If the numbers that Matthias gave in his post are correct and they look pretty close to right, then we can make the following statements.
If you look at the central portion of the 5D sensor that is the same physical size at the 20D sensor then you get the following
Central portion of the 5D sensor:
(22.5*15.0) * 14025 pixels per mm^2 = approx. 4.7MP
This makes sense because the 5D sensor has 2.5X the area of the 20D but only 1.5X as many pixels. This seems to indicate that if you were to crop a 5D image to have the same apparent FOV that a 20D has then the 20D would have 3.5MP more then the 5D image and thus would give better results during enlargement. If you crop a 5D image to the same resolution as a 20D then it represents a larger portion of the sensor and will not have the same (apparent) magnification that the 20D give. So it seems that unless you can get a slightly longer optically equivalent lens then the 20D will yield more pixels then the 5D when comparing FOV.
20D vs. 5D for long lens photography
in Canon EOS Mount
Posted
If the numbers that Matthias gave in his post are correct and they look pretty close to right, then we can make the following statements.
If you look at the central portion of the 5D sensor that is the same physical size at the 20D sensor then you get the following
Central portion of the 5D sensor:
(22.5*15.0) * 14025 pixels per mm^2 = approx. 4.7MP
This makes sense because the 5D sensor has 2.5X the area of the 20D but only 1.5X as many pixels. This seems to indicate that if you were to crop a 5D image to have the same apparent FOV that a 20D has then the 20D would have 3.5MP more then the 5D image and thus would give better results during enlargement. If you crop a 5D image to the same resolution as a 20D then it represents a larger portion of the sensor and will not have the same (apparent) magnification that the 20D give. So it seems that unless you can get a slightly longer optically equivalent lens then the 20D will yield more pixels then the 5D when comparing FOV.