Review: Canon EF 11-24mm f/4L via DXOMark

LSXPhotog said:
I find it amusing that DxO hasn't tested a single lens with the 5DS. Are they afraid to reveal lens scores being better than Zeiss and Nikon?

Yes.

They are waiting to find a way to further cook their data to cripple the Canon glass scores when used on a 5DS.

Figure this...

Several L lenses are the equal to or slightly better to Nikon's top glass by DXO's own measures. However, this is tested on an AA filtered 22.3mp 5D3 vs. a non-AA filtered 36mp D810. The Canon glass has a huge disadvantage, yet is right up there.

It stands to reason that once tested on a 5DS, the scores will clearly beat Nikon. The fact that images coming from a 5D3 are of such high quality indicates the sensor is being pushed to its max by the superior L glass.

Now, DXO, Nikonians and other assorted sycophantic apologists will say that these tests are done in a manner by which to be independent the sensor's resolution. This of course is BS, because as you select higher MP cameras on DXO, the score increases.


DXO continues to weigh in dynamic range even on lens tests!!! Not only that, but they make it a big part of the scoring.

This fact, piled upon a mountain of other facts can lead a person to no other conclusion than DXO is biased against Canon, and heavily favors Sony/Nikon.

Now, in fairness to Nikon. Their top pro glass, made in Japan, is very good.

No, they don't have the wide selection of choices that Canon does.

But they are so close to Canon in the common lenses, that it doesn't matter. Their 70-200 (focus breathing issue aside) on a D810 or D750 is so sharp and offers so much detail, it is more than just about anyone needs. In other words, it's not the gear holding you back. Stunning images are created with their system.

However, Canon does edge them out in optical quality, autofocus speed etc.


The only way to really compare them is to do so without cameras. That's right, measure the MTF!!! Each company publishes their own scores. But I think was lensrentals or some site like that that got their own expensive testing equipment and published MTF scores. Look it up.

The results on average was that Canon is slightly better in optical quality. But it is very close on most of the pro lenses. Considering the prices Nikon charges, Canon is a tiny bit better value in that regard also.


But who uses just a lens? It's a lens + camera system that produces an image. For that, the best way to test and evaluate is NOT with some BS that DXO cooks up. But rather, by evaluating dozens of RAW image tests done in comparison by VARIOUS different testers.

Only then will you see for yourself a trend one way or the other. Or maybe not at all, indicating they are too close that small variables in testing are making the difference.


All that said, the 5DS with 50MP will make the most of Canon's L lenses and this is already evident in the test images provided across the web.

DXO has to protect Sony/Nikon, so they use literally the only thing Nikon/Sony has an advantage on, and that is low-ISO DR and throw that into the mix for scoring lenses. (rolls eyes)

What if the only advantage Nikon had was shutter speed? That is akin to them factoring shutter speed into lens image quality scores to prop them up against Canon.
 
Upvote 0
The Flasher said:
This lens is nothing short of spectacular. That "horrid" ca doesn't seem to be that horrid architectural shots, most shot at f5.6-8, a matter of fact, they don't show up at all. Back-lit scenes look fantastic. The super wide, rectilinear 11mm range takes the frustration out of having to stitch multi 17mmts, images are tack sharp. Use this for video as well, amazing.

Micro peep charts in the corners wide open all you like, not until you start taking pictures do you see the fantastic results in real world application of this photography tool.

Three stitched and shifted TS-e 17L portraits gives an angle of view of around 14ish-mm and a nice proportion to the frame (4x5 I think) but two or three shifted landscapes is well under 11mm, closer to 9-10mm. But the shape of the frame is panoramic and close to 16:9 aspect.

The nice thing about the 11-24mm is that it's very sharp and very easy to use. But it's heavy and expensive and one really needs to use it a lot to justify lugging it about. While the Sigma 12-24 is an inferior product...it's a 1/4 of the price here in the UK when bought new and even less when bought S/H.
 
Upvote 0
K said:
LSXPhotog said:
I find it amusing that DxO hasn't tested a single lens with the 5DS. Are they afraid to reveal lens scores being better than Zeiss and Nikon?

Yes.

They are waiting to find a way to further cook their data to cripple the Canon glass scores when used on a 5DS.

Figure this...

Several L lenses are the equal to or slightly better to Nikon's top glass by DXO's own measures. However, this is tested on an AA filtered 22.3mp 5D3 vs. a non-AA filtered 36mp D810. The Canon glass has a huge disadvantage, yet is right up there.

It stands to reason that once tested on a 5DS, the scores will clearly beat Nikon. The fact that images coming from a 5D3 are of such high quality indicates the sensor is being pushed to its max by the superior L glass.

Now, DXO, Nikonians and other assorted sycophantic apologists will say that these tests are done in a manner by which to be independent the sensor's resolution. This of course is BS, because as you select higher MP cameras on DXO, the score increases.


DXO continues to weigh in dynamic range even on lens tests!!! Not only that, but they make it a big part of the scoring.

This fact, piled upon a mountain of other facts can lead a person to no other conclusion than DXO is biased against Canon, and heavily favors Sony/Nikon.

Now, in fairness to Nikon. Their top pro glass, made in Japan, is very good.

No, they don't have the wide selection of choices that Canon does.

But they are so close to Canon in the common lenses, that it doesn't matter. Their 70-200 (focus breathing issue aside) on a D810 or D750 is so sharp and offers so much detail, it is more than just about anyone needs. In other words, it's not the gear holding you back. Stunning images are created with their system.

However, Canon does edge them out in optical quality, autofocus speed etc.


The only way to really compare them is to do so without cameras. That's right, measure the MTF!!! Each company publishes their own scores. But I think was lensrentals or some site like that that got their own expensive testing equipment and published MTF scores. Look it up.

The results on average was that Canon is slightly better in optical quality. But it is very close on most of the pro lenses. Considering the prices Nikon charges, Canon is a tiny bit better value in that regard also.


But who uses just a lens? It's a lens + camera system that produces an image. For that, the best way to test and evaluate is NOT with some BS that DXO cooks up. But rather, by evaluating dozens of RAW image tests done in comparison by VARIOUS different testers.

Only then will you see for yourself a trend one way or the other. Or maybe not at all, indicating they are too close that small variables in testing are making the difference.


All that said, the 5DS with 50MP will make the most of Canon's L lenses and this is already evident in the test images provided across the web.

DXO has to protect Sony/Nikon, so they use literally the only thing Nikon/Sony has an advantage on, and that is low-ISO DR and throw that into the mix for scoring lenses. (rolls eyes)

What if the only advantage Nikon had was shutter speed? That is akin to them factoring shutter speed into lens image quality scores to prop them up against Canon.
DxO hasn't yet tested the Pentax 645Z sensor, or done lens tests for the TS-E_17mm, TS-E_24mm-II or the 200 f/2L. And still no 5Ds-R lens tests. I wonder why...
 
Upvote 0
You call that horrid for an ultra wide lense? Step away from you computer and camera.

Compare that performance to literally ever ultrawide zoom on the market and you'll find it to be vastly superior to anything else - especially considering it is ELEVEN MILLIMETERS.

dilbert said:
romanr74 said:
dilbert said:
The CA on this lens is horrid. It is also sharper at 11mm than it is at 24mm!

To get an idea for performance, click on the measurements ("Sharpness", "Distortion", "Chromatic Abberation", "Vignetting") and then select the "Field Maps" where available.

Can you show this issue in a horrid picture?

http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=977&Camera=453&Sample=0&FLI=0&API=0&LensComp=977&CameraComp=453&SampleComp=0&FLIComp=0&APIComp=3

Check the corners.
 
Upvote 0
LSXPhotog said:
You call that horrid for an ultra wide lense? Step away from you computer and camera.

Compare that performance to literally ever ultrawide zoom on the market and you'll find it to be vastly superior to anything else - especially considering it is ELEVEN MILLIMETERS.

dilbert said:
romanr74 said:
dilbert said:
The CA on this lens is horrid. It is also sharper at 11mm than it is at 24mm!

To get an idea for performance, click on the measurements ("Sharpness", "Distortion", "Chromatic Abberation", "Vignetting") and then select the "Field Maps" where available.

Can you show this issue in a horrid picture?

http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=977&Camera=453&Sample=0&FLI=0&API=0&LensComp=977&CameraComp=453&SampleComp=0&FLIComp=0&APIComp=3

Check the corners.

People have been praising the Nikon 14-24 for the longest time saying Canon couldn't do anything as good.

Interesting that this comparison is largely ignored, and it is with the Canon wide open and the Nikon stopped down.

http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=977&Camera=453&Sample=0&FLI=2&API=0&LensComp=615&CameraComp=614&SampleComp=0&FLIComp=0&APIComp=2
 
Upvote 0
StudentOfLight said:
DxO hasn't yet tested the Pentax 645Z sensor, or done lens tests for the TS-E_17mm, TS-E_24mm-II or the 200 f/2L. And still no 5Ds-R lens tests. I wonder why...

Can't speak to the 645Z or why they've not tested the Canon 200/2L (but have tested the Nikon version).

I presume their lens testing is primarily for the purpose of generating the lens correction modules for DxO Optics Pro, and since the TS-E lenses aren't really amenable to such corrections (due to unencoded lens movements), not testing them makes sense.

For the 5Ds/R lens tests, it's probably just time, or the people that port the data to DxOMark are on summer holidays.
 
Upvote 0
The proof of the pudding .. and all that.

This is the most challenging lens I have, but it´s getting under my skin. Any claim, especially from those who have not used it, that this is a poor performer ... looks stupid. This is plain and simple a fantastic lens, especially considering it´s going all the way to 11mm.

I´m on travel, so no proper post processing capability here, but I thought this image, even though not great in any way, shows a key quality. If I had told you this was a 24mm shot, you would have believed me. This is shot at 12mm! (f6.3, ISO320 on a 5DSR). Feel free look at corner sharpness and CA and whatever else you suspect is wrong.
 

Attachments

Upvote 0
Eldar said:
The proof of the pudding .. and all that.

This is the most challenging lens I have, but it´s getting under my skin. Any claim, especially from those who have not used it, that this is a poor performer ... looks stupid. This is plain and simple a fantastic lens, especially considering it´s going all the way to 11mm.

I´m on travel, so no proper post processing capability here, but I thought this image, even though not great in any way, shows a key quality. If I had told you this was a 24mm shot, you would have believed me. This is shot at 12mm! (f6.3, ISO320 on a 5DSR). Feel free look at corner sharpness and CA and whatever else you suspect is wrong.
Eldar thanks as usual for you insight and willingness to share your real-world experiences with potential buyers. Given your vast experience with Otii, Arts, TS-E, great-white ...etc I can trust your opinion on how un-"horrib" this ground-breaking 11-24 is.

11mm (and even 12mm for that matter) is a whole other ballgame. My natural tendency towards telescopic-vision and lack of patience means I already find it challenging to work with 14mm effectively.
 
Upvote 0
This is the view of Geiranger, one of the most popular tourist sites we have and a World Heritage place.

This is a rather strange use of such a wide angle (for me), but since the vista point further up was covered in clouds, I thought I´d try using this wide angle to create the same view. Not the same thing, but still ...

5DSR, 11-24 @11mm, 1/200s, f8.0, ISO100
 

Attachments

  • _23A1785.jpg
    _23A1785.jpg
    3.1 MB · Views: 327
Upvote 0
dilbert said:
Eldar said:
This is the view of Geiranger, one of the most popular tourist sites we have and a World Heritage place.

This is a rather strange use of such a wide angle (for me), but since the vista point further up was covered in clouds, I thought I´d try using this wide angle to create the same view. Not the same thing, but still ...

It almost feels unnatural to look at such a wide view...anyone else?

This lens does 126.5 degrees at its widest and human (binocular) vision is 114.

Interesting.

I think you will find that is the perspective of the terrain, not a visual aberration. Certainly I have now shot enough at 11-14mm and realise the lens doesn't 'distort' reality to any appreciable degree, but your perspective and use of juxtaposition can. In this image by Eldar, a straight forward scenic shot with no perspective trickery it is impossible to tell the focsl length used without some knowledge of the actual scene.

Look here http://www.canonrumors.com/forum/index.php?topic=24975.msg529944#msg529944 there might be some projection distortion, but no 'unnatural feeling'.
 
Upvote 0
At 11mm you will have strange artifacts at the edges. That is unavoidable, unless the scenery is discontinued at the right places. I agree that it looks unnatural and this looks a bit funny. But I did so at 16mm with the 16-35 f2l8L IS II also. I have shot a number of other shots, where you don´t get that over exaggerated line distortion on the edges.

However, this is a fantastic lens!!
 
Upvote 0
It would be nicer to see the same places shot with 11-24 and 16-35 for instance - most people can get 16-35 even it is not cheap but still more affordable lens. Showing only images from 11-24 - they are amazing but to be able to appreciate it you have to have something to compare to.

______________________________________________________________________________________________________________

www.OakvilleWeddingArtPhotography.com
______________________________________________________________________________________________________________
Canon 70D, Canon Rebel T3i, Canon EF 70-200mm f/2.8L IS II USM, Canon EF 100mm f/2.8L IS USM, Canon EF 50mm f/1.4 USM, Canon EFS 18-135mm f/3.5-5.6 IS STM, Canon EFS 18-55mm f/3.5-5.6 IS USM, Canon EFS 10-22mm f/3.5-5.6 USM, Canon EFS 55-250mm f/3.5-5.6 IS USM, Canon 430EX II
 
Upvote 0
Eldar said:
This is the view of Geiranger, one of the most popular tourist sites we have and a World Heritage place.

This is a rather strange use of such a wide angle (for me), but since the vista point further up was covered in clouds, I thought I´d try using this wide angle to create the same view. Not the same thing, but still ...

5DSR, 11-24 @11mm, 1/200s, f8.0, ISO100

Now that I see this I agree - these corners are HORRID.
 
Upvote 0