[net.ai] Visual After-effects

fostel@ncsu.UUCP (06/03/83)

    The effects described such as the railroad track and video after
    effects are well know by psychologists, and indeed are one of the
    tools used to study the levels and types of processing in the optic
    system. Most introductory texts on the subject will include a few
    pictures to stare at in certain ways to acheive some of types of
    after effects you noted.  I beleive Scientific American even gave
    away a resubscription freebie on the subject a few (6?) years ago.

    The earliest description of the phenomenon I know of (circa 1910)
    by a reputable psychologist was from a fellow who had a small area
    of his retna with a blind spot.  (Was this Lashley?) He observed once
    at a party, that when a person stood against a highly regular wallpaper
    and their face was in his "spot" their head would be "removed" and
    replaced by the Wallpaper Pattern!  The visual system was simply
    making its best guess of what should be simulated for those bad
    receptors.  A bit of experimenting later, it was shown that the effect
    could be reproduced with anyone by simply fatiguing the receptors
    at one spot (simulating a defect) by staring intently at one object
    without blinking, moving the head or sacading the eyes.  If the level
    of fatigue is great enough and the background suitably benighn and
    predictable, the object stared at will indeed disapear, actually being
    replaced by the visual systems best guess for what the fatigued cells
    would report if they were sending out a better signal.

    My own experience with video games provides some confirmation of the
    "modern" experience.  I play Robotron, occassionally for several hours
    (takes a while to recycle the 9,999,999 score) which involes LOTS of
    little glowing things moving about, some of which must be avoided and shot, and
    some of which must be "rescur
    rescued".  (Sorry, key-board burped.)  After such a binge, I will see
    afterimages of the little Good guys I must recue, but never the bad killer
    robots.  Now THAT is a high level of processing in the optic system:
    it seems to be able to tell good from bad!!
    ----GaryFostel----

mmt@dciem.UUCP (Martin Taylor) (06/09/83)

Actually, the blind-spot game of removing people's heads has a
long history. King Charles II of England used to amuse himself by
seeing how his courtiers would look without their heads. And it is
true that any regular pattern behind will be filled in across either
the normal blind spot or blind spots due to retinal problems.

As for the effect in which objects tend to disappear if stared at,
this is normally studied with special devices attached to the eyeball
(on a contact lens) to ensure that the visual world remains stationary
on the eye. Objects rapidly vanish under these conditions, but reappear
in fragmentary form from time to time. Very slight shifts of viewpoint
tend to make the objects come back, which is probably the reason attending
to a detail "behind" the object makes it return. It is easier to
make things with blurred or diffuse edges go away than things with
sharp edges (so I imagine people with poor eyesight can do it easier
than people with good vision).

The effect of changing letter size after watching for game objects
that change size is another example of the same kind of thing as
the railroad track after-movement effect. It's probably a different
visual channel (we have separate channels for size changes and for
movement) but the principle is the same. Some people claim that the
effect is due to fatigue of the system sensitive to movement in
one direction, leaving the balancing components sensitive to movement
in the other direction to control what is seen when the stimulation
is neutral. (i.e. the other direction is more sensitive after one
is fatigued). I'm not convinced by this explanation. Things are
probably more complicated than that.