P3D Re: eye focusing by chromatic aberration

[roberts@xxxxxxxxxxxxxxxxx (John W Roberts)]

>Date: Sun, 18 Jan 1998 23:49:04 -0700
>From: Peter Abrahams <telscope@xxxxxxxxxx>
>Subject: P3D eye focusing by chromatic aberration

>>From: roberts@xxxxxxxxxxxxxxxxx (John W Roberts)
>>....- it has been reported on P3D that humans apparently
>>use chromatic aberration in the eye (when the colors in the scene make it
>>available) as a feedback mechanism to achieve focus.

>I believe that any color except black would be refracted by a simple lens
>into colors, no particular color would be needed.  Some minimum intensity,
>and some fairly sharp edged objects, might be needed.

I was thinking of situations such as monochromatic lighting, where there
aren't any different colors to divide up (e.g., a parking lot lit by low
pressure sodium lamps). In most cases multiple colors *would* be available.

>> In any event, the aberration degree and direction (if any)
>>are "frozen" in a still photograph.

By which I meant to say that in a still photograph, the point of minimum
aberration remains at a fixed "distance", instead of shifting, as it does
when we look at near and far things in a real scene. As with many of these
subtle focus issues, if the attention of the person viewing the photo can be
drawn to a particular depth, the fixed focus cues may actually enhance the
effectiveness of the photo.

>>Another possible issue - depth of focus as a function of brightness of
>>illumination of the scene [Note]. (Though if you're looking at the depth
>>that was in focus in the original photograph, it might match up.)

>There is something of great interest here, but I lost it by taking this out
>of context.   Can we have an elaboration?

I was thinking of an analogy to the "day for night" shots in older movies,
as "Lefty" describes them (thanks, Lefty). A photo could meet the main
criteria for realism, but a few minor inconsistencies set off alarms in the
various processing centers of the brain that make the picture a little less
believable.

In the case of the movie shot, most people probably don't consciously think
about the contrast ratio between light and shadow in moonlight, and many
people are not consciously aware that cumulus clouds generally disappear at
nightfall, but what they see doesn't match the pattern of what they've seen
before. (Unless they've watched a lot of those shots in movies. :-)

In the case of stereo photos, people have become accustomed to experiencing
a particular depth of focus  for particular lighting conditions. Particular
scenes are associated with particular lighting conditions. Since many cameras
have adjustments for both aperture and shutter speed, the depth of focus
for a particular photo might or might not match what a human viewer would
naturally come to expect.

I don't know whether any of these effects are significant enough to make a
difference. Experimentation is needed, and veteran stereo watchers are
probably poor test subjects:
 - If they still enjoy viewing, then these effects don't bother them,
     regardless of whether the effects might bother anyone else.
 - If some people are naturally bothered by these effects and some aren't,
     then the veteran viewers would be selected mainly from the group that
     aren't bothered by the effects.
 - If most people start out being bothered by these effects, then the veteran
     viewers have learned to ignore them, and might not be able to consciously
     evaluate them in the manner that a new viewer might.
 - The effects might really be insignificant for everybody, but having
     just veteran viewers wouldn't be enough to determine that.

On the other hand, novice viewers may have trouble expressing just what
they're experiencing. Perhaps some experiments with people who started
viewing stereo photos "not too long ago" would provide insights.

It would be a major undertaking to design a system that does adjust focus
in response to eye motions, and it's difficult to predict ahead of time
whether the effort would be worthwhile!

John R


------------------------------