You said that you wanted the walls to radiate one ambient color, but let me talk about a situation in which the color across a wall varied somewhat. Picture the same room, but where the brightness of each wall went down as you moved further from its center. In this case, it might be possible to discern the sphere in the room.
It turns out that the reflectance vector’s angle from the observer is twice the size of the angle between the surface nomal (the “up” direction at that point) and the observer. When the surface angle is 45 degrees from the observer, the reflectance vector is at 90 degrees. As the surface angle gets close to 90 degrees, the reflectance vector gets close to 180 degrees. That is, as you look at points getting closer and closer to the outer edge of the sphere, the two rays from the observer to the sphere and then to the wall get closer and closer to a single ray headed straight towards the wall.
That would mean that the sphere would reflect everything in the room that it did not occlude. Additionally, the color on very edge of the sphere would blend into the wall that it was in front of. So the fact that there was a reflecting sphere in the room would be pretty apparent, but exactly where the sphere ended and the walls began would be a bit difficult to determine. Not completely, because the eye not only detects disconinuities in the color across the view, but also detects dicontinuities of the color gradient. So, the fact that the change in the color across the wall will immediately reverse itself across the sphere might make the border easier to see.
And it would be even easier to see if the sphere’s surface was not perfectly reflecting. All mirrors that I know of don’t reflect 100% of the incident light. With physically realistic materials, there would be a noticable color discontinuity. Also, if the sphere was large enough, there would be some viewing angles where you could get the sphere to occlude a wall’s edge, at which point the location of the sphere’s edge (at least on that side) would be apparent.
I don’t know what game you’re playing, but requiring the characters to make a successful saving throw vs madness (or whatever) in order just to look at the thing might go a long way toward emphasizing the weirdness of it.
Ok, I did some rendering, but there’s nothing to show you (ironic, huh?); a perfectly mirrored sphere in an environment with completely featureless ambient lighting and backdrop is utterly indistinguishable from the background. The only thing that makes shiny things shiny is that they reflect objects that aren’t uniform in appearance and that they are not uniformly lit, leading to specularities such as the classic ‘window’ squares on cartoon balloons.
Cut from its featurelss grey backdrop and inserted into another scene (and assuming that it wasn’t affected at all by the lighting conditions or surroundings of the surrogate scene, the sphere would look like a flat, featureless grey disc - it would actually be impossible to discern that it was spherical except by touching it. (even walking around it, you could be fooled into thinking that it was just a fetureless disc that somehow turned to face you wherever you are).
You might be able to make a spherical mirror look anomalous by having it reflect something that has features, but isn’t its surrounding environment, but even that might not look terribly wrong, because by definition, the things it is reflecting are behind you, out of your field of view and so the contrast might not be all that striking.
Guess he was right after all! Thanks so much for testing this out for me, mangetout. I think I was imagining the object with a single directional lightsource; this would make it look mirrored, right? That’s probably how I’ll revise the description: it looks like it’s reflecting a single light source, but the light source is invisible, and nothing else is visible reflected in it. Think of a pinball.
If you make it so that it reflects the lightsource as a specularity, but doesn’t reflect any other features, it will look polished, but black, more like an 8 ball than a pinball.
What makes silvery things look silvery is the way they reflect everything.
You might get the visual effect you’re looking for with a transparent clear sphere rather than a reflective surfaced sphere. I seem to recall POVRAY being able to deal with clear solids of arbitrary refractive index.
Since this is a roleplaying game, why bother to make it utterly realistic; when your buddy says ‘but that’s physically impossible’, just tell him that’s the whole point.
Have none of you doubters ever seen a ball bearing or a tiny sphere of mercury?
Of course you can have a shiny globe that does not (appear to) reflect the images of other objects:
You simply start with an object so small that any object that is reflected is too tiny and distorted to be recognizable, leaving the “shiny” image of the sphere in the mind, then imagine that shiny object much larger. The reality is, of course, that the larger the shiny sphere became, the less distortion each reflected object would have, leaving images reflected from the sides, just as most of you have noted. However, for the purposes of the OP, such an object, where no reflected image can be discerned, allows one to imagine a shiny, non-reflecting sphere which can then be imagined in a much larger size.
(Of course, we no longer let small children play with tiny globules of mercury, but I also wonder whether mercury might not have some properties of physical structure, reflecting so many tiny images from so many facets that only the “shiny” image is perceived, since no reflected image can be discerned. I do not actually know, one way or another, not having seen globules of mercury in 45 or more years.)
