Hi Slobodan,
My take on the issue is that a larger sensor can collect more photons. A larger sensor will also go with a larger lens,as you would need a longer focal length for the same angle of view. Photons are important, because much of the noise is just statistical variation of the number of photons reaching the pixels.
Let's take a simple example. You go from say 16 MP APS C to say 20 MP "full frame". The full frame sensor will have 1.6^2 = 2.56 the area of the APS/C sensor. So it can collect 2.56 times the number photons (all other factors kept constant).
Now, let's assume that you use a decent quality 35 mm lens on the APS-C camera and use f/5.6, which probably is pretty optimal. That means the area of the aperture will be 31 mm^2.
Now, on the full frame a 55 mm lens would give you the same angle of view. You could still shoot at f/5.6 the area of the aperture would than be 76 mm^2. So using same exposure time 76/31 = 2.5 times more photons would leak trough that aperture. So you would get the same number of photons per mm^2 on both sensors.
Now, if you needed that f/5.6 to achieve a given DoF on the APS-C and would need similar DoF full frame, you would need to stop down to 5.6 * 1.6 = 8.9 (if my memory don't fails me). If you shoot on tripod, you can increase exposure time to compensate, and still get the same amount of photons. Or yo can increase ISO, which means collecting less photons.
As a side note, astronomers like diffraction limited systems, and they don't need things like aperture. More often than not, they are looking at stars and just want to collect light from those stars. Astronomers also seldom have problems with DoF :-), so for astronomers a big lens is a big advantage. A larger lens collects more photons, but a larger lens has also less diffraction, so the spot image of the star will be brighter. As far as I recall, in film times a telescope having twice the diameter would have sixteen times the detection capability. With digital sensors the case may be different as not so much is gained once the star image is smaller that the pixel size of the sensor. Ray perhaps has some interesting input on the issue.
Best regards
Erik
Bill,
Pardon my ignorance and impatience to read every word of that interesting article. I am not sure I understood the concept of "larger lens" or "wider aperture lens." Are you saying:
1. That a lens with a wider max aperture (say f/2) is always going to result in better image quality than a, say, f/4 lens, regardless of the actual aperture used or...
2. That a lens with a wider max aperture (say f/2) will only result in better image quality at apertures wider than the other lens?
