This post is inspired by the work that I have done in the field of computer graphics, where programmers are doing considerable efforts to simulate lighting/shading of objects.
Take a look at this image:
Yes, those are lensflares. I think none of us would say anything else. The link is taken from this
article on wikipedia. In the case of the above image the effect is pretty obvious, however I'd like to take a closer look on it in order to sort lensflares into three cathegories and to compare them with some of the orb images in the other posts:
1.) Direct flares from intensive light sources:
The flare around the sun in the above image would be an example of that. Generally lensflares occur in "high dynamic range" environments, that is lighting environments where the normal content of the image is very dark in comparision to small areas of the image which are extremely bright. The easiest way to reproduce that effect is to take a direct picture of the sun or of a light bulb. I guess we all know that effect, so no need to go into further details.
2.) Indirect flares from specular reflection:
Basically a surface has two types of reflections: specular and diffuse. Diffuse reflection is when a ray of light hits a surface and is not reflected into a single direction but is scattered by microbumps on the surface (imagine the way a piece of paper reflects light). Specular reflection is just the opposite: the ray is reflected in one single direction (a mirror is a good example for this). Again
wikipedia is a good read on this. Real world materials always have diffuse and specular reflections, however for matte materials the diffuse component is stronger, for shiny materials the specular component is stronger. If a specular reflection can be seen on a photograph it will mostly be much brighter than the surrounding (diffuse) part of the image (because the light is not scattered it does not loose much intensity). Also it will mostly only be a very tiny spot on the photograph that catches those specular reflections for curbed surfaces. So basically it yields a high dynamic range image again, a lensflare is likely to occur. The above image shows a lot of those artefacts because the material of that golden foil has a very high specular component and is curbed almost everywhere. Please note how the flares always occur where the curbature of the object is extreme (=edges). Now compare to some of the images in the other posts:
The orb is at the edge of that gravestone. Obviously I cannot know for sure, but to me this is a specular reflection that results in a lensflare. If the image would have been taken from just a little different angle (very tiny change is enough) that "orb" would not be visible.
3.) Indirect flares from bright objects outside of the frame:
This is the classic example of a lensflare. It's caused by light that falls into the optics but is not necessarily focussed onto the image plate. Compare the lensflare on the bottom right of the NASA image with this:
So yeah, to me this is another example of a lensflare (which happens to look like it's sitting on a gravestone).
I hope this wasn't to annoying to read, and of course it's just a theory, could be utterly wrong.