WEBVTT

00:01.500 --> 00:03.700
>> NARRATOR
Should you trust your vision?

00:04.300 --> 00:06.600
Professor Daniel Levin at Vanderbilt University

00:06.700 --> 00:09.300
has a unique test that might make you skeptical.

00:09.400 --> 00:13.000
>> DANIEL LEVIN
What you do is look at a face

00:13.100 --> 00:16.755
and then attempt to match a little square

00:16.855 --> 00:20.984
with the darkness or lightness that you see in the face.

00:21.500 --> 00:24.050
>> NARRATOR
To do this, you need to come up with a sense of the average

00:24.100 --> 00:26.200
or overall brightness you see in the face.

00:26.600 --> 00:29.500
It sounds simple, but it's harder than you'd think.

00:29.800 --> 00:31.500
Try it for yourself.

00:31.461 --> 00:34.200
Adjust the slider bar until the brightness of this square

00:34.300 --> 00:37.211
matches the overall shade of gray you see in the face.

00:40350 --> 00:43.100
Did you choose the same values for both faces?

00:43.200 --> 00:45.200
You should have -- because these two faces

00:45.300 --> 00:47.300
are the exact same shade of gray.

00:47.900 --> 00:50.400
Their overall brightness values are identical.

00:51.000 --> 00:53.800
Don't believe it? Most people don't at first.

00:53.900 --> 00:56.900
>> DANIEL LEVIN
I put a lot of effort into making sure the mean

00:57.100 --> 01:00.900
brightness value of the image files was the same,

01:01.000 --> 01:03.900
making sure that the mean contrast was the same,

01:04.000 --> 01:06.000
making sure that the number of pixels in the image

01:06.100 --> 01:07.100
was very similar...

01:08.250 --> 01:10.000
So what's going on here?

01:10.250 --> 01:12.850
Watch our video on Illusions at Work to learn the science

01:12.950 --> 01:15.550
behind the illusion, and how our misperceptions can

01:15.650 --> 01:19.600
color evaluations that go far beyond shades of gray.