I made a comparison last month between my 5D3+24-70L II and a friend's D800+24-70 AFS.
Here are some screenshots of the images shot at 24mm f/11 to virtually eliminate slight small differences in focussing which could happen at wider apertures, even using LiveView (which incidentally is quite badly implemented in the D800).
Tripod, LiveView, self timer, etc.
NEF and CR2 converted in ACR with zero sharpening and no lens correction.
I then upsampled the Canon file to 36 Mp with bicubic standard and then I applied SmartSharpen (LensBlur) at 200/0,3 for the Nikon and 300/0,3 for the Canon, to partly compensate the upsampling which softens a bit the image.
Differences at f/11 are small but considering the Canon is upsampled I still rate its IQ slightly higher, more so at wider apertures.
Still the 24-70 G is a wonderful lens and while it is heavier and longer it is amazingly well built.
Hi Marco,
Thanks for that direct comparison between the two lenses. Again, those shots were made at f/11, right? Frankly, I am surprised at the difference in presence/absence of the LCA (purple fringing) in the upper left hand corner between the two at that aperture. You further said that the Nikon lens is longer at 24mm? I didn't quite understand that but in other words, the Canon lens is wider at 24mm then?
It reveals what turned out to be a massive disappointment when I started looking into Nikon lenses. Lateral CA really leaves a lot to be desired compared to Canon equivalents, and it goes through the lens lineup. Yes, it can be corrected but not without affecting critical sharpness and negating the resolution advantage of the D800.
This is one of the main reasons why i only use alternative manual focus lenses for my D800E.
1st The Nikon CA is horrible ( did I really ude this therme in an other thread? ).
2nd I would never upsample an image to a higher res. I would do the oposite to compare.
That those sample from the Canon still looks as good or (to me) better shows HOW good it is.
3rd. I would have loved to see the comparisment wide open. Those lenses offer 2.8 and I would like to see how usable they perform there.
I would think f/11 would level the IQ playing field, and that the D800's extra MP rez would slant the results in its favor.
The opposite is true -- even at f/11 (I can only imagine the results compared wide open!), the aberrations are far greater in the Nikon, and it sharpness if far reduced. Even with the superior Nikon (Sony) sensor, the 5D3 with 24-70II is the clear winner in resolution, sharpness, and color.
Any head-to-head outtakes from wide open shots? I've been quite pleased with the shots I got from Canon's when wide open.
A fruit for thought.... could diffraction that becomes more prevalent at a higher MP camera play a role regardless how minor in that comparison? I don't know how complicated it would be but seriously, a direct comparison at wider aperture settings would make it definitely very interesting. Of course, we know about the LCA aspect already but I am referring more to the resolving power. Maybe your friend would be less willing to let you borrow his camera/lens now ?
Both have similar FL at 24mm setting but the Nikon at close range keeps the longest focal lenght more than the Canon (I'd say 70mm vs 65mm of the Canon below 1 meter focussing distance).
@Ralph, I upsampled the Canon just to reveal how little difference there is between the two, partly because of the 24-70L II higher quality but partly because of the very good sensor in the 5Dmk3 which allows for some nice upscaling.
As for the wide open comparison I could provide it, but there's not enough dof to correctly judge corner sharpness. For instance the Nikkor is definitely more blurred in the lower right at f/2.8 but maybe I focussed that lens just a tiny bit backward compared to the Canon, so this could aggravate the issue.
Anyway I'll try to find the crops wide open and post here.
Consider also that some CA could be easily corrected in post.
This doesn't change the fact that the Canon has almost none compared to the Nikkor though.
I don't have access to the raw files here (and maybe I also delete them...) but this is a screenshot I made of the lower right corner wide open for both cameras (Nikon on the left, Canon on the right).
It is way overexposed because I was commenting in another forum about the excellent shadow recovery of the mk3 once you shoot in "ETTR-mode" (expose to the right, meaning overexposing without clipping highlights).
To exagerate the issue, once I normalized both exposures (5D3 originally shot at +0.7, D800 at -0.3 exposure), I pushed the exposure slider in ACR by 4 stops and I found that even with that unreal pushing the noise between the two cameras is almost the same.
Marco wrote:
Joshua sorry for the confusion, with longer I meant that it is a bigger lens... as you can see here...
Got it, thanks! Well, I should have known that since I had that Nikkor lens before and it uses the same or at least similar lenshood system as that of the Canon 24-70mm Mark I, which in my opinion is quite effective in shielding the lens for flare.
AGeoJO wrote:
Got it, thanks! Well, I should have known that since I had that Nikkor lens before and it uses the same or at least similar lenshood system as that of the Canon 24-70mm Mark I, which in my opinion is quite effective in shielding the lens for flare.
Yep, agreed.
I like the small form factor of the new lens hood, but the older works better indeed.
Although the new lens is very flare resistant from my (limited) experience.
Thanks for these wide open crop shots, Marco. It really helps illustrate the superior performance of the new Canon lens.
Regarding the hood: I found the same. The original Canon v. 1 hood and zooming mechanism is better at shielding the lens from flare. But like you guys, I found the new Canon lens is almost impervious to flare even with the inefficient new hood (it does block most stray light, and acts as a convenient lens protector). The old one is enormous and clunky, but does its job and helps the old v. 1 lens hang in there.
BTW -- I just sent my original 24-70 to Canon after comparing to the new lens. Mine was pretty close in quality at the center (and edges, at some FL), so I'm hopeful a small adjustment can be made to improve its IQ. Just trying to eke out another couple years. But the original took a pretty stunning defeat in overall IQ (using center point AF, switching between 5D2 and 1Ds3) to the new version.
Gunzorro, v1 is still an amazing lens... while the v2 has nice bokeh I'm stunned by some images from the 24-70 v1 with regard to threedimensionality (is it a word..?).
I wouldn't be in a hurry to change a good copy of the v1 for the v2 and your choice to make it serviced is indeed a very sensible one.
.
