Love the topic! Thank you Eric!!
Most horns are designed with acoustic amplification as the top priority. A few are designed with low coloration as the top priority, Some of those have the potential of imaging well.
I’ve been working with horns and waveguides for many years now, and during this time I’ve had the privilege of working with Earl Geddes, whose waveguide-based designs (a waveguide being a particular type of horn in this context) image extremely well and totally disappear as the apparent sound source.
Okay time to get nerdy: The ear/brain system gets its localization cues from the first approximately .68 milliseconds of sound. In that time, sound travels about 9 inches (corresponding to the distance around the head from one ear to the other). A poorly designed horn has lots of opportunities to screw things up within this time window.
Reflections and diffractions within the horn will happen during that first .68 milliseconds. Diffraction (from sharp edges and discontinuities) is worse than reflections. Reflections within the horn are inevitable, but some geometries generate more internal reflections than others, and/or direct more of these internal reflections towards the listener than others.
My understanding is that the best horn geometry from the standpoint of internal reflections is the Oblate Spheroid (what Earl uses). I use a Super-Elliptical Oblate Spheroid which comes very close, and which may have a couple of advantages in other areas.
So to sum it up, proper horn geometry is absolutely crucial if imaging matters. Most horns screw up that first .68 milliseconds of sound, but a few are benign in this respect. Those are the ones that have the potential to image well.
Here is a link to a show report that comments on the imaging of one of my systems. My room is covered starting about 1/4 of the way down from the top of this page, and the writers are Tyson and Pez (Jason) from AudioCircle.com, who for several years (before they burned out) were arguably doing the most precise show reports of anybody because they pre-selected their rooms, listened to the same few short revealing tracks in each room, wrote quickly, and then moved quickly to the next room on their list:
https://www.audiocircle.com/index.php?topic=120504.msg1267150#msg1267150
In his subsequent "Best Of" report, Pez had us tied with several rooms for second place. For some reason I can’t post the link. Anyway, he wrote: "This room had the most locked in soundstage and imaging I have ever heard bar none with dynamics to match. The sweetspot is just an incredible experience and really musical top to bottom."
Did you catch that? A HORN system had "the most locked in soundstage and imaging I have ever heard bar none".
Tyson had us tied with one other room for first place and wrote: "Basically did everything as well or better than any other speaker at the entire show... And, amazingly, they did it with zero room treatments."
(We have good radiation pattern control and we aim those patterns intelligently, so we can get away without room treatments if we want to.)
If anybody is going to be at RMAF 2018 next weekend, you are invited to Room 3002 to find out firsthand whether or not "horns done right" can image well, and/or otherwise compete on the basis of freedom from coloration, natural timbre, disappearing as the apparent sound source, precision, and whatever else horns traditionally are presumed to not do well.
Duke
got dogs in this fight!