redfordxx
26th of June 2008 (Thu), 07:00
How exactly CA works?
After reading something about CA I have still one question. I make a simplified example of a situation which will best answer my question.
1) Imagine you are taking a photo, and top right corner of the frame is completely dark, only a small green light is there. The size of the light would be 2x2 pixels on the sensor. Since it is green, it stimulates 2 pixels because of the bayes filter. So it will be let's say pixels p[9,10], p[10,9]
2) Imagine now the same light is red. Because of the CA the red light from same place would stimulate different part of sensor. In reality I don't know, but let's say for the example, I it will be p[5,5].
So far it is clear for me.
3) Now imagine the light is yellor. Yellow is split to 50% green and 50% red I think. So, what would be the result on sensor? I see two options and dont know which one is it or if there is other phenomenon:
a) The light is broken on the lens green will fall where the 100% green was originally and red where was red. So, pixels p[5,5], p[9,10], p[10,9] will be stimulated.
b) The light from the lens will continue still as one light together and will be split only on bayes filter. So, pixels p[7,7], p[7,8], p[8,7] will be stimulated.
c) other option?
So, how is it?
Appreciate your help
R*
After reading something about CA I have still one question. I make a simplified example of a situation which will best answer my question.
1) Imagine you are taking a photo, and top right corner of the frame is completely dark, only a small green light is there. The size of the light would be 2x2 pixels on the sensor. Since it is green, it stimulates 2 pixels because of the bayes filter. So it will be let's say pixels p[9,10], p[10,9]
2) Imagine now the same light is red. Because of the CA the red light from same place would stimulate different part of sensor. In reality I don't know, but let's say for the example, I it will be p[5,5].
So far it is clear for me.
3) Now imagine the light is yellor. Yellow is split to 50% green and 50% red I think. So, what would be the result on sensor? I see two options and dont know which one is it or if there is other phenomenon:
a) The light is broken on the lens green will fall where the 100% green was originally and red where was red. So, pixels p[5,5], p[9,10], p[10,9] will be stimulated.
b) The light from the lens will continue still as one light together and will be split only on bayes filter. So, pixels p[7,7], p[7,8], p[8,7] will be stimulated.
c) other option?
So, how is it?
Appreciate your help
R*