Re: Contrast and Realists

[P3D Jim Crowell <crowell@xxxxxxxxxxxxxxx>]

At 10:15 PM 9/25/96, P3D John W Roberts wrote:
>
>Intuitively, a "hazy" filter should affect the modulation transfer function
>(MTF) of the optical system across the picture,

I think that's right--I seem to remember reading somewhere that haze & fog
have a pretty flat transfer function (as a function of spatial frequency).

> but the effect should be
>most noticeable near sharp boundaries with high contrast. It would add light
>to a shadowed area mostly around the outer edges, with the center part of the
>shadow relatively unaffected. Brighter areas would be somewhat darker around
>the edges than in the centers of the bright areas, but this might not be
>very noticeable to humans because of the relatively low sensitivity to
>transitions of low spatial frequency.
>

Actually, I think it's just the opposite.  Humans are much more sensitive
to low frequencies than high ones; the limitation is imposed by the optical
quality of the eye.  At _very_ low frequencies (< ~5 cycles/deg, depending
on the level of illumination) sensitivity drops again because of the way
signals are processed in the retina, but that effect is much smaller.  The
fuzziness is more noticeable at high frequencies because they're more
likely to be pushed down below threshold.

-Jim C.

----------------------
Jim Crowell
Division of Biology
216-76
Caltech
Pasadena, CA
(818) 395-8337
jim@xxxxxxxxxxxxxxx



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