Wow, Sony trumps everything except on one focal length (24-70).
Almost hard to believe!

Will we get the same results/sharpness if we used Sony lenses with Canon adaptors on a Canon body?

what are you a robot? i just said you can select which camera the lens it tested on.

Here is a voice from the North of Norway again: Whats the best lens i can by to Sony a7RII (or Sony a9) for photographing the Northern Light? For a couple og years I have used my Canon 16-35mm F/2.8 L lens getting some splendid results, but to compete wit nature photos you need the best weapons. Sony a7RII and A9 are good weapons, my Canon 5D Mark IV is ok, but it is my third choice. I am no 100% pleased with metabones adapter, canon 16-35mm F/2.8 L and Sony a7RII, and looks for a sony lens for my main use in this class- Northern Light photo. What lens do you recommend? Do not recommend any lens if you don not believe it will take sharper photos with my best camera than my canon 16-35 mm does! I have read DXO’s scores of the 16-35mm group lenses, but can I trust the scores og not?

Wow. This is a bit unexpected, but very good to know. I always thought, the Canon L lenses are sharper than the Sony fe lenses. Thanks so much for that list 🙂

Why not test all lenses on the Sony A7Rii or iii with the appropriate adaptors? Using different cameras just adds too many variables.

Brian, you act like a pop-up menu……is that all you do all day, defend Sony at any cost ?
I have very limited time but , perhaps I should send some photos and not just numbers.
Sony has NOT discovered the secret formula.
Believe me, you get what you pay and both Canon and Nikon produce fantastic lenses with a lot of character.
I like you, you are a workhorse in photography but Sony is just another good camera and lens maker.
My fav is A7II and have no interest for A7RIII, I will wait for A7RIV.

An interesting review. At 35mm f/1.4 Canon is also better. Results with the Nikon D850 will be better.

Brian, I really want to believe that you are a fair and unbiased judge of image quality, so let me make one followup comment.

You say the lens tests are not about sensor resolution, and the DxO score just shows that Sony makes better glass. I’m pretty sure sensor resolution is a factor in the lens score, but let’s set that aside.

The link to the DPReview studio scene shows that 5DSR with Canon’s 85/1.8 produces a clearer, more detailed image than the A7RIII with Sony’s 85/1.8. And that is true for every ISO. Follow me so far? And yet DxO gives the A7RIII a DxO score that is 14 points higher than the 5DSR. And DxO also gives that Sony 85/1.8 a DxO score that is 11 points higher than that Canon 85/1.8. So … whether you look at the camera or the lens, or the combination of the two, the Sony gear scores way higher even while the Canon gear produces a clearer, more detailed image at every ISO. That’s why I say the DxO score is fishy.

One might rationally think, “A higher DxO score is earned by producing a better image.” However, I think that the DPReview studio scene shows the contrary. It seems to be conclusive proof that a much worse DxO-scoring camera and lens combination produces a distinctly better image. Don’t you see the contradiction there? Doesn’t the evidence show you very plainly that the DxO score is simply erroneous?

I have been a follower and fan of DxO test results for years. However, starting a few years ago DxO shifted its focus from testing to a range of other activities. As a result my confidence in DxO test results has dropped significantly. For example, why hasn’t DxO tested the Nikon 70-200 E lens ?
Do manufacturers pay or enable DxO to test their products and thereby have a way to avoid being tested and published ?
Today it’s very hard to find independent and trustworthy apples to apples lens tests results. The industry needs a new standard for testing lens sharpness independent of sensor resolution and vice versa.
The manufacturers hide behind their own MTF charts which offer an analysis of the relative quality of the lens as a function of the distance from the center of the lens, often tested at a very low line-pair-density, far from the peak sharpness of the lens. This might be of interest to the manufacturer but it does little or nothing to satisfy the consumers need for comparative information.
Most other industries are forced to measure and publish in an informative and comparative way the performance of their products, why not the same for lenses, and digital image sensors ?

DxO has not made an apples to apples comparison of the present generation 70-200 F2.8 lenses.

LensTip has made this comparison and they report a Sony resolution of 737 compared to a Nikon resolution of 1029.

Your reporting on 70-200 F2.8 lenses is not credible.

While i am no expert, what so many people seem to be missing, is that Sony purposefully set out to design their GM lenses to 100 MP future sensors (or at least with that intention), where as Canon were designing to a much lower MP optimization. All of this effects many things, including the cost of the lens, and how easy/difficult it is to manufacture and keep everything in alignment.

Sony also has the benefit of newer lens designs using the latest computer aided technologies and manufacturing processes in partnership with established lens manufacturers.

If being picky, moving a lens, especially a wide angle lens from one system to another, does not account for the different thicknesses and optical characteristics of the filter glasses placed in front of the sensor, with Canon being less than sony (since they are from the film days when no glass was in front of the film), to sony being around 2mm, to i think it was Fuji(?) around 4 mm. There are certain advantages and disadvantages in how the filter glass is done, and each lens manufacture optimizes for theirs.

One area with today’s complex optical designs is how well they can stay in alignment through knocks (including being banged around when shipping), temperature changes, and aging of seals, etc. This would be a nice subject for someone to touch on – particularly with interviews of different manufacturers and designers.