Nikon 24mm f/1.4 Review (Part Two: Optics and Sharpness)
[This article was originally published on MiddleSilver, my old photography website, on January 14, 2014. The article below is a revised version posted to spencercoxphoto.com on August 30, 2014.]
This is the second part of an in-depth review of my favorite lens, the Nikon 24mm f/1.4 lens. This section covers the sharpness, color, distortion, and general image quality of the lens. In short, this is the sharpest wide-angle lens I have ever used, and it has very few optical flaws. I don't want to give everything away here, though, so keep reading my extensive review for more detail!
Part one covers the build quality of the lens.
Part three is a summary of the lens, plus a comparison to others.
One thing that I like is that, because this lens is a wide angle, a decent amount is in focus even at f/1.4, so long as you focus on something somewhat far away. Just look at the photo below (focus was on the boat).
(Click to see on black :: Link to photo in gallery)
The maximum reproduction ratio is rated at 1:5.6, meaning that, at its closest focusing distance, FX users can fill the frame with an object roughly eight inches in size (20cm), while DX users can fill the frame with objects roughly five-and-a-half inches long (14cm).
This is surprisingly good for a wide-angle lens, and it can give your photos a unique perspective (as in the below photo).
As you can see above, the 24mm f/1.4 is pretty perfect. Granted, the sun in this photo is slightly diffuse because this was taken right at sunrise, but it's still amazing performance.
However, when the sun is slightly outside of the frame, it's common to get a little rainbow dot somewhere in the frame. This dot is small and sometimes faint, but it can be extremely irritating. Because of its size, it can be hard to spot in the viewfinder. If you notice, it's easily correctable by moving your hand to the side to block the sun. In theory, the lens hood should eliminate this, but that doesn't always happen.
Luckily, in my photos, the sun has been the only cause of flare that I can find. Even bright city lights at night do not cause flare with this lens, which is a remarkable result for a wide-angle lens. Landscape photographers will find this to be a major benefit over most zoom lenses.
Flare becomes more defined as the lens is stopped down (as shown below). Also, when the aperture is extremely small, this lens is prone to producing a red rainbow pattern between the beams of the light source itself. This is unrelated to flare, but instead results from the interaction of a bright light with the aperture blades of the lens.
Note that the below is an extreme example, of course, which is why there is noticeable flare in the photo. That said, this is much better than most lenses would have done in this situation, especially in terms of the photo's contrast.
The 24mm f/1.4 produces a light-blue flare overlapping harsh light sources near the corner. The example below is a crop from the top-right. The rest of the photo has absolutely no visible flare. Note that this also becomes more distinct as the lens is stopped down.
See the blue dot at the top-right of the sun? That's microflare.
Luckily, microflare is, by definition, micro. It doesn’t really have an effect on most photos, but it would be nice to find a lens that can handle it well. The 24mm f/1.4 doesn't do badly (especially considering its near-perfect flare performance otherwise), but it's not great either.
I figured that this may be my imagination, so I took some sample photos. Look at the samples below to see what you think, but for me it’s enough to prove that there are real differences in color between lenses. Personally, I think that the 24mm f/1.4 produces better colors than the 17-55mm f/2.8.
To flip between the two photos, click on one to see it full size, then switch back and forth between the two photos. This is easier on a computer with a keyboard than it is on a phone.
Both of the photos are unprocessed, RAW photos with identical white balance and tint settings. The camera profile in Lightroom was set to “camera neutral” to show the amount of unprocessed color information captured. The actual lights were very bright and saturated. Clockwise starting from the greenish light at the top left, the colors of the four main lights were supposed to be green, orange, red, and blue.
To me, there is a clear difference between the blue light at the bottom left of each photo. On this light, the 24mm f/1.4 renders more of a traditional blue color than the 17-55, which renders the light with more of a cyan tone.
The green light above it shows more minor differences, but again the 17-55 is slightly more cyan, while the 24mm f/1.4 is closer to the true green color.
The bottom-right red light seems identical to me in both photos.
The orange light at the top-right also shows minor differences, but to me it is a very useful test: the 17-55's rendering is the slightest bit greener than the 24mm f/1.4. It's hard to tell even flipping between the two, but there are definitely differences. Because most indoor light is orange or yellow, this difference is important. If saturation were higher, and if the orange took up more of the photo, the differences would be even more strongly in favor of the 24mm f/1.4. When oranges and yellows are too green, it becomes a challenge to make everything look "right."
And is it just me, or is the 24mm f/1.4 photo more contrasty (with deeper shadows)?
There are three total ways I look at bokeh: the out-of-focus bits in front of the subject, the out-of-focus bits behind the subject, and the transition bits that are just barely out of focus.
On the bright side, the background bokeh of this lens is great. Not just great for a wide angle lens, but truly great. Everything out of focus behind the subject fades into a pleasant, though not unrecognizable, blur (see above photo). There are just a couple slight issues: Some out-of-focus edges turn slightly green-colored, and out-of-focus corners appear slightly stretched-out. These are minor issues, though, and the overall background bokeh is really good.
Despite the great background bokeh, the foreground bokeh isn't perfect. Lenses typically have to be optimized for either foreground or background bokeh, and this lens is certainly better at the latter. That said, who even has out-of-focus foregrounds in wide-angle shots anyway? The crab photo above is a decent example, but that's the best I've got.
The transition bokeh of this lens is good, although transition detail sometimes looks a little hazy, almost as if the clarity slider in Lightroom had been lowered. This only happens at f/1.4 to f/2, beyond which, the transition bokeh is pretty perfect (although good luck getting much out of focus at f/2.8 and 24mm). This hazy blur really gives this lens some character, though, so it's not always a bad thing. In front of the subject, the transition blur is actually fine, though vaguely purple-tinted. Only behind the subject does it become somewhat hazy. This is mainly noticeable towards the lens's closest focusing distance.
To me, it seems to me that Nikon intentionally left some longitudinal chromatic aberration uncorrected in the background to make the bokeh smooth. This makes backgrounds smooth, but the side effect is the lack of “clarity” above and the slight color fringes in the background blur.
The photo below is a torture test which shows the differences between foreground and background bokeh, as well as color fringes. I have never seen it this bad in real pictures.
In total, I have printed five photos on metal from this lens: a 24x36 of the Eiffel Tower (taken at f/1.4), a 20x30 of the cathedral photo (f/2.5), a 16x24 of the graffiti photo (f/1.4), a 16x23 photo of the canal photo under the "aperture" heading (f/1.4) and an 8x12 of the alleyway photo under the "colors" heading (f/1.4). See a pattern? Every single one of the apertures was below what an f/2.8 zoom could achieve.
(And no, the 16x23 size for the canal photo is not a typo. I ordered a custom size because that was what fit my crop of the photo!)
Here's a quick picture of what four of them look like (the special-order photo hadn't arrived yet).
While the above photo obviously doesn't show the full sharpness of each print, it makes my point clear: I trust this lens to deliver sharpness good enough for prints this large, and I have not been let down.
It is important to note, however, that this lens is not that much sharper than, say, the 14-24mm f/2.8 or 17-55mm f/2.8 when shot at f/2.8. Yes, a $2000, two-pound zoom lens can produce images nearly as sharp as the 24mm f/1.4. This is especially true in the center. In the corners, the 24mm f/1.4 pulls away from Nikon's zooms slightly. Below is a comparison with the 17-55mm f/2.8 zoom. (Though somewhat old, the 17-55 is generally agreed to be among Nikon’s best zooms, retailing new for $1400). This is a DX-only comparison of an image.
The below shots are unsharpened raw photos from the D7000 with chromatic aberration, distortion, and vignetting corrected (this section focuses only on sharpness tests, and I would normally correct those anyway). The crop is equal to 1/16 the area of the photo. The first set is from the center of the photo, whereas the second set is made up of crops from the top-left. Framing isn't identical because trees are living, breathing, moving things.
If you're not using a phone, it is a good idea to click on these photos to view them full-screen.
These are the f/4 crops shown above, blown up larger for a better comparison:
The center differences become somewhat noticeable at this size, especially if you click on the photos and flip between them, but they're still subtle.
In the corners, though, the differences are obvious. The 24mm f/1.4 is literally as sharp at f/1.4 as the 17-55 is at f/2.8. By f/4 or f/5.6, the 17-55mm f/2.8 lens hasn't improved much, whereas the 24mm f/1.4 looks simply amazing.
Below are f/4 crops enlarged to see differences more easily.
The corner differences are obvious. The 17-55 has poor contrast and detail compared to the 24mm f/1.4. This is one major advantage of the 24mm f/1.4: corner performance.
Note, also, the color differences between the photos. For me, the 24mm f/1.4 looks better, no contest! I've already gone over that in the colors section, but here's even more proof. The blues actually look blue with the 24mm f/1.4, as opposed to cyan with the 17-55mm f/2.8.
It is important to know that, even though the f/1.4 shot looks pretty fuzzy, that's pretty great for a photo at f/1.4. It's definitely sharp enough for a large print on acrylic or metal. Of the five prints I mentioned above, remember that four were shot at f/1.4 (including the 24x36 photo of the Eiffel tower, which had most of its detail along the edges and corners).
The sharpest apertures on the 24mm f/1.4 in both the center and the corners are f/5.6 and f/6.3, where sharpness and contrast are as high as any lens I’ve ever seen.
At f/11 and f/16, you will start to see contrast reduction throughout the image due to diffraction, and slight loss of detail at f/16. Although contrast can be increased to a degree with sharpening, doing so increases grain and could make the photo look over-processed. For this reason, I tend to avoid f/16 if possible unless I have a specific image goal in mind. Keep in mind, though, that the diffraction monster isn't that bad; at f/16, I'd say that the images are similar to the f/1.4 photo's center in terms of overall sharpness. If you're happy printing an f/16 photo large (I am), you'll probably be fine printing an f/1.4 photo large as well.
Some people online have reported issues with focus shift (saying that the lens focuses slightly differently at different apertures). I did not encounter this issue.
The sunstars on this lens are amazing! Pinpoints of light captured by this lens become beautiful eighteen-pointed stars, thanks to the lens’s nine-bladed diaphragm. Pinpoints of light can turn into rays even starting at f/2 (where the rays are very weak, but definitely noticeable if you make slight contrast boosts). By f/9 or so, the sunstars are very large: you need to be careful that the giant sunstar doesn’t take up too much of the frame! Stopping down to the lens’s minimum of f/16 increases the size and the strength of the sunstar.
Below is a comparison of a sunstar at f/9 (first photo) and at f/16 (second photo).
Interestingly, the sunstar rotates direction very slightly depending upon aperture. As the lens is stopped down, the sunstar rotates slightly clockwise. Flip between the photos to see the difference between f/9 and f/16 in terms of sunstar strength and rotation.
Axial chromatic aberration
See the blue-magenta fringes along the crisp edges of the trees? Magenta chromatic aberration is especially visible along the right side of the two trees.
Chromatic aberration can be removed in most editing programs with a single click, but it is impossible to entirely remove it from a photo without any other effects on the photo. However, these effects are minor (usually a tiny halo along sharp lines near the photo's corners), so I find that chromatic aberration is not as important as other optical issues. For the record, I always leave chromatic aberration correction on unless my photo is monochrome.
Longitudinal chromatic aberration
Out-of-focus edges in the background become green. Out-of-focus edges in the foreground become reddish-purple. This is very subtle and will rarely be obvious in photos. (Even when it is noticeable, only other photographers will notice, never normal people!)
I don't see much coma in this lens, even at f/1.4 in the far corners. That said, I use a crop-sensor camera, and coma is typically more noticeable in the far corners of full-frame cameras. The Eiffel Tower photo shows very, very slight coma the the bottom corners, despite being shot at f/1.4. The largest coma blob at the bottom right takes up maybe a quarter inch on the final, three-foot wide print. Not an issue (but I'll include a crop anyway).
This is a super-extreme crop from the very bottom right of a photo at f/1.4, and I still see almost no coma. If you're looking for it, see how the cluster of lights at the very bottom right look slightly X-shaped? That's coma. Ideally, they would be perfect dots, but they're so small that it's not a big deal.
Be thankful that you aren't using Canon's version of the 24mm f/1.4 if you hate vignetting. Nikon's isn't great, but Canon's has even worse vignetting at f/1.4. This shows me that Nikon did pretty well considering the type of lens this is.
The distortion can be removed in Lightroom at the expense of slight loss of edge sharpness and slight cropping of the corners.
That said, I find that I rarely need to correct this lens's distortion, even when straight lines are in the photo. See the sample below, which has no distortion correction applied.
Focus speed is solid for this lens, although it certainly isn’t lightning-fast. To put it this way, it probably can’t track focus on anything running towards the camera (unlike Nikon's f/2.8 zooms, which are built for that). However, it focuses within a reasonable amount of time considering that it’s an f/1.4 lens. F/2.8 lenses have so much more depth of field that their focusing don’t need to be nearly as precise as an f/1.4 lens. For what it does, it’s good. It’s not ten-out-of-ten good, but it’s solid. Don’t expect it to always be perfect at focus tracking on a moving object, but it will probably work if the subject is not too close to the camera.
A wide-angle, wide-aperture lens will almost always have slight focusing issues. This is something that the photographer needs to learn to work around.
The bottom line is that my personal 24mm f/1.4 is generally accurate at focusing. It's not as consistent as the f/2.8 zooms, especially when I try focusing on distant objects, but this is more the fault of phase-detect autofocus systems than the lens itself.
The Nikon 24mm f/1.4 has an internet reputation for being bad at focusing, and it’s true that there are a few early copies out there with broken focusing modules. However, since the 24mm f/1.4 type of lens historically is not good at autofocusing, it's hard to separate the truth from the half-truth online. This isn’t just a Nikon issue; Canon’s 24mm f/1.4 is also notoriously bad at autofocusing on far subjects (Nikon’s lens has a better, though still not great, reputation for focusing).
Part of the issue is the thin depth of field at f/1.4. If people stop the aperture down to f/2.8, they’ll probably find similar accuracy, though not quite as good, as their zooms. Of course, there’s always the argument that you buy lens that goes to f/1.4 lens to shoot at f/1.4. To the people who argue, I have to say that you've hit the limits of phase-detection. This lens is surprisingly good at autofocus, considering its focal length and aperture.
All that being said, there are still some bad copies of this lens. These have two recurring themes: they literally will not focus past about ten feet, and they won’t respond to autofocus fine-tune. It’s easy enough to see if the lens focuses past ten feet: set it to f/1.4 on single-servo autofocus mode and, using the center point in the viewfinder, focus on a distant building or horizon (make sure these locations are well-lit). If the autofocus goes ANYWHERE past ten feet, whether it actually nails focus or not, you’ve passed the first checkpoint. If the actual focusing point isn’t exactly correct, you can fine-tune it later. Don't go by the distance scale on the lens (it doesn't have a mark for ten feet, and it isn't calibrated properly either), just look at the photos instead.
To find if your lens responds to fine-tuning, the test is more complicated (although intuitive). Visit my link here, which also explains how to fine-tune a lens's autofocus properly without buying a third-party product.
Overall, although this lens isn't as good as Nikon's f/2.8 zooms in terms of autofocus speed or accuracy, it is definitely not as bad as the internet would suggest.
This article is the second of three.
Part one covers the build quality of the lens.
Part three is a summary of the lens, plus a comparison to others.