snake0ape wrote in post #17175215
This is what I need to think about when I consider reach with a crop camera over a full frame.
If I pair the Tamron 150-600mm with my 5diii, DXOmark gives this pair a 17P-Mpix sharpness rating. This means I really have a 17megapixel camera when I use this lens.
If I used a 70D camera, DXOmark gives the camera/lens pairing a low 8P-Mpix sharpness rating. This means that I am really using a 8Mpix camera, because that is all the pair can resolve.
So doing the math (17/8) means that my 5Diii has 2.125 times more resolution than the 70D camera. So lets do the reach calculation. The 70D has a 2.56 (1.6 x 1.6) greater pixel density then the 5Diii. So 2.56 / 2.125 = 120%. So this means that the 70D will only give me 20% more resolution compared to enlarging my 5Diii's photos to match the 70D.
Of course this is not real world comparison and the DXO number is a general single factor number. But it is something that makes me ask the question if the reach of the 7dii worth getting. Well at least with this lens and compared to using my 5diii.
I also own the 70-200ii. With this lens and using DXOmark's numbers, I will have a 170% resolution advantage using the 70D over the 5diii. Now for me, that is enough of an "numbers" advantage to justifying getting a crop camera.
Interesting theory, and a good way of thinking about the benefits of a crop body for a given lens.
Would it be easier though to just think of it as the FF body having 1.6^2=2.56 times the area, therefore taking a Canon APS-C sized crop out of the middle gives 2.56x less resolution, therefore 17MP sharpness cropped to APS-C size is 17/2.56=6.64MP.
That gives the same result (8MP/6.64MP = 120%), but might be easier to visualise.
The only note of caution I'd sound is that the 24-70II is rated as 18MP sharpness, but I've seen a number of claims that it outresolves the 5D3 sensor, and shows noticeably better results on a 36MP A7R. But if it "only" gives 18MP sharpness on a 22MP 5D3 then I'm assuming the rating is a bit more complex. For example, the Sigma 35mm F1.4 DG HSM rates as 17MP on the 5D3 (so appears to be lens, not sensor, limited), but then rates as 30MP on the D800E.
I suppose the presence (or not) of an AA filter would skew results; e.g a 20MP sensor with an AA filter may show <20MP of resolution with a good lens, but that same lens may show higher resolution on another sensor without an AA filter.