You're kidding, right?
I went to DXO and compared out the Measurements>Sharpness and CA field maps and profiles of the FE55 on the A6000, the Sigma 60mm DN A on the A6000, the Otus 55 on the Nikon D7100, and the Sigma 50mm 1.4 Art on the Canon 70D (closest to 24Mp APS-C DXO has tested).
You are not just comparing sensor sizes with these bodies, but also resolution (number of pixels). A larger sensor will give higher resolution, and a greater number of pixels provides higher resolution.
Thank you! I actually have the 50mm. I just wanted to know why there is a difference
Which would be nice.
It is better to think of smaller pixels as recording the blur from aberrations (including diffraction) more accurately. The better this point spread function is recorded, the greater the opportunity for deconvolution software to recover "lost" detail.
I think 46 or 48 MP is more likely, but the difference is negligible. (Except in processing time -- more pixels means fewer frames per second.)
As others have already postulated, its probably more to do with sensor pixel size. The FE lens is obviously designed primarily for the FF sensors pixel size so probably resolves better with it than the asp-c sensor.
The Nex-6 has a 16mp sensor. Expanding that pixel pitch to FF would be 36mp which matches that of the A7R where the FE55 works very, very well. So it seems the Nex-6 and A7R have about the same pixel size. Using the points made about pixel size, the FE55 should theoretically work equally well on both the A7R and the Nex-6.
Nice explanation and you just provided me with a new perspective (that the sensor should ideally out-resolve the lens in order for software correction to be more effective) all blogs and sites I have been to haven't managed to explain or rather did not have this knowledge.
No way. A larger sensor provides higher resolution, all else being equal. More megapixels provide higher resolution, all else being equal. Add both together for A7R vs. NEX6. 35mp has much higher resolution than 16mp APS-C, regardless of the lens used. You can confirm this at DXO using the same lens on a Nikon D800 and D7000.
DXOMark shows clearly that 24mp APS-C has higher resolution than 16mp APS-C, regardless of whether you use a humble lens or the best.
Actually the Nex-6 and A7R have a similar pixel pitch (~4.8um) therefore the same resolution with the Nex-6 just being a crop of the FF sensor. The A7r only has more pixels due to it's size but the resolution would be the same as far as lens performance is concerned.
From a lens performance POV, as requested by the OP, it's the density of the pixels/number of pixels per square mm that's the more important factor rather than the size of the sensor.
No. Having sample pixel pitch does not mean same resolution (per height).
You're welcome. One other difference I forgot to include here is that extra 5mm. It doesn't matter for a lot of people but for me it does. I find that 75mm or 70mm (in FF equivalent) to be a bit short for head shot portraits. The 55mm for my A6000/NEX-6 is close enough for that classic 85mm----excellent for close up portraits.
Nope, forget pixel pitch, it will only lead to the wrong answer. Here's the evidence, using the Otus 55mm:
Look at the DXO results below. The NEX-6 uses the same sensor as the Nikon D7000. The A7R uses the same sensor as the Nikon D800E. The Zeiss Otus 55mm is comparable in performance to the FE 55. The sharpness results shown below demonstrate that pixel pitch is not the answer to the OP's question. The correct answer is a combination of sensor size, increased megapixels, and lack of a blur filter in the A7R.
Thank you for the clear technical explanation.
It will do better (have higher resolution) than on a 36mp FF sensor.
But there will still be a large net gain in resolution, probably around 50% based on these results:
The Otus will easily score twice as high for sharpness on a 54mp FF sensor vs. a 24mp APS-C sensor.
