MOON ILLUSION: Illusion claimed to exist that moon near
or on the horizon looks "larger" than the moon we recall seeing
high up in the night sky on other occasions.
This illusion has been reported for thousands of years by various
cultures. Usually attributed to clues or perspective from objects seen
on horizon [causing a realization moon is further away than we thought
-- therefore, to be that size of a circle in our eye and be that far away,
it must be larger than we thought.] The realization causes us to "see
"what we realize, or a "believing is seeing" illusion.
The illusion above (which I modified from the Ponzo illusion, shown
later below) seems to work (mildly) based on the amount of area surrounding
the circle -- the more area, the smaller the circle will seem by comparison.
The zenith moon may seem small compared to the vast expanse of sky area
surrouding it..
But when I substitute moons in place of the circles, then this is the
view:
The illusion above no longer seems to work. Both moons seem equal sized.
No reason for it comes to mind. So I decided to go back to the original
Ponzo illusion (just below).
Now here (above) the illusion works -- again only mildly, especially
when I think of the converging lines as being just the letter "V"
that fell over. On that "flat" purely geometric basis, the circles
look practically equal in size. The illusion will pop up, tho', and I think
that's because the eye has trouble completely avoiding the converging lines
as a perspective or distance clue, out of habit or visual programming.
But again, when I substitute moons for the circles, then it looks like
this:
Well, now!! It seems that moons here also do NOT create
much of an illusion, or even any illusion, for some reason, but probably
because I'm still seeing the converging lines as a fallen-over letter "V."
But if I look leftward directly at the "V" and
not the moon, and make myself see it as a perspective thing
going to a vanishing point -- namely, a distance clue -- then the
nearest periphreal moon does look a bit larger. Not as easily
seen as the geometric Ponzo circle picture above this pic, but the illusion
is sometimes there (mildly).
But now I think, what if I INCREASE the distance clues
-- what if I pictorially FORCE the converging lines to be
a perspective or distance clue totally unable to be perceived as just
a flat fallen-over "V"?
After I do that, then the illusion should become unquestionably
apparent (scroll down):
Now the illusion is inescapable. Two moons stuck to a stone wall, one
clearly looking "larger" than the other because it is now
strongly interpreted being on A FURTHER-AWAY part of the wall.
Once again, I have used only a distance clue to create an illusion, or
create a stronger illusion than when objects and lines are
ambiguous in their distance interpretations. When ambiguity about
distance is removed, the illusion doesn't just pop in and out (and
at that, only mildly illusory), but now becomes stable, and even
appears "bigger" in the size of it that is perceived.
The more that distance is clearly established, the greater size there
is to the illusion. The more cues used for distance, the more the illusion
appears stronger.
Furthermore, the fact that convergence has been turned sideways (unlike
the railroad track illusion which is as used in the illustrations below),
this does not necessarily prevent one from still seeing the illusions.
It may take a few more seconds to kick in, but the distance cues can
even be upside down and still create the illusion of one moon or object
being larger than another (even tho' drawn the same diameters).
Below, two pics of the moon illusion are created using two methods
of cueing distance to the eye:
In real life, we don't see two moons simultaneoulsy, but see
a horizon moon and a zenith moon, one from memory, the other in view. There
may actually be no "moon illusion" at all. It could be that the
two sightings (horizon and zenith) are independent and free of any illusion,
but because of different cues and processes happening in each sighting,
the feeling of how large the moon is in each experience usually gives rise,
when compared, to believing the horizon moon is larger than
the zenith moon.
I think the Ponzo illusion is NOT clued solely by the converging lines
nor by perspective clues. It is clued by the fact that there is a comparison
made by the eye(s) of the area around the circle. We compare the surrounding
area to the circle's area. Since the ratio or percentage of areas when
compared changes, then, simply, this occurs: The more area around the circle
(or a moon) then the smaller it looks by comparison. See the attached drawing
I made as evidence of this at top. This may play a role in the zenith moon,
as the area of empty sky overwhelms the moon's circle within it, and makes
it look smaller (by recall, when we later look at the horizon moon).
THEN THIS HAPPENS, and this includes a NEW way of analyzing
this whole matter:
The horizon moon already looks absolutely "huge"
when compared to horizon objects and references, and using recall of the
zenith moon is NOT NEEDED for the horizon moon to look "huge"
once it gets compared to RR tracks, buildings, trees, skylines, etc.
But the mind will then turn to another exercise: Now, we recall the
small-looking zenith moon which is now mentally compared to the already
HUGE-looking horizon moon. In this comparison, one can easily understand
why the huge horizon moon therefore looks bigger. And also why it looks
"closer" (because it looks "larger" due to its look
of hugeness). One might think: "Since it's the same moon, it must
have moved closer to the earth in order to appear larger.")
All this occurs despite what (only appears) to be contradictory, in
that it was the realization that the horizon moon is so far away
[which was clued by being over or behind the horizon and its familiar objects],
that made it look so huge. (We need to use the absolute word:
"Huge," not the comparative word: "largER.")
So, while being seen as "far away" being what
makes it look huge; yet this hugeness, now seeming larger than our recall
of the zenith moon, makes it also look "closer" to
us than the zenith moon. Both are true.
The horizon moon appears huge because reference clues tell us it's
very far away, and also perhaps further away than we thought
or recall, yet it also looks "closer" than (only)
the recalled zenith moon, because hugeness also makes it seem larger
than the zenith moon.
In my railroad picture, the greater the distance clues that are present,
the more huge the moon looks. The eye translates that into being "larger"
than the recalled zenith moon, which is seen comparatively smaller by its
vast surrounding sky-area. Flip the picture over, and the distance clues
"fail to compute" (initially) due to lack of familarity when
seen upside down, and the hugeness disappears for a short time. Both moons
then look equivalent.
Note: My version of the railroad illusion doesn't place anything inside
the converging lines. Instead only distance cues are used, namely:
1. One moon is placed near the bottom of the picture, wholly under
the horizon line, as there it will be interpreted as closer.
2. The lower moon also casts a shadow, placing it "in the picture"
(rather than floating) so that it is clearly closest.
3. The other "horizon" moon is partly above the horizon,
where it can be interpreted as the furthest away object in the picture.
4. This upper moon is behind the horizon, also placing it "in
the picture" [rather than floating], so it is clearly furthest from
us.
5. The upper moon is less dark (the "graying" effect from
intercepting amounts of air between eye and object). This is also true
in how I make the RR tracks "fade" in details they converge.
6. The upper moon is less sharp or less detailed, while the nearer
one is more detailed, imitating the effect of focus in determining distance
(as both near and far object cannot both be in focus, usually, therefore
one must be further than the other).
In order to test if the Ponzo illusion plays a role (based only on
spacial comparisons rather than distance clues), I have drawn the second
illustration of the illusion using the sea, rather than the railroad tracks.
I removed the railroad tracks. No converging lines anymore.
No perspective lines, nor vanishing points, anymore.
I removed all Ponzo attributes from the horizon moon (being near to
reference material or objects) -- and gave them to the lower moon instead.
I 'Ponzo'd' the lower moon by putting a box around it, and there is
immense busyness surrounding the lower moon to see if being surrounded
by reference objects will "confine" it and make it seem larger
(like the same am't of food on a smaller plate). Thus there is no vast
empty "space" around it to overwhelm it and make it look small.
All objects (footprints in the sand) etc., are smaller than the lower moon.
Nothing left but familiar distance clues. The upper moon
has nothing but the horizon line near it. More than any object, it has
lots of space around it, and what's more, it isn't even all there,
as part is below the horizon. But despite all this the illusion is still
there that the horizon moon appears larger than the lower-shown moon.
Why?
Because the distance cues (creating a knowledge of how much
distance there is to the moon) trump everything else, even the spacial
or "boxed" modification of the Ponzo effect, enough to still
trigger the eye, tricking it into thinking the retinal or "apparent
size" is larger for the horizon moon.
"Knowledge is power" -- or should I say "believeing
is seeing?" -- as in the "drive-in movie effect"
where knowledge of where the sound "should"
come from of an actor's talking -- will soon trigger, or subconsciously
instruct, the ear to hear it "coming" from the face on the movie
screen even though it really comes from a speaker brought into the car
and hung in the back seat or off to one side.
