Refer to this page as http://www.iki.fi/leopold/Photo/CanonEF24mmF14L2/
![[24mm f/1.4L vs Mark II from side]](090616-210526_2.jpg)
During the last year Canon has revised several of its wide
angle Full Frame lenses as well as introduced some lenses
completely unheard of in the 35 mm world, the best example
being the TS-E17mm f/4L lens that has no equivalents anywhere,
period.
Being a lens without equivalents is of course a description that has always fit Canon's EF24mm f/1.4L USM as well. It has been by far the world's fastest auto-focusing lens in the wider than 35 mm category. Nevertheless, late last year Canon saw it fit to update the then 11-year old design with a new version.
Just as its predecessor, the Canon EF24mm f/1.4L USM II is at the moment, along with the older version I, the only auto-focusing lens that offers a 17 mm aperture and a diagonal field of view of 84 degrees (35 mm Full Frame camera needed). Thus, it offers a unique perspective with some Depth-of-Field (DoF) control especially when shot wide open at close subjects.
After five months of agonizing waiting I was finally able to get my hands on the new 24 mm lens this June. I was fortunate enough to have some time to test my old, original EF24mm f/1.4L USM along with the Mark II, and thus I am now able to show some direct comparisons between these two excellent lenses.
This article doesn't even attempt to be a one-in-all full test of the new Canon lens. It's just a collection of my own experiences. They may be useful, they may be not. For rigorous tests with lots of numbers, have a look at The Digital Picture or some other proper lens test site.
![[24mm f/1.4L's side by side]](090616-180231_2.jpg)
My first impression was in the line of "Wow, this thing is huge!"
From looking at numbers and pictures I had expected the Mark II to be
slightly larger than Mark I. But the difference was larger than
I expected - particularly when the lens is used as it should: with the
lens hood on.
I don't know if the photo on the right comparing the old and new 24 mm
lenses side by side really shows the difference, but believe me: it's there.
The upside of this is that I believe the new, larger lens hood
to be a better lens protector against noses and fingers than the old one.
The new 24 mm feels every bit as solid as the old one. It is heavy, it is large, it is expensive, there is no play at the manual focus ring, auto focus is really fast and quiet and all feels good. All in all, every bit of it is pure L and the lens balances very well on the Canon EOS 5D Mark II (and, yes, the old 5D, too).
Other than the obvious size difference, there are little to separate
the two lenses. The most important of these differences might be the
weather seal rubber ring on the Mark II: where you see a metal ring
on the original 24 mm, you'll see black rubber on the Mark II.
![[24mm f/1.4L vs Mark II from front]](090616-174052.jpg)
When looking at the front of the two lenses you will see that
the Mark II (on the right side (no, it's not the eyes of Wall-E))
had more of the really black
stuff inside. I'd expect this to be one one the several methods
Canon uses to fight flare caused by internal reflections.
As with the original version, the Mark II is a joy to focus
manually when using the Canon 5D Mark II's Eg-S Super High Precision
Matte focusing screen (which I recommend wholeheartedly
to large aperture shooters).
Manual focus was used all the time, and each photo was taken only once.
![[Sharpness test 1]](24mmF14Mk2_Test3_F040.jpg)
To the right is a scene that I am going to show crops of.
My original intention was to show some crops of this scene
at every full aperture between f/1.4 and f/22. However, I
slightly goofed exact focus (still getting the hang of Live View).
Thus, only two images for this demonstration,
taken at f/4 and f/8, then losslessly cropped from various parts
of original camera JPEGs will have to do.
| Tests at f/4 | |||
|---|---|---|---|
| EF24mm f/1.4L USM | EF24mm f/1.4L USM II | Comment | |
| Crop 1 f/4.0 | 
| 
| No big difference here. The Mark II may appear slightly less sharp, but that is upto my focusing error. Sorry. |
| Crop 2 f/4.0 | 
| 
| A bit suprisingly there is no big difference here, either. I would have assumed that there would be a significant difference in chromatic aberration this far from image center. This is a good reminder that the original 24 mm is every bit as good as it was before the Mark II was published. |
| Crop 3 f/4.0 | 
| 
| If there wasn't much of a difference in Crop 2, here the new lens has a definitive advantage. For some people this is reason enough to update. Some other people might just turn the chromatic aberration correction knob on their favourite RAW converter up a bit. |
| Crop 4 f/4.0 | 
| 
| The extreme crop shows two things. First, the new 24 mm
lens has less chromatic aberration and it is sharper.
Second, there's still some CA left.
So, while the new one is better, it is still
not pixel perfect - at least not at f/4.
The better corner sharpness isn't an accident: particularly at apertures between f/2.8 and f/5.6 it was a recurring theme in my other tests. |
Let's see how well the lenses do at f/8.
| Tests at f/8 | |||
|---|---|---|---|
| EF24mm f/1.4L USM | EF24mm f/1.4L USM II | Comment | |
| Crop 1 f/8.0 | 
| 
| At f/8 it is hardly suprising that the lenses look exactly the same in the center. |
| Crop 2 f/8.0 | 
| 
| While chromatic aberration, which is most easily seen as incorrect purplish colours at harsh contrast transitions, are less disturbing with the new lens, the difference isn't actually that huge. |
| Crop 3 f/8.0 | 
| 
| As with the f/4 tests, the biggest difference between the two lenses can be seen in Crop 3 where the original 24 mm lens shows quite a bit of chromatic aberration while the Mark II is capable of offering a remarkably clean performance. |
| Crop 4 f/8.0 | 
| 
| As with f/4, the Mark II is slightly better near the corner, but the difference is by no means huge. Still, this might be important for a landscape photographer. |
Hmmh. The new lens is obviously better than the old one, but one thing must be said: the original EF24mm f/1.4L USM has never been a bad lens.
| Tests at f/8 | |||
|---|---|---|---|
| EF24mm f/1.4L USM | EF24mm f/1.4L USM II | Comment | |
| Crop 1 f/1.4 | 
| 
| At f/1.4 there doesn't seem to be any huge differences between
the two lenses. However, there is an elongated flare blob
right to the left of the tree trunk with the old lens,
and contrast is ever-so-slightly lower.
(By the way, although it may look like it, the funny colouring on the building is not moiré. That's the actual colour pattern of the tiles.) |
| Crop 1 f/2.8 | 
| 
| At f/2.8 the old lens' flare blob on the left of the tree trunk has grown sharper and the contrast difference is slightly easier to see. |
| Crop 1 f/5.6 | 
| 
| By f/5.6 there are several visible yellowish light blobs with the old lens while the Mark II only starts to gather greenish loss-of-contrast mush. I may have inadvertedly helped the newer lens by choosing a green scene: its green flare artifacts will not show as easily as the old lens' more yellow blobs. |
| Crop 1 f/11 | 
| 
| By f/11 it is clear that the new lens handles flare quite a bit better, which is no mean task considering that the old one is also good. |
| Crop 1 f/22 | 
| 
| This final image makes it clear that the Mark II is noticeable better in the flare department. It is not 100 % clear cut, though: if you are shooting something where yellow-purple flare would look less distracting than green flare, you'd might think differently. |
From everything I've read on the Web the new lens, with its new optical design and rounded aperture blades, should be noticeably better. Thus, I ran two different tests at apertures between f/1.4 and f/22. Neither of these tests managed to show any significant difference at any aperture. From what I can see, any differences are subtle enough that I can't tell the two lenses apart without looking at the EXIF.
![[24mm f/1.4L Mark II vignetting at f/1.4 compensated]](Vignetting.gif)
Vignetting between the old and new 24 mm lenses is close
enough that I consider them identical. In other words, vignetting
is very high in image corners when shot wide open, although the
vignetting correction of the 5D Mark II does a good job of making
it vanish. To the right is a very exciting brick wall that has been
shot at f/1.4 with the 5D2's vignetting correction both on and off.
I've never minded vignetting, but I'm not going to cry a
river over its disappearance.
Corner darkening is, of course, only one phenomenon
caused by vignetting, and the easy one to correct. Worse and/or
weird bokeh in the corners is an equally nasty follow-up
and, unfortunately, not correctable with software.
But, all in all, this lens is no better and no worse than the
old one in this regard.
![[Sharpness test 2]](24mmF14Mk2_Test5.jpg)
Now that we've established The Most Exciting Brick Wall in the
Whole Wide World(TM), why not make some
extra use of it? After some consideration I decided to publish these
results, too, even though I cannot be 100% extra double sure that my
focus was absolutely right. Then again, this is sometimes a problem
with the best professional camera and lens reviews, so why worry?
As you can see, I have chosen three crops: one at the center, one
little over halfway to the corner, and the absolute corner.
Shooting distance is somewhere around 3 meters and I at least
think I was reasonably perpendicular to the wall. As you see,
there are many caveats, but I did do my best to avoid a
crappy test.
| Crop 1: The Center | |||
|---|---|---|---|
| EF24mm f/1.4L USM | EF24mm f/1.4L USM II | Comment | |
| f/1.4 | 
| 
| Mark II has a clear advantage here: the old lens shows slight
halation around all details which washes out the contrast.
Still, back in my head is the thought: is my focus absodiddle-lutely correct? Even the slightest error in focus could cause this. |
| f/2.0 | 
| 
| By f/2 both lenses gain lots of extra contrast and particularly the halation of the older Mark I diminishes greatly. Still, Mark II is slightly sharper - or at least it has slightly more contrast. |
| f/2.8 | 
| 
| At f/2.8 I'd be hard pressed to be able to tell these lenses apart. They are so sharp you could cut your eyes with them if you looked too closely. |
| f/5.6 | 
| 
| By f/5.6 both lenses look exactly the same and they are also virtually identical to the f/2.8 images. |
| f/11 | 
| 
| When closed down to f/11 both lenses start to lose contrast so much that both look worse than at f/2. This is something to remember before you stop down to unnecessarily small apertures. |
| f/22 | 
| 
| At f/22 both lenses look identical and significantly worse than wide open. |
Crop 1 condensed results:
Mark II is better upto f/2; after that both
are extremely good. However, is the difference real or a result of
incorrect focus? Other tests seem to indicate it's real, but there
is a danger of doing what is expected instead of what is right.
Remember what happened with the
N rays.
| Crop 2: Halfway to the Border | |||
|---|---|---|---|
| EF24mm f/1.4L USM | EF24mm f/1.4L USM II | Comment | |
| f/1.4 | 
| 
|
Both lenses might have looked good in the center at f/1.4, but
the same is not true for crop 2. Here both lenses are equally
soft.
Again there are the questions of if this is an effect of field curvature, but as I don't have the Mark I anymore there is no way to retest. |
| f/2.0 | 
| 
| Both lenses get sharper by f/2 and they look very similar to each other. |
| f/2.8 | 
| 
| Although both sharpen at f/2.8, the image is still far from clear, so to speak. Again, there are no significant differences between the lenses, except that chromatic aberration of the older lens makes tile pink bleed a little bit at the edges. With stronger colours and higher contrast this bleeding would be more obvious. |
| f/5.6 | 
| 
| Both lenses sharpen up nicely by f/5.6, and like before, they look pretty much the same, except for the pink bleed of the older lens. |
| f/11 | 
| 
| Here something interesting happens: while the older version looks exactly the same as at f/5.6, the Mark II sharpens up considerably. Now there is a significant difference: the Mark II is clearly sharper and misses the chromatic aberration of its older brother. |
| f/22 | 
| 
| As before, f/22 kills all quality we gained during the four previous stops, so again I'd warn against using it unless you really like to play with sharpening tools for that digital look. |
Crop 2 condensed results:
From f/2.8 to f/5.6 the Mark II is slightly better, but at f/11 the
difference is significantly larger.
| Crop 3: Absolute Corner | |||
|---|---|---|---|
| EF24mm f/1.4L USM | EF24mm f/1.4L USM II | Comment | |
| f/1.4 | 
| 
| Unlike with previous crops, here there is absolutely no
contest: the Mark II looks pretty bad, but not nearly
as bad as the older lens.
This is why landscapes generally shouldn't be shot wide open. |
| f/2.0 | 
| 
| There is absolutely no difference between f/1.4 and f/2 in terms of sharpness. The only thing that has changed is that these samples have gotten slightly brighter. This goes on to prove that when changing aperture between f/1.4 and f/2 there is no more extra light to the absolute corners because of vignetting. |
| f/2.8 | 
| 
| At f/2.8 both lenses start to gain a bit of extra sharpness and contrast. Although they still look pretty soft, Mark II is clearly ahead. |
| f/5.6 | 
| 
| While neither lens is perfect at f/5.6, Mark II has gained
sharpness pretty nicely and is completely usable. Only
the last 60 or so horizontal pixels show extreme
softness.
By these apertures it is clear that we are looking at the actual performance of the lenses and not just my potential focusing mistakes. There clearly are applications where the new lens would be acceptable at f/5.6 while the old wouldn't be. |
| f/11 | 
| 
| While Mark I still gains some extra resolution, Mark II
is already suffering slightly from diffraction.
But it still is clearly ahead of Mark I.
The behaviour of Mark II feels a little contradictory: Crop 2 still got sharper at f/11, while Crop 3 already suffers at the same aperture. I don't have explanations for this odd behaviour. |
| f/22 | 
| 
| Diffraction hits both lenses hard, but not hard enough for the Mark II not to look better. It now looks as good (or bad) as at f/2.8. Mark I still looks a great deal better than at f/2.8 and the absolute corner even looks better than f/5.6. |
Crop 3 condensed results:
There are no subtle differences here: Mark II wins hands
down at all apertures. Somewhat contradictory to Crop 2 Mark II doesn't
enhance from f/5.6 to f/11, which may leave a landscape photographer
scratching his head. My suggestion is that you probably just have to
test your lens at all apertures between f/5.6 and f/11 and see which
one suits you best. As always, there can also be copy variations
although they should be relatively small considering this is an
expensive L prime lens.
![[Eclipse 2006 with the original 24/1.4L]](060329_Pimennys2.jpg)
The EF24mm f/1.4L USM II is a great lens. From colours,
through contrast, sharpness and chromatic aberration, it's all
good. For someone who wants to get into the 24 mm low-light
bandwagon there is no better way than to grasp this lens.
Having said that, if there is always infinite amounts of time and
enough light, Canon's new and in its own ways superb manual focus
TS-E24mm f/3.5L II seems to have even better image quality,
particularly as it reportedly doesn't know what the words
"chromatic aberration" mean.
However, the original EF24mm f/1.4L USM is also a great lens - almost as good as the one replacing it. The biggest difference between these two lenses lies not so much on the fast aperture range but at the aperture range mostly occupied by landscape photographers. Here, nearing the corners, the difference may even be called dramatic.
Whether it is worth to update the old one with the Mark II is a difficult question. I updated mine without as much as seeing the new lens, but I also managed to order my copy just before the almost universal price increases and thus my friendly web shop sold it to me for over 200€ less than what it would have cost on the date of delivery. All in all, with the old lens sold for a reasonable price I only lost a few hundred euros with the update.
Still, I'm left wondering whether this really was what I needed. There seems to be some difference between the two lenses when used wide open on a brick wall. That test seems to indicate that the new version is sharper from center to corner, at least at certain apertures. (Then again, I might have botched the setup for this test and now I don't have a chance of repeating it.) Vignetting looks the same on the new lens, as does bokeh. Corner sharpness and chromatic aberration hardly matter when you shoot people at f/2 but a landscape photographer will think differently.
However, if it's sunny in China in a few weeks when the 2009 Total Solar Eclipse is going to take place (090820: it wasn't), and if I manage to take a good picture of The Great Darkness (090820: I didn't, but I did get a picture), I'm certain I'll feel the new lens was worth every penny (look to the right to see the image I took with the 5D+24/1.4L in 2006 that I'm trying to top (and yes, the TS-E24mm would be better suited for that particular image, but I am not going to buy that, too (090820: my new picture isn't even close to topping that one))).
To finish this article, here are my pluses and minuses for the Mark II as compared to the old version. In both lists, the points most important to me are listed first:
| + | Less chromatic aberration |
| + | Significantly better corner sharpness when stopped down |
| + | Better flare resistance |
| + | Slightly better wide open |
| - | More expensive |
| - | Surprisingly much bigger |
| - | Heavier |
Although I know this article is by no means an exhaustive test report of Canon's EF24mm f/1.4L USM II, I hope it has offered some insight into what the new lens is, what it's good for and how it compares with its predecessor.