PDA

View Full Version : Colour perception test


DrFx
05-11-2005, 02:05 PM
Saw this in another forum:
http://www.echalk.co.uk/amusements/OpticalIllusions/colourPerception/colourPerception.html
(http://www.echalk.co.uk/amusements/OpticalIllusions/colourPerception/colourPerception.html)
It shows how your colour perception is affected by the background of stuff. It really is amazing, try it!:thumbsup:

paperclip
05-11-2005, 02:10 PM
That is pretty crazy. I've seen things like that before but it gets me every time.

I wonder if people with more developed color sense can 'see' this right away?

stewartjones
05-11-2005, 02:26 PM
That is mega cool! :thumbsup:

offbeatworlds
05-11-2005, 03:38 PM
Wow, that was amazing. Very cool like thingy there.

jmBoekestein
05-11-2005, 04:03 PM
I'm more amazed at the fact that we are fooled by a 8-bit value range, are you going to tell me it never crossed your mind how many times an image had recurring colours after the artist has carefully thought up a colourplan or scheme.

We're talking 256 differeent greyscale intensities, that's all man.

DrFx
05-11-2005, 05:14 PM
The human senses are pretty limited: Put those greyscale levels side by side in a gradient, can you tell the difference between them? No, it'll be one continuous gradient!
Also, can you tell the difference between 8-bit, 16-bit and 24-bit sound, coming from average speakers? Or between a good jpg and a tiff? A CD and an MP3?

jmBoekestein
05-11-2005, 08:03 PM
Yeah actually most of the time or a lot of the time, also depends on the environmental noise and levels imposed on my senses, since they're more imporant to respond to.

But I do agree that without watching for the subtleties it could easily ellude me.

_slvl
05-11-2005, 08:07 PM
We see the colors that way because we are constantly trying to correct the colors so they seem right to our brains.

A nice example would be Monet (french painter). He develpoped an eye disease which made him see things more red. He used to paint his garden and as the disease developed his painting went from normal green colors to reds. His brains were trying to correct the colors.
(edit: a link about this subject: http://www.intermonet.com/colors/index.htm )

That's why eye "malfunctions" are mostly discoverd fairly late, because we can't notice the degrading eyesight. The brain corrects most of the flaws by filling in the gaps.

jbo
05-11-2005, 08:24 PM
can you tell the difference between 8-bit, 16-bit and 24-bit sound, coming from average speakers?

hell yes i can. if you can't hear the difference between 8 and 16, there's something seriously wrong with your ears or speakers. 24-bit sound is often hard to distiguish from 16, but it's great to record in 24 or 32 bit, because you have plenty of headroom when editing the sound, so it won't degrade past the point where a human would notice... the same reason hdr images are usefull even though you can't actually see them on your monitor.

jmBoekestein
05-11-2005, 08:25 PM
also, in the last one it surprised me that it was denoted as a grey colour I thought for an iontstant that it really was blue. Would have made more sense to me actually. ANybody else saw a blue region there? Or am I a freak?

Tocpe
05-11-2005, 08:46 PM
hmmm on the third one I was seeing blue and yellow not grey and yellow. What's that mean? Am I broken? ;)

glassefx
05-11-2005, 08:48 PM
Thats amazing... Thanks for sharing... As for sound... I can tell the difference from compressed MP3s at 256 kbs comparred to the original... Me and my old roomy were pretty toasty one night and he said he could tell the difference... I have a nice system... Maybe thats the difference...I have KEF 104.5 studio monitors with audioquest midnight hyperlitz cables...He was correct... You can put your fingers GENTLY on the bass-cones which are 12" and the Mp3's bass is more sloppy... the in/out of the original recording is tighter. One really has to hone their perceptions but since the optic system is WAY more esoteric than hearing... The tests just work on everyone... I guess?

SkyZero
05-11-2005, 08:56 PM
also, in the last one it surprised me that it was denoted as a grey colour I thought for an iontstant that it really was blue. Would have made more sense to me actually. ANybody else saw a blue region there? Or am I a freak?

i saw blue in there too, i'm gonna go exhange my eyes for new ones... :eek:

Lunatique
05-12-2005, 02:00 AM
This is an excellent example of how colors are relative, and questions like "What's a good skin tone color to use?" are completely futile.

Ordibble-Plop
05-12-2005, 03:01 AM
If anyone is interested in scientific theories of image perception then this article on Image segmentation and light perception (http://www.nature.com/nature/journal/v434/n7029/full/nature03271.html) is applicable. It doesn't deal with colour but has similar optical illusions to the greyscale ones linked to above and a couple of videos.

I recently edited a paper that summarised the one linked above and it mentioned colour:

"The layers concept exemplifies a computational strategy known as inverse optics. The intensity at each point in the image is the product of a combination of factors: the proportion of light reflected by the surface at that location (called reflectance), the
intensity of illumination incident on that surface, and certain properties of the intervening
media, such as those of fog or filters. By the laws of optics these factors become entangled in the image. In principle they can be disentangled by hypothetical brain
processes that are inversely related to the optics of entanglement.

For example, a red book on the dashboard of your car casts a red reflection in the windshield. Through the reflection you perceive distant objects, including
green grass, in their normal colors. Light from the green grass and the red reflection physically mix to produce yellow. The yellow is observed when seen through a
small hole punched in a piece of cardboard held up so it blocks out the surrounding context. Without the cardboard, however, no yellow is seen, only the red and green
layers. The brain is thought to split the yellow light into the red and green layers using rules that invert the usual rules of color mixing. This is called scission."

Gord-MacDonald
05-13-2005, 04:39 AM
If anyone is interested in scientific theories of image perception then this article on Image segmentation and light perception (http://www.nature.com/nature/journal/v434/n7029/full/nature03271.html) is applicable. It doesn't deal with colour but has similar optical illusions to the greyscale ones linked to above and a couple of videos.

I recently edited a paper that summarised the one linked above and it mentioned colour:

"The layers concept exemplifies a computational strategy known as inverse optics. The intensity at each point in the image is the product of a combination of factors: the proportion of light reflected by the surface at that location (called reflectance), the
intensity of illumination incident on that surface, and certain properties of the intervening
media, such as those of fog or filters. By the laws of optics these factors become entangled in the image. In principle they can be disentangled by hypothetical brain
processes that are inversely related to the optics of entanglement.

For example, a red book on the dashboard of your car casts a red reflection in the windshield. Through the reflection you perceive distant objects, including
green grass, in their normal colors. Light from the green grass and the red reflection physically mix to produce yellow. The yellow is observed when seen through a
small hole punched in a piece of cardboard held up so it blocks out the surrounding context. Without the cardboard, however, no yellow is seen, only the red and green
layers. The brain is thought to split the yellow light into the red and green layers using rules that invert the usual rules of color mixing. This is called scission."

Interesting stuff! I will *try* to read the articles after Master Servant

Gord

Jukebox
05-13-2005, 11:08 AM
to be honest i didn't get number 2
number 1 was trippy......
pretty crazy..

jmBoekestein
05-13-2005, 04:34 PM
Well...:wise:...they're not jokes actually.

The second one illustrates how contrast increase has your rbain thinking that it should "reset" 50% grey in your head. At least that's how I get it.:)

alexyork
05-13-2005, 04:53 PM
This is an excellent example of how colors are relative, and questions like "What's a good skin tone color to use?" are completely futile.

Absolutely! I've always maintained that colour choices are entirely relative. for instance, it is perfectly possible to produce a shot with a bright blue apple in it. in theory our brain tells us that a blue apple is weird/wrong/incorrect, but with colours in the surrounding space that work properly, relative to the blue apple, the entire shot can work perfectly (if a little weirdly). The movie 'Hero' is a very good example of this.

This stuff is fascinating to me.

paperclip
05-13-2005, 05:22 PM
If anyone can produce a picture with a blue apple that looks red, I would be amazed. Not that I don't believe that it can't be done, just...it sounds so unreal.

alexyork
05-13-2005, 05:56 PM
If anyone can produce a picture with a blue apple that looks red, I would be amazed. Not that I don't believe that it can't be done, just...it sounds so unreal.

it's not that the apple looks red - it's more that the image works as a whole, because the colour relationships work regardless of the subject matter.

DrFx
05-13-2005, 06:38 PM
If anyone can produce a picture with a blue apple that looks red, I would be amazed. Not that I don't believe that it can't be done, just...it sounds so unreal.

Ok, I've tried it, but it's far from perfect. But I think I got the trick, it works better if the surrounding colour is the opposite of (in this case) red, and perhaps if you have got some colour variance throughout the scene (which is not the case). Oh, well 10 min. wasted precious labor time! :applause:

alexyork
05-13-2005, 07:07 PM
woah! my point has been totally taken out of context haha! interesting challange tho! far more interesting than my original point... :scream:

carry on!

paperclip
05-13-2005, 09:27 PM
It looks more like a plum to me....let me see if I can do anything. Watch this space!(JUST the colors, remember, not the actual apple needs to be worked on, can you make a blue look red just by combining other colors? I bet Linda could do it!)

jmBoekestein
05-13-2005, 09:50 PM
Why don't we all give it a shot? Seems a lot of fun. I'm going to try too. :wip:

alexyork
05-13-2005, 11:29 PM
You know... if this thing gets popular, it might be pretty cool to see if we can come up with some mini challanges along the same lines.

I might have accidentally started a potential revolution! :D

Hey, I guess we should wait and see what people post, and if this thing really works 1st :)

paperclip
05-13-2005, 11:37 PM
This is the closest I can get (for now :twisted: )

http://img50.echo.cx/img50/6708/blueintored1jc.jpg

umbrellasky
05-13-2005, 11:41 PM
Pretty cool. I've seen one of these before amazing how easily we can be fooled, guess the whole neutral and chromatic grays thing works in the same way.

weird...

jmBoekestein
05-14-2005, 12:32 AM
http://img66.echo.cx/img66/344/yellabastard1lj.jpg (http://www.imageshack.us)

'nuff said, but I'll keep trying. I went over the yellow border a wee bit. I cheated.:shrug:This really is a good learning experience actually.

leeyiankun
05-14-2005, 07:22 AM
You've heard of the term that 'it's all relative' right?
That is how human find themselves.
How far apart we are from other ppl/things.
How can you tell a position of an object?

You tell it by the co-ordiated space in relations to others.

I think one green apple in a lone universe has no color at all actually.
Because you'll have nothing in which to judge it by.

As for this wonderful example of color perception?
Well, this is a good demonstration how we humans base are good at adapting.

Lets try and think of a system that human use exact color to percieve by.
And how confusing it is going to give us.

Hmm... That's an interesting thought indeed.

DrFx
05-16-2005, 09:28 AM
Well, I think the next CG Challenge should be "The Blue Red Apple"! What do you think? :)

GOTgraphic
05-17-2005, 03:23 AM
These kinds of brain-teasers blow my mind.

I've heard of people being raised in a primitive setting in jungles and then they take these individuals into cities and they can't get around because their perception of depth is non-existent. So one day I was in a parking lot of a grocery store that was uphill from where I parked. I concentrated on the scene and worked at imagining the whole thing as a flat surface and I was merely climbing it to arrive at the doors of the store. Suddenly the whole thing flattened out and I had to stop walking because it was freaky looking... the brain is POWERFUL! I tried the same thing while riding a bus to work one morning and it can be repeated. I haven’t done that again though… its too crazy… I like my perception of depth :D

paperclip
05-17-2005, 09:56 AM
Freaky!

I guess that is what we see after all, a 2d image- the further away, the higher up it gets (if you're on the ground obviously). Strange how we learn to intepret that so flawleslly.

The miracle of God, eh?:)

Son_of_Skeletor
05-17-2005, 01:48 PM
Wow, these brainteasers hurt my head.

It boggles my mind that our brains view colour this way.

mr3dguy
05-18-2005, 03:36 AM
Relative Contrast


After reading this thread I was inspired to try something I've wanted to try for awhile now.
What if you increased the contrast of an image with each pixel being relative to its surrounding pixels. So that a gray pixel surrounded by black becomes white, but surrounded by white it becomes black.


Here is my test image
http://members.iinet.net.au/~mr3dguy/pics/RC.jpg
And the result
http://members.iinet.net.au/~mr3dguy/pics/RC-result.jpg

First I took the original image and made a blurred copy of it (to get each pixels surrounding pixels)
Then I subtracted the blurred image from the original image (results need to be unclamped, values below 0 darken the image whilst values about lighten it).
In order to tone down the effect a bit the results may be divided at this stage.

Then I added the original to the results of the subtraction.

Here is an example on a photo

Original
http://members.iinet.net.au/~mr3dguy/pics/The_Rose_by_Catw0man.jpg
Result
http://members.iinet.net.au/~mr3dguy/pics/The_Rose_by_Catw0man-result2.jpg

This does a bit more than just changing the contrast in photoshop, because it works relatively it brings out details, where contrasting just adjusts the whole image.
Bringing out details can be bad in a lot of cases, for one it makes highly compressed jpegs look even worse and for two not many people like having their pimples standing out from the rest of their face in photos.
This process works to exaggerate the way the eye views differences in lightness/darkness relative to their surroundings.
This is may be available in some software packages I am not aware of. Or even photoshop.

RonB
05-18-2005, 04:13 AM
The 3rd example of the color test is not correct...those two sword center part color values ARE different. To check that for yourself, do a screen capture of the image and check them with the Eye Dropper sample tool in Photoshop, I used a 5 X 5 sample rate. The others are pretty cool though as far as relative perception of tone goes...neat stuff...

dogyears
05-18-2005, 05:54 AM
Check out

http://en.wikipedia.org/wiki/Color_constancy

Basically the fact that we still see "colours" in a variety of different lighting condition in the real world suggests that "percieved" colour and "real" color can be radiacally different. This probabily won't come as a surprise to us as artists - but what the theory suggests here is the fact that the images appear green is because there is a white border ...

Do not imply, however, that this means "all colour is relative", percieved colour can be relative - but true colours, their wavelengths, still exist ... colours are absolute, so to speak, relative to white.

KayosIII
05-18-2005, 11:05 AM
I used to I had my desktop in dark blue... Then I switched back to a neutral dark grey. It took a few weeks till I wasn't seeing yellow - Scarey huh.... Neutral colours from now on

CGTalk Moderation
05-18-2005, 11:05 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.