction systems.

The STL/STEP files

STL files are included because 3D printing the horns from those parts remains a perfectly viable option — particularly when a large, high-quality printer is available. At the same time, all kits also contain STEP files, allowing the horn designs to be readily adapted to any other manufacturing process as needed. The (printed) horn is intended to be used free-standing, supported at three points: two beneath the mouth and one below the driver flange (rubber feet work effectively). If a very heavy compression driver is used, the driver should be supported instead.

Each kit is designed as a three-part assembly. The mouth rim can be fabricated either as a separate component (potentially from a different material) or integrally with the main body — both versions are included. The driver mounting flange is provided as a separate part. How are the parts further subdivided to fit the available 3D printer build volume is left to the builder. One possible support-free division is illustrated in the following picture; additional subdivisions are naturally possible.

DIY Gallery - Opus 12 implementations so far