Upload & Sell: Off
| p.4 #6 · 5D3, 6D and D600 true RAW (dcraw) comparison |
Well, no. Depth of field, yes, exposure, no. The FF sensor doesn't gather "1.38 f-stops" more light, it gathers more light in total, but spreads it over a larger area, and in the end is no further ahead. If a shot is properly exposed at f/4 on crop, then it is properly exposed at f/4 on full-frame, with the same shutter speed and (real) ISO setting. There is no "equivalence" measure here.
This is one of the advantages of the smaller formats. Smaller lenses, same light gathering ability, for the correct exposure. For example, the Olympus 150/2.0 is really one stop faster (i.e. light gathering) than the "equivalent" (angle of view) 300/2.8 full-frame lens, but the smaller lens will have more depth of field, in spite of its smaller max aperture.
We are exchanging "aperture" and "f-stop" terms too liberally (well, this is normal at any talk!). And we must not forget the ISO advantage of bigger sensors in the equivalence equation. BTW the truly equivalent lens in FF to the four thirds 150mm @ F2 would be the 300mm @ F4.
A bigger sensor always captures more light (all other exposure parameters equal) at the same f-stop (thanks to using a bigger effective lens aperture diameter). What I tried to mean in my post is that when using a bigger sensor, one can always use a smaller f-stop ("downgrading" to the same aperture) to exactly mach the smaller sensor, of course using an equivalent (not the same) focal length. Better with a numerical example, remembering the old definition of f-stop (f-stop = focal length / aperture):
A 35 mm lens in FF using f-stop 4 has an aperture diameter of 35/4 = 8.75mm. To get the same shot in 1.6x Canon crop (same depth of field, same angle of view, same diffraction, etc) we need a 22mm lens at f-stop 2.5 (because 4/1.6=2.5) and fortunately this also yields an aperture of 22 / 2.5 = 8.75mm. Yes, both systems (FF at F4, crop at F2.5) gather exactly the same light as long as they use a "light cone" of the same 8.75mm width, and the proper lens to spread that light over their respective sensor surfaces. Of course, the light is spreaded in FF in a sensor surface 2.6 times greater. Thus, for the same shot (same shutter speed) in FF it would appear underexposed, unless the FF shooter increases the ISO 1.33 fstops to push the exposure 1.33^2=2.6 times to compensate (for example, by using ISO 250 instead of ISO 100). Now the two shots are identical, and because the full frame sensor has also a 1.33 f-stop advantage in ISO (its pixels are 2.6 times bigger and count 2.6 times the photons, increasing the signal to noise ratio) both match even in the noise matter. As it should be, because both sensors have the same performance in our assumption and the same amount of light was used taking the picture.
In crop cameras, in addition to multiply the focal lengths by the crop ratio to get the FF-equivalent focal lenght, we need to divide the f-stop by the crop ratio to achieve the same depth of field and diffraction effect. But we can also multiply the f-stop by the crop ratio to determine the FF f-stop with a equivalent light gathering ability (of course, this requires entering the ISO into the equation, which is fair if we also want equivalent noise levels).
The equivalence of systems is true, despite usually it is not available (smaller sensors usually have no comercially available fast enough lenses to be able to make certain shots that bigger sensors can -ok, trading depth of field-). Or the equivalence is made available at the expense of no advantage of the smaller format in terms of equipment size, like the example of the Zuiko 14-35 F2 vs Canon 24-70 F2.8 weight/size (or the bulky 35-100 F2 lens of the same manufacturer vs Canon 70-200 F4). Or the otherwise bulky designs are partially mitigated by creating lenses which desperately need heavy correction by in-camera software (many mirrorless lenses).