Canon EF 14-24 f/2.8L [CR2]

[Jettatore]

I'm not quite sure what to do with those comparisons, with widely different settings. Were you handheld or on a tri-pod for these? Something(s) seems a bit off, I would say hopefully non-offensively, that this samples makes both lenses seem unimpressive (please don't take any offense to that, as none is intended).

What exactly do you mean by widely different settings? If you speak of the flare on the Canon lens -- that's simply b/c the sun had set further during the time between shots on the Nikon vs. the Canon. But sharpness is still quite easy to judge. If you're talking about my different apertures-- that was the whole point: that at f/2.8 the Nikon lens exceeds f/11 performance on the Canon. Additionally, I posted both f/2.8 lenses.

These are all taken on a tripod.

I also don't see anything unimpressive about the Nikon shots... care to clarify? The Nikon shots are pretty much sharp edge to edge!

Your shot looks good, but it doesn't speak to the edge-to-edge performance of the lens. You have continually varying distance of subject from the bottom of the frame to the top, with not much/any detail for comparison at the top (mostly just sky). My tests were designed to evaluate the flat field performance of the lens, something of interest typically to landscape photographers. And my tests are hardly showing anything surprising or new for those of who've actually compared the performance of the Nikon 14-24 against lesser lenses. For example, the guys over at 16-9.net.

Also keep in mind: often a wide angle lens can't maintain focus at infinity across the field (when you've intended to focus at infinity) because of field curvature. This'll often result in closer subjects coming in to focus around the edges if you focused the center at infinity. Therefore your image with subjects of varying distance is not indicative of the ability of the lens to focus on distant landscapes across the field-of-view.

Also remember that a 12MP camera will not show defects as well as a 22MP camera, or 36MP, etc.

Specifically there is a very large difference in shutter speeds between your two comparisons, any little movement or instability on your Canon side of testing could easily impact sharpness to a much larger extent then the Nikon side of the test (I see that all of the cars have motion blur on the Canon test which had a much slower shutter. That's why I asked if you were handheld or not. Also there seems to be a lack of clarity in detail to an extent that I find uncharacteristic of the 16-35 from my own personal use.

I suspect perhaps in camera noise reduction as well as perhaps .jpg shooting are being used (or some other sort of in-optimal post processing), but that's just a guess. Something about the characteristics of the pavement suggest some sort of processing/filtering going on that isn't working out too well. I like to have Noise Reduction in camera turned off completely and shoot in RAW personally.

The Nikon shots in your test are sharper than the Canon shots, but on it's own, outside of this comparison I wouldn't call those results 'sharp'. Chromatic aberrations are all over the shot, hampering it's sharpness to my eyes. To me it's apparent on the street pavement details as well as the safety rail (look for reddish tints if you are having trouble seeing what I'm seeing, it's noticeable on the shadows of the safety rail and elsewhere, at least to me). There's also a lack of contrast in fine details from this image, which comes unexpected at ISO 100.

Perhaps each lens needs to be learned and set individually to pull out it's real world strengths rather than trying to compare them in any sort of controlled or uncontrolled way just to make them compete.

 

[sarangiman]

Specifically there is a very large difference in shutter speeds between your two comparisons, any little movement or instability on your Canon side of testing could easily impact sharpness to a much larger extent then the Nikon side of the test

Easy to disprove... here's the center portion of that Canon f/11 image:

Full-size image here: http://cl.ly/HQcB/Canon16-35_f11_Center.jpg

Tack sharp. Which I'd expect w/ mirror lock-up, 2s delay, Gitzo 2541 carbon fiber tripod w/ RRS BH-40 ball head.

Also, note that I did post the Canon f/2.8 shot (1/250s) side-by-side w/ the Nikon f/2.8 shot (1/320s). Hope you're not trying to say that 1/3EV longer shutter speed screwed over the Canon shot

Also there seems to be a lack of clarity in detail to an extent that I find uncharacteristic of the 16-35 from my own personal use.

As I mentioned, that's due to flare b/c I was shooting into the sun. And these are converted RAWs w/ no tone curve applied.

I suspect perhaps in camera noise reduction as well as perhaps .jpg shooting are being used

Absolutely not. All shots are RAW & processed in exactly the same manner. Shot on the same body (5D Mark III). The only thing uncontrolled about this test is the flare. I could've redone the comparison on a flat cloudy day like I usually do for my tests (not hard to find in Seattle!), but the results were so overwhelmingly obvious that I didn't bother.

The Nikon shots in your test are sharper than the Canon shots, but on it's own, outside of this comparison I wouldn't call those results 'sharp'. Chromatic aberrations are all over the shot, hampering it's sharpness to my eyes.

CA is rather easily removed in LR/ACR. And all my Nikon shots are at f/2.8. CA cleans up by f/5.6. No extra sharpening applied besides the default ACR setting of 25. Out of curiosity, are you used to seeing pixel-level detail on 21-22MP RAW files, or have you only worked w/ the 12MP 5D? Because those examples are pretty tack sharp, especially for edges, at f/2.8, with no extra sharpening applied... I could show you the center, which is even better, but that's not the point of this comparison.

There's also a lack of contrast in fine details from this image, which comes unexpected at ISO 100.

As for lack of contrast, these are RAW files, with no contrast adjustments done at all & just default ACR sharpening (25) applied, so, naturally they are flat.

Perhaps each lens needs to be learned and set individually to pull out it's real world strengths rather than trying to compare them in any sort of controlled or uncontrolled way just to make them compete.

I wanted an unbiased test of sharpness between these two lenses, & that's all this test was. No more, no less. To me, the results speak for themselves, and *everything* was controlled for save for the flare due to the sun setting. If I were publishing this as a review, yes, I would've redone the shots on a cloudy day where the lighting was not an added variable.

 

[sarangiman]

BTW, feel free to PM me if you want links to the original RAW files.

 

[Woody]

This sort of performance is something Canon UWA zoom users can only dream of. Let's hope Canon fixes that w/ this rumored 14-24.

The fact that Canon EF-S 10-22 beats the c**p out of Nikon DX 10-24 gives us some hope that Canon will be able to pull off some magic with this rumored 14-24. Canon's EF-S 17-55 f/2.8 and recent EF 24-70 f/2.8 II further prove they are capable of great things.

The only issue as far as I can tell is cost.

 

[sarangiman]

The fact that Canon EF-S 10-22 beats the c**p out of Nikon DX 10-24 gives us some hope that Canon will be able to pull off some magic with this rumored 14-24. Canon's EF-S 17-55 f/2.8 and recent EF 24-70 f/2.8 II further prove they are capable of great things.

Haven't tried those lenses but I suppose I should mention that I've been quite happy w/ the edge-to-edge performance of my 24-70L (version I). The 70-200 f/2.8L II is stunning. I'm not saying that Canon is incapable of making amazing lenses; that'd be an irresponsible blanket statement. I'm just staying that their 16-35 & 17-40 lenses do not stack up against the competition. Not even close.

Oh, and please, please start using 9-blade apertures in all your lenses Canon.

 

[D_Rochat]

I have no doubt that if Canon produces this lens, it will be nothing short of amazing. I just wonder how long it would be before we see something like this hit the stores.

For using the Nikkor 14-24 on a Canon, is it only af that you lose?

 

[sarangiman]

For using the Nikkor 14-24 on a Canon, is it only af that you lose?

You also have to manually adjust aperture. Metering works pretty reliably for well-lit scenes, but starts to fail in dim lighting or at small apertures (both of which result in less light reaching the exposure sensor, which is why it starts to fail... you have to be within the linear operating range of the meter for it to function properly).

I've found that metering is more reliable in Live View mode, but also fails at very small apertures or when the lighting is extremely dim. That being said, I haven't found this to be a huge issue b/c we're not shooting film here You can re-evaluate the exposure after you've taken the shot!

The Novoflex adapter that allows you to control the aperture will also set you back ~$250 or so.

 

[Jettatore]

sarangiman,

This is getting too complicated for my tastes so I won't fully respond. Thank you for the clarifications/corrections on your settings, though I don't feel that even the center Canon shot you posted is tack sharp or any representation of what the lens is capable of, I have this lingering hunch that somethings altogether are off in both images... I am used to both 18MP crop body shots and 12MP full frame shots. I would like to see the RAW file of the Canon if you wouldn't mind sending it.

 

[sarangiman]

though I don't feel that even the center Canon shot you posted is tack sharp or any representation of what the lens is capable of, I have this lingering hunch that somethings altogether are off in both images...

Yeah I'm not sure how to respond at this point either, since that center crop is pretty much as sharp as I see for any lens at an ACR sharpening setting of 25 (default). I have a feeling you're not liking it b/c of the flare... but I just sent you a PM w/ both RAW files, so you can sharpen & adjust contrast to your heart's content

 

[Jettatore]

Thank you sarangiman, cheers.

 

[Etienne]

Thank you sarangiman, cheers.

Good god ... you guys can go on forever analysing one photo ...

... there are a ton of reviews on both of these lenses, much more in depth than this one photo taken at different settings can ever provide.

Go to photozone.de for a starter.

It's no secret that the Nikon is sharper than the Canon. But it also flares more easily, it's heavier, more expensive and it doesn't go to 35 mm. The Canon is cheaper, lighter and more versatile.

Pick your weapon and get to work.

Otherwise you'll have to wait for the Canon to see how it performs.

 

[sarangiman]

... there are a ton of reviews on both of these lenses, much more in depth than this one photo taken at different settings can ever provide.

How were my photos taken 'at different settings'?

Go to photozone.de for a starter.

Yup, photozone's numbers are great, & I'm a firm believer in quantitation. But side-by-side images can sometimes tell you what those numbers mean in your imaging system. Which is why I performed the test on my own to begin with. It's just one way of representing the data, & was particularly helpful to me anyway.

 

[Radiating]

This would be a seriously amazing lens to have from Canon.

Combine that with releases of a 35mm 1.4L + 50mm 1.4L & 135mm 1.8 L IS and I would be totally content.

 

[aznable]

The fact that Canon EF-S 10-22 beats the c**p out of Nikon DX 10-24 gives us some hope that Canon will be able to pull off some magic with this rumored 14-24. Canon's EF-S 17-55 f/2.8 and recent EF 24-70 f/2.8 II further prove they are capable of great things.

The only issue as far as I can tell is cost.

and it's beaten by tokina and sigma latest wide angle aps-c zooms...so sigma and tokina are better than canon and nikon in order to deliver aps-c wide zooms? dont think so; it depens which are specs/target price they want to deliver to market.

when they wanted to deliver a superior 70-200 they did, i amconfident they are doing with the new 24-70 and maybe they will do with a fast wide angle zoom...of course the superior quality comes at a PRICE, but this is true for every manufacturer

 

[Ricku]

This would be a seriously amazing lens to have from Canon.

Combine that with releases of a 35mm 1.4L + 50mm 1.4L & 135mm 1.8 L IS and I would be totally content.

Yes, but 50mm 1.4L?

1.4?

 

[dirtcastle]

While it's always appreciated to see example shots, I think we should be able to agree that one person's images don't make a definitive argument for or against a lens.

Even DxO is controversial.

 

[Woody]

and it's beaten by tokina and sigma latest wide angle aps-c zooms...so sigma and tokina are better than canon and nikon in order to deliver aps-c wide zooms?

That's not my point. I am trying to repudiate the common belief that Canon is incapable of producing decent wide angle lenses.

 

[briansquibb]

and it's beaten by tokina and sigma latest wide angle aps-c zooms...so sigma and tokina are better than canon and nikon in order to deliver aps-c wide zooms?

That's not my point. I am trying to repudiate the common belief that Canon is incapable of producing decent wide angle lenses.

I understand.

Try the TSE-17 and TSE-24 ....

 

[aznable]

and it's beaten by tokina and sigma latest wide angle aps-c zooms...so sigma and tokina are better than canon and nikon in order to deliver aps-c wide zooms?

That's not my point. I am trying to repudiate the common belief that Canon is incapable of producing decent wide angle lenses.

yes i know you didnt want to say that...i tried to advance one or two step further reasons of your, in my bad english (sorry but i am italian and i am not talented for foreign languages)

 

[Radiating]

This would be a seriously amazing lens to have from Canon.

Combine that with releases of a 35mm 1.4L + 50mm 1.4L & 135mm 1.8 L IS and I would be totally content.

Yes, but 50mm 1.4L?

1.4?

Maybe a 1.3, but no manufacturer has ever been able to create a sharp 50mm lens that's faster than f/ 1.4 for full frame. Canon would be well off to lower the aperture and increase the image quality like they did from the 50mm f/1.0 L.