The way I think about the relationship between FX and DX is:
Assume the FX crop has twice the area as a DX crop.
If I shoot a large gray card, both FX and DX crops will have the same, exciting shade of gray. So, the intensity of light--the photons per mm^2--is the same on each crop. Obviously.
But, the total amount of light in the FX crop is twice that on the DX crop because the FX crop has twice the area.
If, instead of a uniform gray card, an actual object whose image just fills the DX crop is exposed, the image in the FX and DX crops will be the same in size and brightness. Duh.
If I zoom in and shoot again at the same exposure with the image now filling the FX sensor, the image in the FX crop will now look the same as it did before within the DX crop; Just bigger with the same shades and tones. However, overall, there is twice the amount of light--twice the number of photons--in the FX crop because the FX crop has twice the area.
This factor of two in light collection for frame-filling crops, between FX and DX, is the basis for claiming a 1-stop difference in noise performance.
Hope this helps.
Assume the FX crop has twice the area as a DX crop.
If I shoot a large gray card, both FX and DX crops will have the same, exciting shade of gray. So, the intensity of light--the photons per mm^2--is the same on each crop. Obviously.
But, the total amount of light in the FX crop is twice that on the DX crop because the FX crop has twice the area.
If, instead of a uniform gray card, an actual object whose image just fills the DX crop is exposed, the image in the FX and DX crops will be the same in size and brightness. Duh.
If I zoom in and shoot again at the same exposure with the image now filling the FX sensor, the image in the FX crop will now look the same as it did before within the DX crop; Just bigger with the same shades and tones. However, overall, there is twice the amount of light--twice the number of photons--in the FX crop because the FX crop has twice the area.
This factor of two in light collection for frame-filling crops, between FX and DX, is the basis for claiming a 1-stop difference in noise performance.
Hope this helps.