How good are you at picking shades?

Here's a discussion that stemmed from people attempting to work out
'real world' colours from black and white pictures...

See what you think. Square A and B are exactly the same shade, but would
you have guessed it? Use the colour picker in photoshop if you dont
believe it... I was fooled!

http://www.network54.com/Hide/Forum/thread?forumid=47211&messageid=1075085896

Scott.

 

http://www.network54.com/Hide/Forum/thread?forumid=47211&messageid=1075085896

Hi Scott,

Yes, very interesting indeed. I certainly couldn't see it until I did as
others did and drew the line from A to B. Of course this has generated some
crunching in my limited grey matter about the implications of this and any
possible relationship with my photography.

Cheers,
David
P.S. I never did get back to you to thank you for your long dissertation on
making a pan setup. I'll have a good think about all that and let you know
how I go. Many thanks!

 

The guy who did that is so cool

has more demos here
http://www-bcs.mit.edu/people/adelson/illusions_demos.html

his main page is here
http://www-bcs.mit.edu/people/adelson/

 

It's nice to hear someone praise psychology !

 

Actually, the A and B squares are NOT the same shade. The A square is
black, whilst the B shares is white. The shadow cast onto the B square
doesn't change it's underlying colour, and your brain is smart enough to
realise this and thus report to you that they are a different colour.

If you look at the green clyinder your brain will probably report to you
that it is a constant colour, with the different shadings being caused by
the angle the light is hitting it. Obviously photoshop will disagree with
you as the left hand side of it is far darker than the right hand side...

Your brain doesn't simply report to you on what it sees, but it first
interprets that information for you into a more useful form. Hold two pens
in front of you, one twice as far away as the other. Your brain doesn't
tell you that the pens are different sizes (as they appear to be), but
instead it tell you (correctly) that they are the same size, only one is
closer than the other.

That said, cool illusion

Scott.

 

The right side of my brain agrees with you, but the left side
tells me that it's a 2D computer-generated graphic
(giving the illusion of 3D), and that no shadow actually exists,
just the illusion of a shadow.

 

Yep, and here's some more, for people who are interested more in the
theory of how they work:

http://www.colorcube.com/illusions/illusion.htm

Well, that one doesnt have much theory. The rest do though!

Here's some info on why they happen:

http://www.du.edu/~jcalvert/optics/illusion.htm

This page also talks about Poggendorff's illusion of the intersection of
diagonal and vertical lines:

"Poggendorff's illusion is very famous. Line 2 is actually
the continuation of the line on the left, although line 1
appears to be. This illusion is counteracted in the British
Union Flag by displacing the arms of St. Patrick's cross
on either side of St. George's cross so they appear to be
in the same line. Greek temples were designed with
deliberate distortions to make the building appear
correctly. Columns were given entasis, a slight swelling
in the middle, so they would look straight, and architraves
were cambered up slightly in the center so they would appear
straight"

Apparently, they still have no explanation for why that works...

Here's some more of them, along with a PDF paper about it.

http://persci.mit.edu:16080/gaz/

 

Ok, with my sciency objectiveness hat on....

....Yes, they are!

There is no shadow! Its the picture that is in question here, not the
'original scene'... but that said, I know what you mean

That doesn't change the fact that it is NOT the same colour all the way
around. Your brain is tricking you into thinking that it is the same
colour, because it is used to seeing it that way. If it were a real
object, then it sould be the same colour. But it's a picture. The same
as in a photo - the colour is defined by the pigments, not by the real
world.

No, that's not right. Your brain distorts the physical reality.

Shut one eye and it makes it harder But this is a different issue
from the colour perception.

True... and although your brain fools you into thinking the colours
arent the same, you'd have a hard time if it didnt work that way

By the way, apparently, some painters are able to see the colours for
what they are, rather than what most people's brains tell them it
'should' be.

Scott.

 

Yes - and one thing I find interesting is that, even though part of your
brain 'knows' that there's no shadow there, you still can't convince
yourself that the colours are the same... This is different from the
'ambiguous figure' illusion. Do you know the 3D representation of three
faces of a cube, drawn so that you're looking directly at the point of
one of the corners? You can think about it so that you're looking at it
as though it from the inside or the outside, if you know what I mean.
This makes the 3D figure look 'sunk in' or 'raised up' from the paper. I
hope that makes sense. For that one, you can convince yourself that it
goes either way.

Scott.

 

The concept of a shadow doesnt alter the mechanism.

 

If there is no shadow, then there are no "squares". If you're looking
at a 3D image, then you can consider the 4-sided boxes to be squares, the
green lines-and-a-circle-thing to be a cylinder, and the colour difference
to be from a shadow.
If you're looking at a 2D image then there are no squares, there is no
cylinder, and there is no shadow. You can't have it both ways

Regardless of what there is, your brain sees squares, it sees a cylinder,
and it sees a shadow. If not, why did you call them squares?
Photoshop has no brain (as proven by what it does when you select
Auto-levels), and thus sees a 2D object with no difference in colour between
the A area and the B area.

Scott.

 

Certainly... but the shapes of them dont matter. In fact, you're right,
they're not squares.

Again, it doesnt matter what shape they are.

What defines the colour of those shapes? I define it physically (it
doesnt have to be defined this way, but that's just how I choose to
define it). It's the wavelength of light emitted from the object being
viewed. You're viewing it on your screen, and the screen is shining the
same coloured light at you when you look at both 'shapes'. It's the same
colour!

Anyway, we're arguing the same point here! Really, our discussion is
about how we're defining the 'colour' of the picture. I'm saying that
it's colour is decided by the picture.

Scott.

 

Colour is and always will be relative. In addition to which you can "see"
the difference between two green objects and swear they are the same hue,
till you see one of the objects next to a darker strip. Then you may set the
two green hues apart.

 

Still, that's in agreement with what we're saying. Still, it depends on
how 'colour' is defined. They'r only relative because our brains modify
what is actually being detected and give us a 'faulty' version of what
is really there. Faulty, but nonetheless useful most often.


Scott.