Full Version: Canon EF 35mm f/1.4 L II corners vs Sigma AF 35mm f/1.4 Art corners
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Pages: 1 2 3

Jan S

Hello, may I question results from your test of canon 35mm f1,4 mark II because it is inconsistent with other tests on internet. When I compare your MTF(resolution) result from canon 35mm f1,4 II on 5DS R and Sigmas 35mm f1,4 on 20 megapixels than sigma outperforms canon on the edges at f1,4 despite that 5DS R have more megapixels and has AA filter cancelation.
When I compare this same setup on Dxomark (links below) than sigma has pronouncedly worst corner performance. I know, they use different approach but when I also compare this setup on "the-digital-picture.com" (link below) I can optically see pronouncedly better performance with canons 35mm mark II. The difference can be seen in both comparison of 1Ds mark III R vs 1Ds mark III and 5DS R vs 1Ds mark III. By this I meant no offence, I just want to know that your finding are correct and if they are correct than how should I interpret them or what cause the difference. Thank you.
Sincerely Jan.
At Dxo I compared sharpness on field map. Measurements -> sharpness -> Field map
1st of all, "20mp" is false. Not sure why PZ calls it "20mp" and not 21mp.


Other things to consider:

DXO is very, VERY unreliable. Look at the Canon Ef 17-40mm f4 L USM for instance: http://www.dxomark.com/Lenses/Canon/Cano...ments__795 Wow, amazing performance at 17mm and wide open, right?


On top of that, their field map is silly, they only have 4 test points per lens and the rest is just from interpolation and rotation.


PZ uses sharpening in their MTF evaluation. Sharpening will have more effect on certain frequencies than others, so some frequencies are preferred over others, making the cross sensor results a bit hard to judge. And obscures diffraction effects somewhat.


The digital picture crops feature has its very own issues. 


But apart from all that, consensus from less problematic comparisons seems to be that the Canon EF 35mm f1.4 L USM II performs noticably better than the Sigma in the borders/corners wide open, partly because of much better CA performance. And that the Sigma 35mm f1.4 Art anyway can be called a sharp lens.

Let's not be hasty. As soon as Photozone tests the Sigma we can say more and more sure, how big the difference in reality is. When comparing the predecessor Canon 35 to the Sigma, I see on Sigma's results border and corner performance much closer to centre performance, than the old Canon 35 was able to deliver.

Given the fact of quite huge differences between borders/corners of the new Canon 35 I would not dare to conclude that much of superiority in the outer areas. Bokeh fringing or not, when accurately focused, the Sigma could still be on par.




When lensrentals compared the new Canon to the competitors, Roger said

“The new Canon lens is clearly and definitely better than the very good lens it replaces, both on and off axis. It is also better, from an MTF standpoint, than the Zeiss 35mm f/1.4. When we look at the average MTF curves, it's a tiny bit better, even, than the Sigma 35mm f/1.4 Art lens, although the difference is small. The two lenses are very close with the Canon being slightly better off axis (14 to 20mm) at 10 and 20 line pairs frequency. Could you tell the difference in a photograph? I doubt it, they'd be nearly identical. Other things like vignetting, focus accuracy, out of focus highlights, cost, etc., would be much more important to you than this slight MTF difference.”

What's the deal with having 50mps if all it does is show up $2,000 lenses? and "having to stop down to F8" to get very good results at the edges and by the time you've done that the center has lost it's superlative sharpness.


Quote from review......To be fair though - "at large apertures you don't tend to require sharp corners anyway"


 In reality this should read...... at it's price point "we should seriously expect" sharp corners!! 


Well done Sigma!
"Lets not be hasty". 




Can very clearly tell the difference in photographs.
Once more I'm totally impressed.

By the Sigma. At this price, like Roger said, "they''d be nearly identical".
Regardless how you look at it - this review does not reflect my opinion. 

It reflects what Imatests tells me. And I can visually confirm the results.


The issue that we are seeing here is that the lens (also the others) can scale massively in the center. But not in the corners.

This is emphasized by the lack of an AA filter in the 5Ds R. 


The charts are simply illustrating this "spread". On older cameras, this never showed up because the center quality always hit the resolution limit of the sensor "very early" so it could never show its potential. Now it can. Actually we had this discussion before .... :-)


The question may be whether I have to go for a "near center" reading now just to avoid the extremes of the dead center. Alternatively I could switch to a logarithmic scale or so.


Normally I prefer not to argue with this but it also helps to have a look at the LW/PHs here.

Compared to the 5D II numbers they are much better. 

e.g. they peak at 3800 LW/PH for the 35L II vs 3100 LW/PH for the Sigma on the 5D II.

Now, it is likely that the Sigma will beat the Canon on the 5Ds R because of scaling effects but the 5Ds R also shows that the scaling effects are gone now so the question is how much is in there.
Quote:"Lets not be hasty". 




Can very clearly tell the difference in photographs.
Which way?
Quote:Which way?
By looking at the crops side by side?
Another two things to consider: the comparison was done with camera on a tripod and at base ISO. Now, a 35 mm is often called a reportage lens which then means higher ISO, no tripod. The 5DS R is not exactly a high ISO machine, if I believe what a couple of testers said while the Sigma can be used on any of those great and sensitive sensors, be it Sony or Nikon. In my opinion the Canon lens is slightly better under perfect conditions. In everyday handhold shooting there's just no Canon body up to the competition, and more than that: with the new A7xII series there's a HR sensor with high ISO capacities plus in body stabilisation which will max a Sigma or Zeiss lens in much worse conditions than the Canon body ever will be able cope with.

The other thing is sample variation. Canon does a very good job in that, Sigma less so. But as Roger's OLAF charts show, the worst Sigma was still better at some frequencies than the best old 35 from Canon. Now I wonder, where in this range was the compared sample of Marvin Anthony? If on the "low end" you could get a sample which is even closer to the Canon (but will not exceed). While if it's on the higher end one could see a bit more difference with an average sample compared to the Canon.

And the third thing just occurred while typing this post. Maybe Canon's AF is excellent at all distances equally, who knows. I use my 35 mm sometimes at close or closer distances and it turns out, the possibility of adjusting AF on 4 different distances with the dock is quite useful and gave me more keepers than before. One can say "I expect a lens to be perfect out of the box", fine, dream on. Personal experience: not even Zeiss at Zeiss prices is able to do that, very especially not in 800€ regions.

But okay, this advantage would mean nothing, if the 35/1.4 L II would be adapted on a Sony mirrorless.
Pages: 1 2 3