Do not shun away from the crappy stuff. On the night of 28/29 November 2016 diamond dust turned at one stage snowfall like, losing its glitter and much of its halos. My camera was on tripod snapping away shots and when the swarm suddenly became matte, I though of stopping photographing. But there was something that kept me going: I could discern faint long arcs on both sides of the subsun. So I let the camera run - photos might turn up something interesting.
Indeed so. The long arcs in the matte phase - shown in the single 30s exposure above - do not cross at subparhelia, but shoot past them on the outside. This is demonstrated in the collage below, where this image (lower left) and another image (upper right) from the earlier stage with subparhelia are superposed (upper left). Thus they are not the traditional arcs of Schulthess from Lowitz orientation.
In addition to the unorthodoxy of the arcs, the disappearance of parhelia and subparhelia during their occurrence is also interesting. Futher below is an animation of most frames photographed in that session (only the final decline of the display was cut out). The arcs can be discerned in photos for about 6 minutes, but the peak was short, lasting only 1-2 minutes.
So what are these arcs? I have only to give the usual answer: no idea. Some later displays, which I will post in time, seem to have them too and there are elements in those observations that point to the possibility of some kind of divergent light effect. But how that would be, divergent light simulations I have seen have not given out anything like this. So it is a mystery any which way you look at it. If they happen to be a real deal, then it must be assumed we already got photos of them in solar displays.
In the time lapse below, the arcs that are seen between the parhelia and subparhelia before the long problem arcs emerge seem themselves tricky as well. Why they do not cross over subparhelia like properly behaved Schulthess arcs are expected to do?