Cars-N-Cans
Addicted to Fun and Learning
Yeah it does depend very much on how the system images and speakers will tend to be unstable if there is inter-ear interaction and reflections. On my nearfield system I referenced it has passive cross-talk cancellation and control of reflections (via adding furniture and other items at strategic locations) and that is the one system that does it relatively accurately since the ears are involved as well. Since it’s two-channel it experiences quite a lot of aberration beyond about +/- 60 degrees, but inside that range it works well to locate things.Agree this demo is provocative. My KRK desktop monitors were very idiosyncratic; they changed their spatial characteristics but they didn't match the video in 3D space at all. However, my DCA Aeon 2 Noire phones matched the video perfectly in 3D space, move for move and angle for angle.. Most interesting, I heard the image go behind my head when the YouTube demo showed it, although I have not found another music recording which spatially leaves the frontal, 9-to-3 o'clock hemisphere with those 'phones, until this demo.
My experience is in general headphones will tend to localize somewhat behind the head, along the axis of the two ears since depth cues are not present with the drivers firing directly at the ears. An odd one as I said was my 560S headphones with features inside the cups to aid in interaction with the ears. These actually moved the image over top of my head. This was nice in music listening, but by itself it was an odd effect.
Its an interesting topic. As others have said, there is a lot of research on it, but it’s not as known in the public domain in general as things like speakers and their characteristics. Maybe one day a good reference will be made available that summarizes how headphones behave with respect to imaging. I’ve seen it referenced here and there but it definitely deserves its own treatment since it’s quite nuanced.