I'm assuming for most devices, we only need the output color to match the appearance of the target color for human perception. If that's the case, would a photo look different to what the object looks like in the real world to a creature that sees different wavelengths than us?
Staffjamesfong1
@reinaw1012 Great question. The answer is yes! You may have even experienced this at one point. If you take a picture of a display's output, the colors in that picture may not match correctly (especially for older cameras and displays). This is because the spectral sensitivities of a camera's sensors can be very different from the LMS sensitive which these displays are intended to fool.
If you don't count a camera as a "creature", there are also animals with >3 cone types. If those creatures built displays, they might just want >3 color channels.
saltyminty
Do cameras also typically also have 3 color channels/spectral sensitivities for their inputs, similar to displays? Or do they follow a different process in order to capture color?
geos98
@saltyminty, I believe most cameras would measure only 3 color channels, similar to display (but one measures light, one emits light). This is because 1. it is sufficient as human only have three types of cone cell, 2. building one with more is expensive. However, I believe there are exception, like this Sony RGBE camera.
joeyzhao123
Speaking of color reproduction, is there some way to easily make our monitors look the same without buying specific tools to measure? I know when moving from my laptop to desktop, there is a visible change in how it looks.
I'm assuming for most devices, we only need the output color to match the appearance of the target color for human perception. If that's the case, would a photo look different to what the object looks like in the real world to a creature that sees different wavelengths than us?
@reinaw1012 Great question. The answer is yes! You may have even experienced this at one point. If you take a picture of a display's output, the colors in that picture may not match correctly (especially for older cameras and displays). This is because the spectral sensitivities of a camera's sensors can be very different from the LMS sensitive which these displays are intended to fool.
If you don't count a camera as a "creature", there are also animals with >3 cone types. If those creatures built displays, they might just want >3 color channels.
Do cameras also typically also have 3 color channels/spectral sensitivities for their inputs, similar to displays? Or do they follow a different process in order to capture color?
@saltyminty, I believe most cameras would measure only 3 color channels, similar to display (but one measures light, one emits light). This is because 1. it is sufficient as human only have three types of cone cell, 2. building one with more is expensive. However, I believe there are exception, like this Sony RGBE camera.
Speaking of color reproduction, is there some way to easily make our monitors look the same without buying specific tools to measure? I know when moving from my laptop to desktop, there is a visible change in how it looks.