Review: Canon EF 35mm f/1.4L II at DXOMark

[Canon Rumors]

Let the fun begin, DXOMark has completed their review of the Canon EF 35mm f/1.4L II along with a bunch of numbers.</p>
<blockquote><p>With the introduction of high-resolution bodies such as the Canon EOS 5Ds and 5DsR, with their ability to highlight lens flaws, it comes as little surprise to see Canon roll out new versions of some of the older models in their line-up. The original 35mm f/1.4 wasn’t a bad performer, but it was beginning to show its age. The new model greatly surpasses the original in several key areas. It is generally sharper, but it’s the improvement in geometry and uniformity that really impresses. Not only does it have low field curvature and astigmatism, but Canon has also dramatically lowered lateral chromatic aberration, with the overall effect of greatly improving imaging performance in the outer field (where the older model was weak).</p></blockquote>
<p>According to the review, the Canon EF 35mm f/1.4L II doesn’t quite score as well as the Sigma 35 f/1.4 Art in sharpness. However it seems to perform well in all the other categories against the Sigma and Zeiss 35 f/1.4 Distagon.</p>
<p>Comparison below (Click for larger):</p>
<p><a href="http://www.canonrumors.com/wp-content/uploads/2015/10/dxo3514lii.jpg"><img class="aligncenter size-medium wp-image-23132" src="http://www.canonrumors.com/wp-content/uploads/2015/10/dxo3514lii-728x471.jpg" alt="dxo3514lii" width="728" height="471" /></a></p>
<p><strong>Canon EF 35 f/1.4L II $1799: <a href="http://www.bhphotovideo.com/c/product/1180801-REG/canon_9523b002_35mm_f_1_4l_ii_usm.html/BI/2466/KBID/3296" target="_blank">B&H Photo</a> | <a href="http://www.adorama.com/CA35142.html?utm_term=UbK24x0al34oSlvW4eT8QxjoUkX3mDVXeWC-Ug0&utm_medium=Affiliate&utm_campaign=Other&utm_source=rflaid64393&cvosrc=affiliate.64393" target="_blank">Adorama</a> | <a href="http://amzn.to/1Uehm5w" target="_blank">Amazon</a></strong></p>

 

[Rahul]

Ever since the DXOMark fiasco over their "score" of 70-200 IS II, I lost respect for them. Over time, I have increasingly felt that they are way too biased for my liking. Having an opinion is fine, but indulging in pseudo-science of translating data into numbers using a "magical formula" (which they won't disclose) and present "facts" doesn't appeal much to me.

That being said, both the 35L II and the Sigma 35A are excellent lenses. Comparing it with the Sigma is like splitting hairs.

Personally, I got the 35L II for instead of Sigma so as to remain within the Canon ecosystem, better / consistent AF, beautiful bokeh and the red ring.

 

[Maiaibing]

Let the fun begin, DXOMark has completed their review of the Canon EF 35mm f/1.4L II along with a bunch of numbers.

No doubt a great lens. Price is the issue.

1.800$ in Japan and the US, somewhat more in Europe. Still holding on to my 35L (I) for now...

 

[Corneria]

Ever since the DXOMark fiasco over their "score" of 70-200 IS II, I lost respect for them. Over time, I have increasingly felt that they are way too biased for my liking. Having an opinion is fine, but indulging in pseudo-science of translating data into numbers using a "magical formula" (which they won't disclose) and present "facts" doesn't appeal much to me.

Don't look at the scores, check the graphs. These are unambiguous. There you see that the Sigma is sharper when stopped down, but at more open apertures, the Canon is better.

I personally have the Sigma for my Nikon, and I'm very happy. That sharpness in the corners is not so important for me in a wide-angle lens. What I do have problems with, is AF consistency and lack of weather sealing. That having said, Canon users have now the 'best' 35mm available without any concessions!

 

[Rahul]

Ever since the DXOMark fiasco over their "score" of 70-200 IS II, I lost respect for them. Over time, I have increasingly felt that they are way too biased for my liking. Having an opinion is fine, but indulging in pseudo-science of translating data into numbers using a "magical formula" (which they won't disclose) and present "facts" doesn't appeal much to me.

Don't look at the scores, check the graphs. These are unambiguous. There you see that the Sigma is sharper when stopped down, but at more open apertures, the Canon is better.

Aye, but for me buying a large aperture prime is about shooting at wide open. Stopped down, most lenses are sharp enough for it to not matter much.

 

[Corneria]

I agree! But on 36MP the Sigma is very sharp wide-open, and I think that wouldn't be much different on the 5Dsr.

 

[Sator]

I agree with those who express immense scepticism about DXO Mark. Alex Cooke at F Stoppers put it extremely well for me:

https://fstoppers.com/originals/are-your-lenses-suddenly-obsolete-85888

I take anything that comes from DxOMark with a healthy grain of salt. Here’s the problem: they refuse to publish detailed methodology. There’s a reason science is an open community: science is hard... That’s why we publish, critique, evaluate, question, test, test, test, retest, and replicate results before they are accepted as fact. The “because I said so” method simply would not fly and it’s why I highly suspect I see sparse mention of DxOMark when I search peer-reviewed academic journals, despite the scientific importance and industrial reach of the study of modern optics.

When someone does publish results in which methodology is laid open to scrutiny, it fails to corroborate DXO Mark's results. I point you to Lens Rental's recent study:

http://www.lensrentals.com/blog/2015/09/35mm-f1-4-shootout-canon-35mm-f1-4l-ii-vs-sigma-35mm-art-vs-canon-35mm-f1-4l-i

This demonstrates that the MTF plots for the new Canon 35mm are slightly better than those of the Sigma. Notice that this study does show you mean and standard variation (I assume they show 95%iles), as well as declaring samples sizes, just as I would expect of an author if I were to review a scientific paper prior to publication.

I guess with my science background where the importance of scrutinising methodology was hammered into me from my first undergrad year through into my PhD lab years, and beyond, I look at the "because I said so" method of DXO Mark results as being little better than witchcraft and voodoo. Worse still, my more cynical side wonders if DXO Mark keep their methods a secret to hide the fact that they make up their results depending on how much money each company "donates" in advance to a secret "independent research foundation". While this may seem too harsh, I have seen similar sorts of scandal come to light in biased scientific trials before, and until DXO Mark lay bare their methodology for independent scrutiny one should sadly assume no less.

 

[ecka]

DxO blah blah blah...
http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=994&Camera=453&Sample=0&FLI=0&API=0&LensComp=829&CameraComp=453&SampleComp=0&FLIComp=0&APIComp=0
How did this happen? :

 

[PureClassA]

I own the Sig 35 and 50. Though I rarely use the 35 anymore, it's a fantastic lens. I haven't used the new 35 yet from Canon but I suspect it's just as sharp as the Sig but with better bokeh (always where sig falls behind Canon) and of course more reliable and faster AF. For people who demand absolute reliability and durability, the Canon is the way to go. As soon as Canon announces the new 50mm (like the Sigma, I expect to be even more impressive than the 35mm variation) my Sig 35 and 50 are going up for sale.

 

[PureClassA]

And that may even be just a difference in focusing accuracy between a Canon lens and third party lens. I tend to think it's just the Canon being sharper, but either way, it doesn't matter. The Canon still wins that comparison wide open, which is the only reason I'd own such a lens anyway. If I wanted to shoot smaller apertures, I'll grab my 16-35 f4, which is just ridiculously good glass.

DxO blah blah blah...
http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=994&Camera=453&Sample=0&FLI=0&API=0&LensComp=829&CameraComp=453&SampleComp=0&FLIComp=0&APIComp=0
How did this happen? :

 

[Loswr]

And that may even be just a difference in focusing accuracy between a Canon lens and third party lens. I tend to think it's just the Canon being sharper, but either way, it doesn't matter. The Canon still wins that comparison wide open, which is the only reason I'd own such a lens anyway. If I wanted to shoot smaller apertures, I'll grab my 16-35 f4, which is just ridiculously good glass.

DxO blah blah blah...
http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=994&Camera=453&Sample=0&FLI=0&API=0&LensComp=829&CameraComp=453&SampleComp=0&FLIComp=0&APIComp=0
How did this happen? :

No, autofocus is not used in the tests. The normal procedure is to manually focus bracket then pick the sharpest shot.

 

[Maximilian]

Lovely! I really love DxO for entertaining us so much.

Either DxO got a cherry picked Sigma and/or the worst EF35L II out there in the market or they'll have a really good time explaining to us why their scores and rankings don't fit to about 80% of the reviews and comparrisons and samlpe pics out there.

As far as I've seen the reviews the Sigma is a great lens, especially for its price, but the 35L II is even better.
On test charts as well as in rl shooting. If that advantage is worth the higher price depends on each own taste.
But as far as I know DxO didn't put the price into their calculation of optical values.

By the way: Still just testing with the 5D3 and not with the 5DS/R. : : :

 

[PureClassA]

Thanks for the info! And like I said in my next line, I tend to think it's just the Canon being even sharper than the Sig. I really don't use the 35mm focal length much if ever anymore outside of my 16-35 f4. I favor the 50mm, so I expect the new Canon 50L to be spectacular and I'll be first in line for a preorder.

And that may even be just a difference in focusing accuracy between a Canon lens and third party lens. I tend to think it's just the Canon being sharper, but either way, it doesn't matter. The Canon still wins that comparison wide open, which is the only reason I'd own such a lens anyway. If I wanted to shoot smaller apertures, I'll grab my 16-35 f4, which is just ridiculously good glass.

DxO blah blah blah...
http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=994&Camera=453&Sample=0&FLI=0&API=0&LensComp=829&CameraComp=453&SampleComp=0&FLIComp=0&APIComp=0
How did this happen? :

No, autofocus is not used in the tests. The normal procedure is to manually focus bracket then pick the sharpest shot.

 

[GMCPhotographics]

More DXO fiction...

The more we laugh at them, the more likely they are to take down their appalling and misleading web site.

 

[PureClassA]

They test it with the 5D3 first because it's a control against their previous tests. I'm sure they will add the 5DSR pretty soon. The Digital Picture has shot tests with the 5DSR and the 35L MkII

Lovely! I really love DxO for entertaining us so much.

Either DxO got a cherry picked Sigma and/or the worst EF35L II out there in the market or they'll have a really good time explaining to us why their scores and rankings don't fit to about 80% of the reviews and comparrisons and samlpe pics out there.

As far as I've seen the reviews the Sigma is a great lens, especially for its price, but the 35L II is even better.
On test charts as well as in rl shooting. If that advantage is worth the higher price depends on each own taste.
But as far as I know DxO didn't put the price into their calculation of optical values.

By the way: Still just testing with the 5D3 and not with the 5DS/R. : : :

 

[LSXPhotog]

I have absolutely no idea how they come up with numbers and overall scores! Haha

From what I've seen the Canon has looked sharper with less vignetting... According to DxO, it is softer with MORE vignetting? I really don't care for this company's opinions on lenses or camera sensors.

 

[Chaitanya]

More DXO fiction...

The more we laugh at them, the more likely they are to take down their appalling and misleading web site.

Remember when Dxo posted a review of their camera for Crapple idevice?

 

[Canon Rumors]

More DXO fiction...

The more we laugh at them, the more likely they are to take down their appalling and misleading web site.

Remember when Dxo posted a review of their camera for Crapple idevice?

I missed that they reviewed their own device. I wonder if they liked it....

 

[Loswr]

More DXO fiction...

The more we laugh at them, the more likely they are to take down their appalling and misleading web site.

Remember when Dxo posted a review of their camera for Crapple idevice?

I missed that they reviewed their own device. I wonder if they liked it....

It's got better IQ than some FF dSLRs*.






* Refers to DxO's SuperRAW™ image format, which averages four exposures for multishot NR, but we will ignore the usual caveats about merging successive exposures and pretend it's just a single image for the purpose of giving it a higher Score with our undisclosed methodology.

 

[ahsanford]

They test it with the 5D3 first because it's a control against their previous tests. I'm sure they will add the 5DSR pretty soon. The Digital Picture has shot tests with the 5DSR and the 35L MkII

Agree for comparative purposes, but we're at the ~ 120 day mark since they 5DS / 5DS R review so one would expect lens testing on the 50 MP rigs soon, at which point even modestly sharp EF glass will vault over any Nikon F mount score. Their scoring system for lenses is horrifically tilted (above all else) to reward how many pixels are sitting behind the lens.

I honestly can't wait, not for Canon pride at all so much as DXO owning a poor rating system and having to explain why a (for example) 70-200 f/2.8L IS II -- heretofore the 1,096th ranked lens in their database -- is now sitting 900 places higher.

- A