If I have a truly yellow object, one that reflects only the light with wavelength between 570-590 nm, and photograph it with a camera that detects only RGB to put it together as a color image, would the object in question appear black? As it doesn’t reflect any truly Red or Green light. If that is be the case, aren’t pictures taken with regular cameras showing a very distorted version of what we would actually see? And, is the reason yellow objects appear yellow in the processed picture the fact that they all reflect red and green light?
$\begingroup$
$\endgroup$
Add a comment
|
3 Answers
$\begingroup$
$\endgroup$
It depends on the properties of the RGB sensors (or filters) in the camera. Human retina cells, for example actually detect spectra that overlap. So, 580 nm light is sensed by two different types of retinal cells. Many color vision systems are designed to have the same R, G, and B spectral sensitivity as the typical human eye.
answered Nov 23, 2019 at 11:44
$\begingroup$
$\endgroup$
If the cells in your eye and the sensors in your camera only registered "true" red, "true" green, and "true" blue, and their combined frequency response consisted of three sharp spikes, then yes, we wouldn't be able to see or photograph "true" yellow - it'd seem black. Or rather a sort of grey, because we also have "monochrome" sensors on the back of our eyeballs that register light across the full range of human-seeable light.
The monochrome cells are more sensitive than the R, G, B colour cells, which is why when we look at things in low light, at night, most of the colour seems to disappear. It's still there, we just can't register it unless it's sufficiently bright.
In reality, the red, green and blue cells in our retinas each respond to a band of frequencies, centred on R, G and B, and as S. McGrew says, these bands overlap. If a color triggers the red cells a lot and the green cells a little, we see orange. If it registers the red cells a bit less and the green cells a bit more, we see it as yellow, a little bit more and we see yellowy-green. If it registers on the red and blue cells, but NOT the green, we see some sort of purpley colour.
This makes it difficult to tell the difference between true colours of a single wavelength and clusters of nearby colours that average out at that wavelength - we find it difficult to tell the difference between pure yellow, a mix of red and green, and a mix of orange and yellowy-green.
On the plus side, this disability allows us to fool the human eye into thinking it's seeing a fairly complete recorded spectrum of natural colours just by using cameras with R, G, B sensors and TV screens with R, G, B LEDs. The eye can probably do slightly better than this by comparing the R,G,B and monochrome signals and subtracting one from the other ... this is probably why normal screens can find it impossible to replicate impressions of some very complex colours, such as complex browns ... but for most purposes, RGB monitors do a good job of fooling us.
$\begingroup$
$\endgroup$
3
RGB is special because red, blue and green are the primary colors. Yellow is a secondary color which is composed of green and red with varying proportions. So basically that yellow object is reflecting both red and green colors from the white spectrum of light incident on it. That's why the object appears to be of yellow color.
answered Nov 23, 2019 at 11:43
-
2$\begingroup$ This is incorrect, yellow is a wavelength, so you can have a pure yellow object. $\endgroup$– JavierNov 23, 2019 at 15:45
-
$\begingroup$ @Javier Yellow is a wavelength, no doubt. Any object which emits or reflects radiation in the range 580-600 nm will appear yellow. But for a camera which processes its image using RGB pixels, the contents of a pixel can have varying proportions of red, green and blue only. So, for that camera, there is no "pure" yellow state. Crudely speaking, that yellow state is a superposition of red, green and blue colours. In this case, it is red and green. So, in spite of yellow being a pure wavelength, a RGB camera can detect its color only in terms of R,G and B. That's what I have mentioned here. $\endgroup$ Nov 23, 2019 at 22:28
-
$\begingroup$ @SchrodingersCat But to be picky that is not what you mentioned. From the question it is obvious that OP understands yellow can be a combination of green and red but asks for the case where yellow is a pure yellow wavelength. This would depend on the spectral width of the detectors. $\endgroup$ Dec 9, 2019 at 7:08