Contrast Between Different Colors

[Bacle2]

Hi, All:

My cable company recently changed its color setup, and I'm having trouble distinguishing

information from the background. I was wondering if there is some scientific basis for

my preferences, and to explain my problems. Specifically, e.g., the timer/clock used to

consist of white numbers in a black background. I had no trouble making up the numbers

from the background. Now the numbers and letters are still in white, but the background

is sometimes blue, sometimes light green, and I'm now having trouble distinguishing

the letters/numbers from the background. My theory (with little basis) is that a pair of colors that

are farther appart in the spectrum are easier to distinguish than a pair that is closer. Is this

correct?

 

[L4xord]

Hi, All:
My theory (with little basis) is that a pair of colors that

are farther appart in the spectrum are easier to distinguish than a pair that is closer. Is this

correct?

Yes... I'm pretty sure that's what complementary colours are.

 

[ZapperZ]

I have moved this out of the physics forum. Take note that light detectors that can detect such a spectrum have no difficulties in making that distinction. So this has more to do with the human eye perception. That is why the question is more relevant in the biology forum than in the physics forum.

Zz.

 

[A.T.]

My theory (with little basis) is that a pair of colors that are farther appart in the spectrum are easier to distinguish than a pair that is closer.

Brightness differences are easier to distinguish than hue differences, even for people which are not color blind. A good screen designer always tests if his color scheme works well in grayscale mode too. This ensures that it doesn't relay on hue differences to create contrasts.

 

[Andy Resnick]

<snip> My theory (with little basis) is that a pair of colors that

are farther appart in the spectrum are easier to distinguish than a pair that is closer. Is this

correct?

Color vision appears to use the principle of 'opponent coding': that is, not only are there 'direct' RGB outputs from the individual cones, but there are also neurons that divide one signal from another. Red-green color opponency has been localized to the parvocellular layers within the lateral geniculate nucleus, but blue-yellow color opponency appears to occur with its own special pathway (the koniocellular pathway), suggesting that red-green and blue-yellow vision systems developed independently.

Visual thresholds for color has been studied much than brightness thresholds, there is not a lot of data out there.

 

[pumila]

Original black/white colours have high brightness ratio. In human perception brightness is dominated by the red-green colour range more than the blue. If you switch from white/black to white/light green then while for white to black you have a brightness ratio of 1 to 0, to light green you have a ratio of 1 to perhaps 0.75, light green being white with (say) 50% red and 50% blue removed; the blue does not count towards brightness so the perceived brightness is 50% red plus 100% green, a mean of 75%.
That would drop discrimination by a factor of four.

 

[sWozzAres]

Color vision appears to use the principle of 'opponent coding': that is, not only are there 'direct' RGB outputs from the individual cones, but there are also neurons that divide one signal from another. Red-green color opponency has been localized to the parvocellular layers within the lateral geniculate nucleus, but blue-yellow color opponency appears to occur with its own special pathway (the koniocellular pathway), suggesting that red-green and blue-yellow vision systems developed independently.

Interesting since that fits in with evolution of colour vision where the mammals that ran about during the age of dinosaurs, lost colour vision due to being noctural. They then regained colour vision as dichromats. Most mammals are still dichromats but catarrhine apes and Old World monkeys are trichromatic suggesting they regained 3-colour vision in two separate events.