CANON 16-35MM F/2.8 III VS F/4 ON MEDIUM FORMAT IN-DEPTH REVIEW
Enjoying the blog? Support us with a PayPal donation:
IN MY QUEST FOR WIDE-ANGLE LENSES TO USE ON MEDIUM FORMAT X1D, THIS TIME I WENT FOR ZOOMS: A COMPARED CANON 16-35MM F/2.8 III VS CANON 16-35MM F/4 IS IN-DEPTH REVIEW!
Why a compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review on medium format Hasselblad X1D? I love Medium Format, love the look of the images, the colours and the tonal transitions between them; in particular, I love the essentiality of the Hasselblad X1D, the amazing files it produces and the fact that it offers a native 21mm lens. However, 21mm with its 17mm field-of-view (FOV) equivalent in so-called “full frame” terms, is often not wide enough for my work. No medium format manufacturer currently offers a true ultra-wide angle native prime or zoom lens, such as a 12mm or 14mm FOV equivalent prime, or such as a 12-24mm or a 14-24mm FOV equivalent zoom lens; the only option is the Laowa 17mm, with its 14mm FOV equivalent, which is sadly available in Fuji GFX mount only.
More, all medium format systems on the market today feature gaps between primes in the wide-angle range, sometimes large ones. Every manufacturer offers some of the “classic” FOV equivalent focal lengths but misses others; sometimes, a zoom takes care of at least part of that gap, but that isn’t always the case.
Let’s have a look at what is available at the time of writing. In FOV equivalent terms, Hasselblad offers 17mm, 24mm, 35mm and 28-59mm FOV equivalent, missing 21mm completely and 28mm with a prime. FujiFilm offers 19mm, 24mm, 35mm and 26-52mm FOV equivalent, also missing 21mm completely and 28mm with a prime. Leica offers 19mm, 24mm, 28mm, 35mm and 24-72mm FOV equivalent, missing 21mm completely. The nearly dead and long unsupported Pentax 645 system offers just a zoom covering 22-35mm FOV equivalent. PhaseOne offers 17mm, 22mm, 28mm, 34mm and 25-50mm FOV equivalent, missing 24mm with a prime.
Of course, the above line-ups might already be perfectly fine for you and your kind of work, and of course you can adapt lenses to cover the gaps. However, the Leica S, Pentax 645 and PhaseOne XF systems are fairly limited in the number of ultra-wide lenses you can adapt to them, generally allowing only the use of some other medium (or larger) format lenses and some so-called “full frame” Tilt-Shift lenses.
PhaseOne, while limited in their adaptable lenses if you use their XF cameras, allows for the use of a wider range of adapted lenses using their digital back on a tech camera.
On the other hand, Hasselblad X1D and FujiFilm GFX users, thanks to the mirrorless design of their cameras and their short flange distance, can adapt pretty much any lens designed for any so-called “full frame” system, albeit with some caveats.
Personally, I worked very happily for one and a half year with the lenses offered me by the Hasselblad X1D system, adding the quite amazing Voigtlander 15mm to it to cover the ultra-wide range. However, the Voigtlander 15mm forces you to crop, and therefore lose resolution, and it also features some colour fringing at the extremes of its image circle, which I don’t particularly favour spending time on removing (see my VOIGTLANDER SUPER-WIDE HELIAR 15MM IN-DEPTH REVIEW ON THE HASSELBLAD X1D for more information).
Therefore, I kept looking for a viable ultra-wide angle FOV equivalent solution. Among today’s ultra-wide lenses adaptable on my Hasselblad X1D and on Fuji GFX with ease, the Canon 16-35mm zooms caught my eye. Both lenses in this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review, when used on a 44 x 33mm medium format sensor, will cover a FOV equivalent in so-called “full frame” terms of about 13-28mm. This is definitely wide enough for my purposes and has the added bonus to nicely fill all the gaps in the Hasselblad XCD’s and FujiFilm GFX’s wide-angle line-ups as well. More, in the case of the Canon 16-35mm f/2.8 III, no native ultra-wide-angle lens for medium format is as fast. This is a great plus for night-time landscape photography, both in terms of ease to focus and to keep a fast-enough shutter speed to stop stars, if needed, without having to go too high on ISO.
However, what counts is whether they project a large enough image circle to be useable on medium format, and most importantly whether their performance is good enough for me to consider them for my work. The Canon 16-35mm f/4 IS has a good reputation for a small and sharp lens, while the Canon 16-35mm f/2.8 III is arguably one of the best, if not the best, ultra-wide angle zooms out there today for so-called “full frame”. So, I decided to get them both and give them a try, comparing them at various focal lengths to see if they were viable for me and my work. As always, I got both lenses from the great guys at NEWOLDCAMERA in Milan, and I’d like to thank them all once more for their impeccable service, starting with the owner, the great Ryuichi Watanabe.
Before we start, let me remind you that this, just like all my reviews, is a review made under the point of view of a Fine Art landscape photographer. If your genre of photography is different, my findings might not apply to you and your work.
Disclaimer: At the time of writing, I am not affiliated with Canon, Hasselblad, Techart or NewOldCamera in any way. I am a professional photographer looking for the best equipment for my work, I buy all my gear with my hard-earned cash and I don’t get paid by anyone to write articles for my blog.
Let’s start and see in this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review how these lenses performed!
THE ADAPTER
There are various adapters allowing you to use Canon EOS lenses on the Hasseblad X1D. I choose the Techart TCX-01 because is the smallest, lightest, more “transparent” and simpler in use, is powered by the camera, and last but not least is not as expensive as other choices.
Before getting the Techart TCX-01 I knew that it didn’t officially support either lens in this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review, thus possibly introducing some limitations in its functionality, but I decided to take a gamble and to go for it anyway.
For this review, I used Firmware 4.0.0 on the Techart adapter and Firmware 1.2.0 on my Hasselblad X1D II, the latest for both devices at the time of writing.
As it turned out, the adapter works pretty well with both zooms, allowing you to control your aperture from the camera as you would with a native Hasselblad XCD lens, and recording all shooting information in EXIF as a native Hasselblad XCD lens would.
However, there are three limitations you need to be aware of: 1. Autofocus doesn’t work; 2. Lens make & model show as “Hasselblad HC 80”, but the focal length is actually the correct one (see next point); 3. To record the correct focal length in EXIF you need to set it on the zoom and then restart your camera. If you don’t, the focal length recorder in EXIF will be the one your lens was first set at when you turned the camera on.
I contacted Techart asking them if they planned to add support for either of those zooms but haven’t heard from them at the time of writing. I will update the review if and when that happens.
BUILD, CONTROLS, SIZE AND WEIGHT
Both lenses in this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review are well built, and their plastic bodies feel solid enough. Gently shaking both lenses, you can hear some moving parts with both. Both are weather-sealed, albeit Canon recommends the use of a front filter to “to ensure adequate dust and drip resistance” with the Canon 16-35mm f/2.8 III.
Both lenses feature internal zoom and internal focus. This means that either lens’ barrel won’t extend with zooming, and that either lens’ front element won’t turn when focusing it.
The lens barrel’s controls are similar on both lenses, featuring two rings, the AF/MF button and, in the case of the Canon 16-35mm f/4 IS, the image stabilization On-Off button as well.
On both lenses, closer to the camera body you’ll find the zoom ring. Focal length markings are 16, 20, 24, 28 and 35mm on both lenses and aren’t extremely precise, much less so than e.g. on the Hasselblad XCD 35-75mm.
Closer to each lens’ front element you’ll find the focus ring, with a distance scale on both lenses. Both lenses use USM, meaning that in manual focus you can turn the focus ring between close focus & infinity, past which points the focus mechanism will decouple and the ring will keep turning.
I love having distance scales on my lenses, even though both lenses in this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review feature markings at 0.28cm, 0.5cm, 1m and infinity only, not allowing for a great degree of control.
Focus throw is short but allows enough control in manual focus, and is very similar on both lenses.
The Canon 16-35mm f/2.8 III features 16 elements in 11 groups, has a diameter of 88.5mm, a length of 127.5mm, weighs 790 gr, and has a filter thread of 82mm. The much smaller Canon 16-35mm f/4 IS, on the other hand, features 16 elements in 12 groups, has a diameter of 82.6mm, a length of 112.8mm, weighs 615 gr, and has a filter thread of 77mm.
Adding the Techart TCX-01 adapter, which makes for a more meaningful comparison since you’ll need to use one anyway, the Canon 16-35mm f/2.8 III becomes 170mm tall and weighs 954 gr, while the Canon 16-35mm f/4 IS becomes 155mm tall and weighs 777 gr (my measurements, with caps).
For reference, let’s now compare them in their lens + adapter configuration with the Hasselblad XCD 21mm, the FujiFilm GF 23mm and some medium format zoom lenses. IF = Internal Focusing, IZ = Internal Zooming, OIS = Optical Image Stabilization. Here we go:
– Canon 16-35mm f/2.8 III + Techart TCX-01: 170 x 88.5mm, filter size 82mm, 954 gr, IF, IZ, no OIS;
– Canon 16-35mm f/4 IS + Techart TCX-01: 155 x 82.6mm, filter size 77mm, 777 gr IF, IZ, OIS;
– Hasselblad XCD 21mm f/4: 106 x 83mm, filter size 77mm, 600 gr, IF;
– FujiFilm GF 23mm f/4: 103 x 89.8mm, filter size 82mm, 845 gr, IF;
– Hasselblad XCD 35-75mm f/3.5-4.5: 145 x 85mm, filter size 77mm, 1115 gr, IF, IZ, no OIS;
– FujiFilm GF 32-64mm F/4: 116 x 92,6mm, filter size 77mm, 875 gr, IF, no IZ, no OIS;
– Pentax DA 645 28-45mm F/4.5: 151,5 x 99 mm, filter size 82mm, 1470 gr, IF, no IZ, OIS;
– PhaseOne Schneider 40-80mm LS f/4.0-5.6: 150 x 111 mm, filter size 105mm, 1860 gr, IF, no IZ, no OIS;
– Leica Vario-Elmar-S 30-90mm F/3.5-5.6: 113,5 x 101 mm, filter size 95mm, 1275 gr, IF, no IZ, no OIS;
As you can see, both zoom with the Techart TCX-01 adapter are appreciably longer than the native primes they would replace but are in line with them in terms of diameter and weight, especially comparing them against FujiFilm GF lenses, which are generally heavier than equivalent Hasselblad XCD ones. The comparison with medium format zoom lenses is interesting, despite the difference in coverage. Since medium format users are used to carry and use zoom lenses of a certain size and weight, they likely wouldn’t mind using either of the adapted Canon wide-angle zoom reviewed here.
USE OF FILTERS
The ability of using my 100mm square filter system is fundamental for me when I choose a lens. Going with the larger, heavier, more cumbersome 150mm square filter system is not an option for me, if I can help it. The lenses in this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review both allow me to use my beloved Formatt-Hitech Firecrest Ultra kit with either an 82mm or a 77mm standard ring.
Starting October 2021, I am a proud Ambassador for H&Y Filters, in my opinion simply the best filters out there with the most innovative, easy to use, practical filter holder (see my BEST FILTERS FOR LANDSCAPE PHOTOGRAPHY and BEST 100MM FILTER HOLDER FOR LANDSCAPE PHOTOGRAPHY articles for more info).
Thanks to my relationship with H&Y Filter, I am happy to offer you a 10% discount on H&Y FILTRI, their Italian distributor’s website. Just use code Vieri10 at checkout to enjoy your discount!
IN USE: FOCUSING AND DIAPHRAGM
At the time of writing, the Techart TCX-01 adapter won’t allow the use of autofocus using either lens in this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review on the Hasselblad X1D II. However, the Hasselblad X1D’s manual focus magnification feature works perfectly with both lenses, allowing you to focus with great precision and ease. While this is more than enough for my Fine Art landscape work, if you plan on using these lenses for other kinds of photography this might be something you want to keep in mind.
Like with all native Hasselblad XCD lenses, aperture can be controlled via a wheel on the camera body, or via the touchscreen, in all shooting modes supporting aperture control with both the Canon 16-35mm f/2.8 III and the Canon 16-35mm f/4 IS used with the Techart TCX-01 adapter as well.
Please note that using a different adapter on the X1D, or using either lens on a FujiFilm GFX camera, might return different results.
SHARPNESS AT INFINITY
For this review, I decided to take my test shots at 16mm, 18/19mm which is the shortest focal length with no mechanical vignetting on both lenses, 21mm to compare with the Hasselblad XCD 21mm, 28mm which is the longest focal length where the Canon 16-35mm f/4 IS is useable before vignette kicks back in, and 35mm on the Canon 16-35mm f/2.8 III alone.
Methodology: using my usual “real world” test scene, I manually focussed with the lens wide open on the trees on the far ridge in the middle of the frame using the maximum focus area magnification for precise focus. I then prepared 900 x 600px, 100% crops of the center, bottom left corner and mid-right side of the frame at full-stop apertures ranging from wide open to f/16. My Hasselblad X1D II had Firmware 1.2.0 installed. Adobe Camera Raw’s profiles for the Canon 16-35mm f/2.8 III and Canon 16-35mm f/4 IS have been applied. Upon various forum users’ request, I decided to add a set of corner images taken with each lens after removing a light baffle present close to the lens mount. Removing the baffle is extremely easy, you can pop it off with a finger; without the baffle, both lenses vignette much less, especially at the long end, but this often comes at the expense of corner sharpness, unless you stop down.
16mm
Let’s start with field of view and useable coverage. On a so-called “full frame” camera, a 16mm lens will cover an angle of view of about 108 degrees. On the medium format Hasselblad X1D’s sensor, thanks to the 0,78x crop factor, this should be equivalent to a 12,4mm lens. However, at 16mm neither lens in this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review covers the whole Hasselblad X1D’s sensor as a native XCD lens does. To give you an idea of what to expect in terms of coverage, I prepared a full image at 16mm with overlays of what I think are pretty much the maximum usable areas for 1:1, 5:4, 4:3, 3:2, 16:9, 2:1 crops, as well as the resulting image sizes & resolution after cropping. Canon 16-35mm f/2.8 III first, followed by the Canon 16-35mm f/4 IS (click on the images to enlarge):
As you can see, if you go for the square image ratio that so many Hasselblad photographers are used to and love, you’ll be able to use either lens at 16mm with no need for cropping. If you go for a 2:1 or 16:9 panoramic ratio, you either won’t lose any resolution or you will lose just an irrelevantly tiny bit. Choosing an image ratio in between square and panoramic, on the other hand, you will lose a bit more resolution compared to a native XCD lens. See below, full size X1D images first, Canon 16-35mm f/2.8 III crops second and Canon 16-35mm f/4 IS crops third:
– 1:1. No difference, you’ll always get 38,4 Mp;
– 5:4. 48 Mp vs 38,7 Mp vs 41.5 Mp;
– 4:3. 51,2 Mp vs 38 Mp vs 41 Mp;
– 3:2. 45,6 Mp vs 36,6 Mp vs 39,6 Mp;
– 16:9. 38,4 Mp vs 34,2 Mp vs 37,4 Mp;
– 2:1. 34,2 Mp vs 32,8 Mp vs 34,2 Mp.
As you can see, the differences become smaller as you get farther away from 4:3, where of course the native lenses have the larger advantage. Choosing 16:9 or 2:1, you pretty much don’t lose anything, and even going for a 4:3 crop you’ll end up with a very respectable 38 Mp or 41 Mp, according to which of the two Canon zoom you’ll choose.
Now, let’s compare 4:3 images taken at 16mm after cropping out the mechanical vignette against images taken at 18/19mm and left un-cropped. Canon 16-35mm f/2.8 III first followed by the Canon 16-35mm f/4 IS (click on the images to enlarge):
As you can see, if you plan on using a 4:3 image ratio, you’d be better off zooming in until the mechanical vignette disappears rather than shooting at 16mm and cropping. Images taken at 18/19mm and un-cropped actually cover a larger field of view than the images taken at 16mm and cropped, and you’ll end up with more resolution as well. The only time I’d use either lens at their widest zoom setting is if you plan to crop 1:1, 16:9 or 2:1. Other than that, zooming in to around 19mm will be your best solution.
For my sharpness analysis at 16mm, I chose to go with 4:3 crops, the worst-case scenario for both lenses. Since by cropping to either 1:1, 16:9 or 2:1 you’ll end up using a more central area of the lens, if you plan on doing that you can expect better performance than the results you’ll see below. In my detailed assessments of the crops below, I will compare both lenses with the Voigtlander Super-Wide Heliar 15mm III, so you might want to open my VOIGTLANDER SUPER-WIDE HELIAR 15MM IN-DEPTH REVIEW ON THE HASSELBLAD X1D in a new tab to easily compare crops.
Let’s start with the center of the frame, Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
At 16mm, in the center of the frame, the Canon 16-35mm f/2.8 III is flawless, starting razor sharp wide-open at f/2.8 and staying like that up to f/11. Apertures smaller than f/16 start suffering from diffraction. The Canon 16-35mm f/4 IS also starts razor sharp in the center wide open at f/4, stays that way at f/5.6 but sharpness starts declining after that. Comparing it to the Voigtlander 15mm, the Canon 16-35mm f/2.8 III wins in that it provides the same level of sharpness at comparable apertures, but with a much faster aperture; on the other hand, the Voigtlander is better than the Canon 16-35mm f/4 IS along the aperture range.
Let’s now examine the bottom left corner, Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
First, as you can see both lenses do vignette quite a bit wide open; however, the Canon 16-35mm f/2.8 III clears vignette earlier than the Canon 16-35mm f/4 IS. The Voigtlander 15mm showed less vignette than either Canon lenses. As far as sharpness goes, the Canon 16-35mm f/2.8 III starts very sharp wide-open at f/2.8, becoming razor-sharp between f/5.6 and f/11. Apertures smaller than f/16 start suffering from diffraction. On the other hand, the Canon 16-35mm f/4 IS starts much softer than its faster sibling, and never becomes as sharp. The Voigtlander 15mm is as sharp as the Canon 16-35mm f/2.8 III at comparable apertures, and is sharper than the Canon 16-35mm f/4 IS. However, in the Voigtlander crops you can easily see the colour fringing I mentioned at the beginning of this review, which is all but absent in both Canon lenses.
Finally, let’s check the mid-right side of the frame, Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
In the mid-right of the frame, the Canon 16-35mm f/2.8 III starts very sharp wide-open at f/2.8 and becomes razor-sharp between f/5.6 and f/11. After that, diffraction takes its toll. The Canon 16-35mm f/4 IS, on the other hand, starts very sharp at f/4, and becomes razor-sharp between f/5.6 and f/8. At smaller apertures, diffraction takes its toll. The Voigtlander 15mm is sharper wide-open than either Canon lens wide-open, but at comparable apertures all three lenses are equally sharp here.
18/19mm
Let’s now compare the widest setting at which both lenses are clear of hard mechanical vignette, 19mm for the Canon 16-35mm f/2.8 III and 18mm for the Canon 16-35mm f/4 IS. Let’s see the full images first, to give you an idea of the relative coverage of both lenses at their widest useable zoom setting without cropping. Canon 16-35mm f/2.8 III first, with baffle and without baffle, followed by the Canon 16-35mm f/4 IS equally with baffle and without baffle (click on the images to enlarge):
You can still see some vignette in the Canon 16-35mm f/4 IS’s image but is soft enough that you can either leave it in or easily correct it, making the Canon 16-35mm f/4 IS useable at a slightly wider setting than the Canon 16-35mm f/2.8 III.
Let’s now examine the crops in detail, starting with the center of the frame. Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
At 18/19mm, in the center of the frame, the Canon 16-35mm f/2.8 III is again flawless, being already razor sharp wide-open at f/2.8 and staying like that up to f/11. Apertures smaller than f/16 start suffering from diffraction. The Canon 16-35mm f/4 IS also starts razor sharp in the center wide open, but sharpness starts declining gently at f/8.
Let’s now examine the bottom left corner, Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Now, the Canon 16-35mm f/2.8 III without the baffle (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
Now, the Canon 16-35mm f/4 IS without the baffle (click on the images to enlarge):
Again, as at 16mm, you can see here how both lenses do vignette quite a bit wide open. In this case, however, the Canon 16-35mm f/2.8 III clears vignette later than the Canon 16-35mm f/4 IS. As far as sharpness goes, the Canon 16-35mm f/2.8 III starts pretty sharp wide-open at f/2.8, if you manage to see into the deep vignette. It then becomes very sharp at f/5.6 and razor-sharp between f/8 and f/11, where I can definitely recommend it for critical work both in terms of sharpness and vignette. Apertures smaller than f/16 start suffering from diffraction. On the other hand, the Canon 16-35mm f/4 IS starts very soft wide open at f/4, and while it is useable in the corner at f/11, it never becomes critically sharp.
Comparing the no baffle series, you can definitely see how removing the baffle helps getting rid of vignette. As far as sharpness goes, it seems to me that both lenses suffer wide open from the absence of the baffle, losing about one stop of sharpness without it: the Canon 16-35mm f/2.8 III becomes razor sharp at f/11, while the Canon 16-35mm f/4 IS never becomes really sharp without the baffle as well.
Finally, let’s check the mid-right side of the frame, Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
In the mid-right of the frame, both lenses start extremely sharp wide-open. The Canon 16-35mm f/2.8 III is razor sharp between f/4 and f/11, and as expected apertures of f/16 and smaller start suffering from diffraction. The Canon 16-35mm f/4 IS becomes razor sharp between f/5.6 and f/8, with apertures of f/11 and smaller getting softer again due to diffraction.
21mm
Let’s now examine 21mm, an interesting focal length for us Hasselblad X1D users since here we have the Hasselblad XCD 21mm, a truly impressive lens. In my detailed assessments of the crops below I will compare both lenses with it, so you might want to open my HASSELBLAD XCD 21MM F/4 IN-DEPTH REVIEW in a new tab to easily compare crops. Let’s start with the full image, to give you an idea of the relative coverage of both lenses at their EXIF focal length of 21mm. Canon 16-35mm f/2.8 III first, with baffle and without baffle, followed by the Canon 16-35mm f/4 IS equally with baffle and without baffle (click on the images to enlarge):
Looking at the full images above, you’ll notice something interesting. Of course, I didn’t change camera position between both series, switching lenses while leaving the camera body on the tripod, unmoved. While both lenses cover pretty much the same field of view at their nominal 21mm setting, it looks like the image field “shifted” somehow when changing lenses. This was already visible at 16mm and 18/19mm, but I preferred to leave this comment for 21mm, a setting where both lenses are used at exactly the same focal length and uncropped.
Let’s now examine the crops in detail, starting with the center of the frame. Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
At 21mm, both lenses performed fantastically well in the center, being razor sharp pretty much all over the aperture range up to f/11. The Hasselblad XCD 21mm offered an equally impressive performance over the aperture range. Apertures smaller than f/16 start suffering from diffraction on all three lenses.
Let’s now examine the bottom left corner, Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Now, the Canon 16-35mm f/2.8 III without the baffle (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
Now, the Canon 16-35mm f/4 IS without the baffle (click on the images to enlarge):
Interestingly, 21mm is not a strong focal length on either Canon zoom in the far corner. The Canon 16-35mm f/2.8 III starts slightly soft wide-open at f/2.8, becomes sharp at f/5.6, critically sharp at f/8 and razor sharp at f/11. Interestingly, f/16 is as sharp as (or possibly even sharper than) f/11, telling me that the lens features some pretty serious field curvature at 21mm. The Canon 16-35mm f/4 IS shows a very similar performance, including field curvature and all, but while useable at f/16, it never gets critically sharp in the corner at 21mm. In comparison, the Hasselblad XCD 21mm starts sharper wide-open than the Canon 16-35mm f/2.8 III, but at comparable apertures from f/8 on, the Canon is actually sharper than the Hasselblad. Compared to the Canon 16-35mm f/4, on the other hand, the Hasselblad XCD 21mm wins easily over the whole aperture series.
Comparing the no baffle series, also at 21mm you can clearly see how removing the baffle helps getting rid of vignette. As far as sharpness goes, at 21mm it seems to me that both lenses suffer much less from the absence of the baffle than at 18/19mm, keeping the relative performance described above even without the baffle.
Finally, let’s check the mid-right side of the frame, Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
Performance at mid-right side is in line with what both lenses showed in the center, with both lenses being razor sharp pretty much all over the aperture range up to f/11, with the Canon 16-35mm f/2.8 III being slightly better than its slower sibling. This reinforces my consideration about field curvature being responsible for the weaker performance in the corner. On the other hand, the Hasselblad XCD 21mm showed a much weaker performance over the aperture range at mid-right side compared to its amazing rendition in the center, and here is slightly weaker than the Canon 16-35mm f/2.8 III and on par with the Canon 16-35mm f/4 IS.
28mm
Let’s now examine both lenses in this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review at 28mm. Interestingly, 28mm is the last useable focal length on the long side with the Canon 16-35mm f/4 IS. At 28mm and above, in fact, this lens starts showing some mechanical vignette again, as you can see in the full image below. While slightly longer at 30mm, in my detailed assessments of the crops below I will compare both lenses with the Hasselblad XCD 30mm, so you might want to open my HASSELBLAD XCD 30MM F/3.5 IN-DEPTH REVIEW in a new tab to easily compare crops. Let’s start with the full image, to give you an idea of the relative coverage of both lenses at their EXIF focal length of 28mm. Canon 16-35mm f/2.8 III first, with baffle and without baffle, followed by the Canon 16-35mm f/4 IS equally with baffle and without baffle (click on the images to enlarge):
As mentioned above, you can see how mechanical vignette is starting to creep back in the Canon 16-35mm f/4 IS’s images. You’ll also notice how removing the baffle definitely makes a difference with the Canon 16-35mm f/4 IS. As with the 21mm images, of course I didn’t change camera position between both series, just switching lenses while leaving the camera body on the tripod. Again, while both lenses cover pretty much the same field of view at their nominal 28mm setting, at this focal length as well it looks like the image field moved somehow switching lenses, shifting on the vertical axes.
Let’s now examine the crops in detail, starting with the center of the frame. Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
At 28mm, again both lenses performed fantastically well in the center, being extremely sharp wide-open, becoming razor-sharp closing down one stop (at f/4 and f/5.6 respectively) and staying that way pretty much all over the aperture range. Apertures of f/16 and smaller started suffering from diffraction on both lenses. Comparing them with the Hasselblad XCD 30mm, both zooms hold their own, being as good if not better than the Hasselblad prime here.
Let’s now examine the bottom left corner, Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Now, the Canon 16-35mm f/2.8 III without the baffle (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
Now, the Canon 16-35mm f/4 IS without the baffle (click on the images to enlarge):
At 28mm in the lower left corner, the Canon 16-35mm f/2.8 III starts very sharp wide-open at f/2.8, becoming razor-sharp at f/4 and staying that way until diffraction kicks in at f/16 and smaller apertures. The lens features some strong vignette wide open; vignette clears stopping down, becoming pretty much irrelevant by f/8-f/11 and is gone by f/16. On the other hand, the Canon 16-35mm f/4 IS starts off very soft, and while it improves stopping down it never becomes critically sharp. Interestingly, with the Canon 16-35mm f/4 IS vignette is nearly absent wide-open and increases stopping the lens down. The Hasselblad XCD 30mm is in between the zooms here, being definitely better than the Canon 16-35mm f/4 IS all over the aperture range, but not better than the Canon 16-35mm f/2.8 III until f/8. The Hasselblad XCD 30mm, as expected, performs better than both zooms in terms of vignetting.
Comparing the no baffle series, at 28mm you can clearly see how removing the baffle helps getting rid of vignette with both lenses. As far as sharpness goes, with the baffle removed the Canon 16-35mm f/2.8 III suffers a bit, becoming razor sharp at f/8. The Canon 16-35mm f/4 IS, on the other hand, replicate the weak performance showed with the baffle, never becoming really sharp over the aperture range.
Finally, let’s check the mid-right side of the frame, Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
At 28mm in the mid-right of the frame, again the Canon 16-35mm f/2.8 III performed wonderfully, being extremely sharp wide-open and razor-sharp stopping down to f/4. The lens stays that way all over the aperture range until diffraction kicks in at apertures of f/16 and smaller. The Canon 16-35mm f/4 IS, on the other hand, starts off slightly softer than its faster sibling; sharpness improves stopping down to become razor-sharp at f/8. Apertures of f/16 and smaller suffer from diffraction. Again, both zooms outperform the Hasselblad XCD 30mm here until f/8 – f/11 where the Hasselblad prime gets really close to both zooms, becoming perhaps slightly better than the Canon 16-35mm f/4 IS but never as good as the Canon 16-35mm f/2.8 III.
35mm
At 35mm, the vignette that we started to see on the Canon 16-35mm f/4 IS at 28mm became too hard to be able to use full size images at this focal length. Since it doesn’t make any sense to shoot at 35mm and then crop, when you can just zoom out and use all the frame at full resolution, I decided not to examine crops taken with the Canon 16-35mm f/4 IS at 35mm with the baffle on. Removing the baffle made the Canon 16-35mm f/4 IS perfectly useable at 35mm, so I added only no baffle corner crops here.
As you can see in the images below, some mechanical vignette started to show at 35mm with the Canon 16-35mm f/2.8 III as well. However, the vignette is gentle enough to be easily removed in post-production should you decide to so do, making the lens perfectly useable at this focal length. In my detailed assessments of the crops below, I will compare the Canon 16-35mm f/2.8 III with the Hasselblad XCD 35-75mm at 35mm, so you might want to open my HASSELBLAD XCD 35-75MM F/3.5-4.5 IN-DEPTH REVIEW in a new tab to easily compare crops. Let’s start with the full image, to give you an idea of the vignette I am talking about. Canon 16-35mm f/2.8 III first, with baffle and without baffle, followed by the Canon 16-35mm f/4 IS equally with baffle and without baffle (click on the images to enlarge):
At 35mm, you can clearly see how removing the baffle dramatically reduces vignette on both lenses, and especially so on the Canon 16-35mm f/4 IS. With the baffle removed, both lenses became perfectly useable at 35mm, in terms of vignetting.
Let’s now examine the crops in detail, starting with the center of the frame. Canon 16-35mm f/2.8 III only for this focal length (click on the images to enlarge):
As you can see, interestingly at this focal length the lens starts off pretty sharp, but sharpness declines at f/4 and f/5.6 to be back at f/8 and f/11, before declining again at f/16 due to diffraction. This tells me that the Canon 16-35mm f/2.8 III suffers of focus shift at 35mm, while it didn’t show any such behaviour at shorter focal lengths. Comparing it with the Hasselblad XCD 35-75mm, the XCD zoom wins, being razor-sharp all over the aperture range without any focus shift. However, at apertures where the Canon 16-35mm f/2.8 III is not affected by focus shift, such as f/2.8 and between f/8 and f/16, both lenses are equally sharp.
Let’s now examine the Canon 16-35mm f/2.8 III in the bottom left corner (click on the images to enlarge):
Now, the Canon 16-35mm f/2.8 III without the baffle (click on the images to enlarge):
And the Canon 16-35mm f/4 IS without the baffle (click on the images to enlarge):
In the corner, the lens doesn’t show any focus shift, starting off extremely sharp, becoming razor sharp at f/4 and staying that way until diffraction kicks in at apertures of f/16 and smaller. While there is some noticeable vignette wide-open, vignette clears off nicely by f/8. You’ll notice that vignette is back at f/16, as we saw with the Canon 16-35mm f/4 IS at 28mm above. Comparing the Canon 16-35mm f/2.8 III with the Hasselblad XCD 35-75mm at 35mm, you’ll notice that the Canon lens might be slightly sharper wide open, but by f/5.6 both are razor sharp. The Hasselblad XCD 35-75mm at 35mm shows almost no vignette, if that is important to you.
Comparing the no baffle series, at 35mm you can clearly see how removing the baffle helps getting rid of vignette with both lenses. As far as sharpness goes, however, with the baffle removed the Canon 16-35mm f/2.8 III definitely suffers, starting off soft to become razor sharp only at f/8 – f/11. The Canon 16-35mm f/4 IS, on the other hand, is unusable with the baffle on, but while removing the baffle makes it viable at 35mm, performance in the corner doesn’t get acceptable until you stop it down to f/11, thus limiting the possible application for this lens.
Let’s now examine the Canon 16-35mm f/2.8 III in the mid-right side of the frame (click on the images to enlarge):
In the mid-right side of the frame, the lens starts off extremely sharp, becoming razor sharp at f/4 and staying that way until diffraction kicks in at apertures of f/16 and smaller. The Hasselblad XCD 35-75mm at 35mm in the mid-right side of the frame offered an equally amazing performance.
SHARPNESS AT CLOSE FOCUSING DISTANCE AND BOKEH
To examine sharpness at close focusing distance, as well as to see how the lenses in this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review draw out-of-focus areas, I focused on the Phillips screw right under the mailbox’s red flag in the lower right corner of the frame (about 40-50cm from the lens barrel, depending on focal lengths). Together with the full frame images, I included 900 x 600px, 100% crops taken at the point of focus, to check out sharpness, and center crops to see what happens in out-of-focus areas near infinity.
In this series, I examined both lenses at 19mm, 21mm, 28mm and at 35mm again using the Canon 16-35mm f/2.8 alone. Since zooming out to 16mm and crop will result in less field of view than using either zoom at 19mm (see the infinity section above) I decided to skip the 16mm setting here.
19mm
Let’s start with the full images first, to give you an idea of the relative coverage of both lenses at their widest useable zoom setting without cropping. Canon 16-35mm f/2.8 III first, with baffle and without baffle, followed by the Canon 16-35mm f/4 IS equally with baffle and without baffle (click on the images to enlarge):
Let’s now examine sharpness at the point of focus, Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Now, the Canon 16-35mm f/2.8 III without the baffle (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
And the Canon 16-35mm f/4 IS without the baffle (click on the images to enlarge):
At 19mm and focussed up close, the Canon 16-35mm f/2.8 III starts off quite sharp; sharpness further increases stopping the lens down, reaching its peak between f/8 and f/16. The Canon 16-35mm f/4 IS follows a similar pattern and is at its sharpest at f/11. I would call both lenses critically sharp between f/8 and f/11, but neither lens becomes razor-sharp at any aperture here.
Comparing the no baffle series, at 19mm and focussed up close you don’t see much difference in terms of vignetting against the baffle images. In terms of sharpness, however, contrarily to what happened focussing at infinity here you can see a small, but visible, improvement shooting without the baffle.
Let’s see now how these lenses render out-of-focus areas at 19mm and at far distances in the center of the frame. Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
Both lenses draw pretty beautifully at 19mm, with the Canon 16-35mm f/2.8 III staying gentler and drawing smoother over the aperture range than the Canon 16-35mm f/4 IS.
21mm
As at infinity, in my detailed assessments of the crops below I will compare both lenses with the Hasselblad XCD 21mm, so you might want to open my HASSELBLAD XCD 21MM F/4 IN-DEPTH REVIEW in a new tab to easily compare crops. As always, let’s start with the full images first, to give you an idea of the relative coverage of both lenses at 21mm. Canon 16-35mm f/2.8 III first, with baffle and without baffle, followed by the Canon 16-35mm f/4 IS equally with baffle and without baffle (click on the images to enlarge):
Let’s now examine sharpness at the point of focus, Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Now, the Canon 16-35mm f/2.8 III without the baffle (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
And the Canon 16-35mm f/4 IS without the baffle (click on the images to enlarge):
At 21mm and focussed up close, the Canon 16-35mm f/2.8 III starts off very sharp; sharpness then increases stopping the lens down, and between f/8 and f/11 the lens is razor-sharp. The Canon 16-35mm f/4 IS behaves similarly and is razor-sharp at f/11. As expected, apertures smaller than f/16 start suffering from diffraction. Comparing these performances with the Hasselblad XCD 21mm, which outputted a tremendous performance in the close corner, both Canon zoom fall slightly shorter, with the Hasselblad besting them both from wide-open to f/11, where all three lenses performed equally well.
Comparing the no baffle series, at 21mm and focussed up close – as happened at 19mm – you won’t see much difference in terms of vignetting against the baffle images. In terms of sharpness, contrarily to what happened at 19mm, it seems to me that both lenses lost a bit of sharpness when used without the baffle.
Let’s see now how these lenses render out-of-focus areas at 21mm and at far distances in the center of the frame. Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
Again, I find both lenses to be drawing pretty beautifully at 21mm, showing pretty nice bokeh for wide-angle zooms. Comparing them to the Hasselblad XCD 21mm, I prefer the rendering of both zooms here, because they follow a more progressively sharpening curve along the aperture series. The Hasselblad XCD 21mm seems to feature a bigger sharpening step between f/4 and f/5.6, almost no difference between f/5.6 and f/8, then more steps, for a less gentle sharpening curve along the aperture series.
28mm
Again, in my detailed assessments of the crops below I will compare both lenses with the Hasselblad XCD 30mm, so you might want to open my HASSELBLAD XCD 30MM F/3.5 IN-DEPTH REVIEW in a new tab to easily compare crops. Once more, let’s start with the full images first, to give you an idea of the relative coverage of both lenses at 28mm. Canon 16-35mm f/2.8 III first, with baffle and without baffle, followed by the Canon 16-35mm f/4 IS equally with baffle and without baffle (click on the images to enlarge):
Let’s now examine sharpness at the point of focus, Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Now, the Canon 16-35mm f/2.8 III without the baffle (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
And the Canon 16-35mm f/4 IS without the baffle (click on the images to enlarge):
At 28mm and focussed up close, the Canon 16-35mm f/2.8 III starts off quite sharp; sharpness then increases stopping the lens down, and between f/8 and f/11 the lens is razor-sharp. The Canon 16-35mm f/4 IS shows an even better performance, being razor-sharp at f/5.6. As expected, apertures smaller than f/16 start suffering from diffraction on both lenses. Comparing these performances with the Hasselblad XCD 30mm I think that both zooms outperformed the Hasselblad prime here.
Comparing the no baffle series, at 28mm and focussed up close both lenses show a pretty sizeable difference in terms of vignetting against the baffle images: removing the baffle removes vignetting as well. In terms of sharpness, however, removing the baffle takes away some sharpness from the Canon 16-35mm f/2.8 III, while the Canon 16-35mm f/4 IS seems to be less affected by the removal and performs pretty much as it did with the baffle on.
Let’s see now how these lenses render out-of-focus areas at 28mm and at far distances in the center of the frame. Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Let’s see now how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
Once more, I find both lenses to be drawing beautifully at 28mm, showing some very nice bokeh for wide-angle zooms. The Hasselblad XCD 30mm also draws beautifully here, so I’d say that all three lenses performed equally well.
35mm
Once more, let’s start with the full images first, to give you an idea of the relative coverage of both lenses at 35mm. As you can see in the images below, as at infinity, the Canon 16-35mm f/4 IS’s vignette makes it unusable at this zoom setting at close focussing distances as well. Removing the baffle made the Canon 16-35mm f/4 IS perfectly useable at 35mm, so I added only no baffle corner crops here too. While some mechanical vignette starts to show with the Canon 16-35mm f/2.8 III at 35mm as well, the vignette is gentle enough to make the lens useable at this focal length. Removing the baffle also clears the Canon 16-35mm f/2.8 III from whatever vignette there was.
Canon 16-35mm f/2.8 III first, with baffle and without baffle, followed by the Canon 16-35mm f/4 IS equally with baffle and without baffle (click on the images to enlarge):
As at infinity, in my detailed assessments of the crops below, I will compare the Canon 16-35mm f/2.8 III with the Hasselblad XCD 35-75mm, so you might want to open my HASSELBLAD XCD 35-75MM F/3.5-4.5 IN-DEPTH REVIEW in a new tab to easily compare crops.
Let’s now examine sharpness at the point of focus, Canon 16-35mm f/2.8 III only (click on the images to enlarge):
Now, the Canon 16-35mm f/2.8 III without the baffle (click on the images to enlarge):
And the Canon 16-35mm f/4 IS without the baffle (click on the images to enlarge):
At 35mm and focussed up close, the Canon 16-35mm f/2.8 III starts off sharp enough. Sharpness increases stopping the lens down and by f/5,6 the lens is critically sharp to become razor-sharp between f/8 and f/11. Comparing it with the Hasselblad XCD 35-75mm, which showed an amazing performance focussed up close at 35mm, the Canon zoom falls short here up until f/8-f/11, where the Hasselblad’s sharpness declines while the Canon’s increases.
Comparing the no baffle series, at 35mm you can clearly see how removing the baffle helps getting rid of vignette with both lenses, dramatically so with the Canon 16-35mm f/4 IS which became perfectly useable. As far as sharpness goes, however, with the baffle removed the Canon 16-35mm f/2.8 III definitely suffers, starting off Sofer than with the baffle on, to become razor sharp only at f/8 – f/11. The Canon 16-35mm f/4 IS, on the other hand, not only became usable without the baffle, but showed a pretty strong performance as well becoming razor sharp around f/8.
Let’s see now how the Canon 16-35mm f/2.8 III renders out-of-focus areas at 35mm and at far distances in the center of the frame (click on the images to enlarge):
Once more, I find the Canon 16-35mm f/2.8 III to be drawing beautifully at 35mm as well, showing a similar sharpening curve as the Hasselblad XCD 35-75mm at 35mm. While I wouldn’t choose a wide-angle zoom to do bokeh, I’d be happy with the performance of either lens here.
SHARPNESS CONCLUSIONS
As a result of this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review’s long sharpness section, we can clearly see a pretty dramatic difference in performance between these two zoom lenses on the Hasselblad X1D medium format sensor. While in the center of the frame both lenses generally performed impressively at any focal length and aperture, for landscape photography work where you need sharpness all over the frame the Canon 16-35mm f/2.8 III consistently outperformed the Canon 16-35mm f/4 IS at pretty much any focal length and at any aperture, making it a clear choice for use on 44 x 33mm medium format sensor cameras, be it Hasselblad X1D or FujiFilm GFX.
While I wouldn’t choose the Canon 16-35mm f/4 IS for my work, I was really impressed by the performance of the Canon 16-35mm f/2.8 III. Considering that I put this lens through a true torture test here, using it well outside the so-called “full frame” image circle for which it has been designed, the Canon 16-35mm f/2.8 III showed a truly outstanding sharpness performance. At infinity, I would feel comfortable using this lens all over its useable range, at any focal length and at pretty much any aperture, keeping past f/5.6 – f/8 to clear vignette and for maximum sharpness. At close range, the lens performed also very well, showing just some weakness at wider apertures on the far corners of the image at some focal lengths.
Comparing it to the extremely good Hasselblad XCD primes & zoom of same or comparable focal length, and with the impressive Voigtlander 15mm, the Canon 16-35mm f/2.8 III certainly holds its ground, sometimes being even better than these lenses. The only parameter where the Canon 16-35mm f/2.8 III consistently performed worse than native Hasselblad XCD lenses, as expected, is vignetting. However, this is not a huge problem for landscape work, since if you use the Canon 16-35mm f/2.8 III at the standard apertures used in landscape photography, such as f/8 – f/16, the lens is just flawless. A truly impressive performance.
WITH OR WITHOUT BAFFLE?
To answer the questions I received on various forum regarding using these lenses without the baffle, I believe that while removing the baffle helps at the long end of both lenses in terms of removing vignette, this comes with a serious penalty in terms of sharpness. It is of course up to you to decide which lenses you choose; personally, after testing both lenses with the baffle removed I would still choose the Canon 16-35mm f/2.8 III with baffle over all other possible combinations. True, it vignettes slightly at 35mm, but that is a focal length I can cover with the 35-75mm zoom, and if I hadn’t got the zoom I would still prefer to clear vignette rather than working with an unsharp image.
Even with the baffle removed, I still wouldn’t choose the Canon 16-35mm f/4 IS for my work.
VIGNETTE AND COLOUR RENDITION
As mentioned, and as you could easily see in the corner crops above, both lenses in this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review showed vignetting on the 44 x 33mm medium format sensor at any focal length. At some focal lengths, this vignette was definitely heavy, and particularly so at larger apertures. However, stopping down to f/8 and beyond generally cleared vignette enough to make it not noticeable in real-world use. Personally, I often like to add back some soft vignetting to my images at a later stage in post-processing, so this doesn’t disturb me much. Of course, your preferences might be different from mine, and with them your assessment of these lenses for your work.
Colour rendition is beautiful to my eye, at least as far as I could see during my test shots. Compared against Hasselblad XCD lenses, both Canon zooms tested here showed a cooler rendition, which you might or might not prefer.
DISTORTION
To check for distortion, as always, I photographed my uninspiring but very useful garage door which, with its straight metal lines, provides me with a good test scene. For your convenience, to help you check for distortion easier I slightly increased contrast, added straight red lines in PP, turned the image horizontal and cut the frame in two, leaving just the top half of the frame. Adobe Camera Raw’s profiles for the Canon 16-35mm f/2.8 III and Canon 16-35mm f/4 IS have been applied.
Let’s see now how these lenses deal with distortion. Canon 16-35mm f/2.8 III first (click on the images to enlarge):
Let’s now see how the Canon 16-35mm f/4 IS behaves (click on the images to enlarge):
As you can see, the lenses in this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review with their lens profile applied behave nicely when it comes to distortion but aren’t completely distortion-free all over the focal length range. This is truer in the corners, areas that fall out of the so-called “full frame” coverage these lenses have been designed for. Conversely, Hasselblad XCD lenses covering the same or similar focal lengths are completely distortion-free once you applied their respective profiles, and so is the Voigtlander 15mm. Luckily, whatever distortion is left with the Canon lenses tested here after applying the lens’ profiles is fairly mild, and easy to correct if needed. Again, the Canon 16-35mm f/2.8 III clearly outperforms the Canon 16-35mm f/4 IS as far as distortion.
FLARE AND CHROMATIC ABERRATIONS
During this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review’s flare test, on purpose I didn’t use the provided, removeable lens hoods. Since I almost always have a filter holder on the lens when I work, I never use a hood in the field, and I want my lens tests to mimic my working conditions as closely as possible. In the torture-test images below, I included the sun in the frame, keeping it close to the image’s corner. As well, I included a crop of the area surrounding the sun to check for loss of contrast and for chromatic aberrations.
Let’s start with full images first, taken at 19mm, 21mm, 28mm and 35mm and starting with the Canon 16-35mm f/2.8 III (click on the images to enlarge):
As you can see, the Canon 16-35mm f/2.8 III showed very high contrast and resistance to flare at all tested focal lengths, with no “ghost” images of sorts except for a small ghost image on the top left at 19mm.
Let’s examine now the Canon 16-35mm f/4 IS (click on the images to enlarge):
In this test as well the Canon 16-35mm f/4 IS proved to be slightly weaker than its faster sibling, showing evident and multiple ghost images at both 19mm and 21mm. Longer focal lengths showed better resistance to ghost images.
Let’s move to the sun crops, starting with the Canon 16-35mm f/2.8 III (click on the images to enlarge):
Let’s examine now the Canon 16-35mm f/4 IS (click on the images to enlarge):
As you can see, both lenses showed good contrast and no fringing or chromatic aberrations at any focal length in high-contrast areas close to the sun.
CONCLUSIONS
Covering an useable “ultra-wide” zoom lens range of 19-35mm, with a FOV equivalent in so-called “full frame” to 14,8-28mm (not considering 1:1, 16:9 and 2:1 crops allowing you to use both lenses at 16mm, or 12.5 FOV equivalent, without any penalty in resolution), the lenses in this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review offer a solution to the lack of native medium format lenses in this range.
While the Canon 16-35mm f/4 IS might be a very strong performer on a so-called “full frame” camera, on the 44 x 33mm medium format sensor the lens definitely falls short if you need a strong performance all over the frame, and I wouldn’t consider it for my work. On the other hand, the Canon 16-35mm f/2.8 III showed a truly impressive performance here. When it comes to sharpness, it performed either as well as Hasselblad own XCD superb lenses or better. When it comes to vignetting and distortion, on the other hand, Hasselblad XCD lenses have a clear edge over the Canon 16-35mm f/2.8 III. Stopping down the Canon 16-35mm f/2.8 III to what I consider the standard apertures for landscape photography work fixes vignette almost completely, whereas for distortion you’ll need to add some corrections, when needed. Whether this is important for you or not, of course, only you can say.
On the other hand, what no Hasselblad XCD native lens can do, at the time of writing at least, is covering the range the Canon 16-35mm f/2.8 III does. No Hasselblad XCD lens goes as wide as 14,8mm FOV equivalent, or 12,5mm FOV equivalent if you shoot square or 16:9 and 2:1 panoramas. No Hasselblad XCD lens will cover all the focal lengths in-between 21mm, 30mm and 35mm as the Canon 16-35mm f/2.8 III does. More importantly, no ultra-wide-angle lens for medium format is as fast and as wide at the same time, which is great for night photography, northern lights, and the like. Again, whether this is important for you, only you can say.
Under a technical point of view, the Canon 16-35mm f/2.8 III is an amazing performer, both in general and especially so considering that we are using it on a 44 x 33mm sensor. At infinity, it’s razor-sharp in the center at all focal lengths and all apertures, and it’s perfectly sharp all over the frame at all focal lengths and apertures as soon as you stop down a bit, depending on your chosen focal length. At close distances, the lens is perfectly sharp all over the frame at all focal lengths if you stop down past f/5.6-f/8. Between f/8 and f/16, the standard working apertures used for near-far compositions in landscape photography, the Canon 16-35mm f/2.8 III will output perfectly sharp images all over the frame and all over its useable focal length range.
More, the lens features great macro and micro-contrast as well, is extremely resistant to flare, produces no ghost images and has no colour fringing or chromatic aberrations to speak of. With its Adobe Camera Raw lens profile applied, the Canon 16-35mm f/2.8 III features very little distortion, which you won’t find disturbing in real-world landscape use, but you might need to correct for architectural work. Vignette is heavy at larger apertures, especially so at the wide end of the range and again at 35mm, but clears up nicely stopping down. Using the lens wide-open, you’ll need to correct it in post-processing if you prefer vignette-free images.
Artistically speaking, keeping in mind that I haven’t used it in the field yet, I think that the Canon 16-35mm f/2.8 III draws beautifully, with a gentle and smooth transition between in-focus and out-of-focus areas and beautiful, if a bit cool, colour rendition. What is most important, however, is that this lens opens up some unique possibilities in terms of reaching into the ultra-wide range for medium format users. Whether you use a Hasselblad X1D systems like me, or a FujiFilm GFX based system, the Canon 16-35mm f/2.8 III will fill a gap in available medium format lenses that no other lens can fill at the time of writing.
LIMITATIONS
As a non-native lens, using the Canon 16-35mm f/2.8 III on the Hasselblad X1D means that you’ll need to use the camera’s electronic shutter. More, since autofocus doesn’t work with the Canon 16-35mm f/2.8 III (at the time of writing, at least), you’ll need to use manual focus. Last, if you want EXIF to report the correct focal length, you’ll need to remember to zoom to your selected focal length and then restart your camera for the lens’ focal length to be recorded correctly.
While slightly annoying compared to using a native lens, all these limitations are of no consequence for the kind of landscape work I do. Of course, if you do a different kind of photography your requirements, and therefore your conclusions, might vary.
Please note that the above limitations might change using a different adapter, and that using this lens on FujiFilm GFX cameras, thanks to their focal plane shutters, you won’t need to use electronic shutter if you prefer not to.
A MUST-HAVE LENS? THOUGHTS ON THE FOCAL RANGE & USE VS PRIMES
As always, the choice of focal lengths for our lens setup is a very personal matter. Personally, I love working with prime lenses, for the visual discipline they give me and because they help me to “focus” my vision, so to speak. As I mentioned in previous articles, in the past I went as far as using my zoom lenses as “stepper” lenses rather than as “real” zooms, e.g. using my 24-90mm zoom lens setting it to 24 – 35 – 50 – etc and trying not to use in-between focal lengths.
However, zoom lenses offer a convenience that prime lenses don’t. Since I do landscape only, very often working in unfriendly environments and under bad weather, using primes isn’t always the best choice. More, when it comes to the Canon 16-35mm f/2.8 III’s focal length range, there aren’t many medium format primes that could replace it to begin with.
When I work, I always carry two X1D II bodies, both to have a backup and for more flexibility in the field. Since I got the Hasselblad XCD 35-75mm, my setup became having the Hasselblad XCD 35-75mm f/3.5-4.5 permanently glued to one body and having a second lens (most often the Hasselblad XCD 21mm) attached to my second body. For destinations requiring wider lenses only, I’d keep the 21mm on one body and use my Voigtlander 15mm on the other body. I also carry the Hasselblad XCD 135mm with me for those situations when I need a longer option.
The Canon 16-35mm f/2.8 III would replace both the Hasselblad XCD 21mm and the Voigtlander 15mm in my bag. Dedicating one body to this lens, and the other to the Hasselblad XCD 35-75mm, would allow me to minimise lens changes even more. At the same time, it would give me more reach into ultra-wide territory, adding some missing focal lengths in the Hasselblad XCD line-up as well, such as a 21mm FOV equivalent. As well, it would bring back the 24mm FOV equivalent that I loved with the Hasselblad XCD 30mm and that I sacrificed when I moved to the Hasselblad XCD 35-75mm.
However, the only thing that counts, in the end, is image quality. Will I lose image quality replacing the 15mm Voigtlander + Hasselblad XCD 21mm with the Canon 16-35mm f/2.8 III?
After this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review, my answer is that while I would never consider the Canon 16-35mm f/4 IS for my work on the Hasselblad X1D, choosing the Canon 16-35mm f/2.8 III instead I would not lose image quality at my usual working apertures. Vignette and some minor distortion aside, which can be fixed easily in post-processing, the optical quality of this zoom is surprisingly good on the 44 x 33mm sensor, rivalling Hasselblad XCD and Voigtlander primes at equivalent focal lengths and sometimes besting them.
THE LAST WORD
Overall, the pros of the Canon 16-35mm f/2.8 III, for me, outweigh the cons. I will keep the lens and bring it with me during my next outings to try it in the field side-by-side with the Hasselblad XCD 35-75mm lens and see if it will earn a permanent place in my bag.
Last but not least, there is something very important that this article proved. After seeing what the Canon 16-35mm f/2.8 III can do, I am fairly confident that it is possible to make a high-performance, ultra-wide-angle zoom with a range of 16-32mm for FujiFilm, 16-35mm or 17-35mm for Hasselblad, covering the 44 x 33mm medium format sensor while keeping its size and weight compact enough, especially if you go for a relatively slow f/4 aperture. Hasselblad & FujiFilm, please take note: I am pretty sure that the first out with such a lens, if done properly, will get a lot of business from us landscape photographers. I, for one, will definitely get it, and very likely I’ll get it no matter who does it first.
Thanks for reading this compared Canon 16-35mm f/2.8 III vs Canon 16-35mm f/4 IS in-depth review, I hope you enjoyed it! Why don’t you share it with your friends, or drop me a comment to let me know how you feel about this?
Have a great day, and don’t forget to SUBSCRIBE TO MY NEWSLETTER!
Enjoying the blog? Support us with a PayPal donation:
Thanks for providing this useful information.
You are very welcome, glad to help. Best regards,
Vieri
Hi Vieri,
Thanks for this test! Have you ever used a Canon TS-E 17mm on X1D? I heard that it will cover medium format sensor quite well.
/Christoffer
Hello Christoffer,
thanks for your comment, happy you enjoyed the review. No, I never tried the Canon TS-E 17mm during my time with the X1D, but talking with people who did, I got the impression that, while the 17mm would definitely cover the sensor, the 24mm TS-E would be a much better option in terms of lens quality.
Best regards,
Vieri
Hi Vieri,
Do you consider to test it on X2d camera?
Hello Wilson,
sadly, since moving to Phase One in 2020, I haven’t got any more Hasselblad equipment, nor the lenses in this review, which I sold – so, I am afraid I won’t be able to test this lens with the X2D, sorry.
Best regards,
Vieri
Did you end up using this setup for landscapes? I’m curious because I’m interested at it for my x1dii. That or just go for the more expensive 30mm xcd lens. It’s hard for me to decide since everything is so expensive and this is a hobby. But i love the quality from the camera. I do a lot of waterfalls here in Arkansas where I live. And with film I’ve got by with my Pentax 67ii and 55mm lens. Which is around a 28mm. I have needed something wider in certain situations. Thanks for any input
Hello Justin,
yes I did, and it worked just fine. Hope this helps! Best regards,
Vieri