Response from Prof. B.

(Dave's comments in blue)

That is one good advertisement for a cognate course. Does Cog Sci 77 have its own home page? Is that where the link to the illusion gallery came from?

No, no home page... If anyone's interested, we meet at 11 on M,W,F in the Moon Lab; the whole class is taught using a multimedia presentation program, so there's lots of nifty demonstrations.

Although I am quite willing to have my victims encounter my room illusion from a distance of ten or more feet, I wonder whether or not the Cog Scientists might devise refined illusions that will be effective even at closer range. I am reminded of Cinemascope, a technology of the fifties that coexisted with early 3D films that required throwaway glasses. The Cinemascope screen was large and curved, so a fairly good-sized audience could receive a pretty effective illusion of wraparound participation. Just as the 3D genre seemed to work best when the actors were throwing spears or other objects at the audience, the less intrusive curved screen technology appealed to epics like "Ben-Hur" and "The Robe". My point is that if we simulate our room on an appropriately curved surface, it could fool someone at five feet (but then probably not at ten). The scenario is always that someone comes around a corner and immediately encounters the illusion, of a room interior or a really big horse, or the ceiling of a matrimonial chamber. Only after the viewer is allowed to leave the encounter spot is the illusion susceptible to disclosure. Or am I wrong? Will the occularmotor skills give enough exploratory motion even at the micro level to dispel any closeby illusion?

The illusion will probably work for just an instant if it's up close, but I do think that stereopsis and occularmotor cues will detect that it's a photograph, even if it's curved. In fact, I think stereopsis may reveal that it's a photo even if the viewer doesn't move her eyes. (Unless she happens to be Lisa Eckstein).

Over the past month I have been three times in the chair of my ophtalmologist who is trying to fit me with some glasses that will enable me to spend the several hours each Sunday night reading the computer screen. When he gets me into that contraption that holds me while he tests lenses ("Which is better, this, or this?") I'll bet he could make be believe I was seeing anything he wanted to put over on me, and at any distance that pleased him. Am I wrong?

If he's providing separate stimuli to each eye, he can make you believe pretty much anything....

By the way, who teaches Cog Sci 77? Sooner or later someone over there is going to become interested in the perception problems that we face here in Math 8 all the time.

Cognitive Science 77- Vision: From Brain to Behavior is taught by Michael J. Tarr.

-david stanke