Old wide-angles (not so wide on APS-C, actually) on modern mirrorless
NOTICE: I am not an English speaker. Some phrases in this text may be incorrect or sound odd,
but I think that this sort of article must reach the widest public: not many
are interested in trying legacy lenses on modern, digital cameras,
and if I were writing in Italian, the not-so-many would become a few ones!
Following
this
test with some legacy lenses, where the test was performed on a target near the camera (less than 2 meters,
not useful for the purpose), I wanted to try the wide-angle primes also against a more proper subject: a building
at least 5 meters away. The distance has been definitely surpassed, because the target building was more or less
12 meters away. That is fine to judge a lens also for focus to infinity.
Introduction, and the Fujifilm way
When the first X-trans sensors were out, I was very intrigued; for some time I wished I had a
Fujifilm camera with one of them onboard.
At the same time, I feared that the first models were not yet "there", so I waited till the XT-20
was on sale, though the XE-2 was already a strong temptation (but in the meanwhile I purchased my
beloved Pentax K-50, it was meaningless to try something else so soon).
Eventually, I became a Fujifilm fan, in a way.
If you like the feeling for reflex cameras (everybody who comes from the film era perfectly knows
what I'm talking about), don't buy this camera. In general, mirrorless photography with small body
cameras is another story.
Again, I love the XT-20, including its compactness, but... man, for one who always used
"traditional" DSRL cameras, those controls all crammed in such small space (that you activate
or slide by mistake quite often). And what about the forest of options in the menus.
Overwhelming, sometimes confusing.
But you'll get used to them, with practice.
The only real issue with this wonderful item is the grip. Too short, you never feel the hand
is at ease. Sorry, much better could be done and I'll soon start looking for an accessory
that solves the problem. I'll put it off only when maximum compactess is required.
And just one last thing... it is a bit on the expensive side. I mean, all of the Fuji
equipment is! And the scarcity of third-party lenses doesn't help.
On the other side... well, you probably know by online reviews:
- the sensor is fantastic
- in-camera JPEG processing is unmatched. Really, compare to any camera of any other brand and you'll see!
- it's small. Sometimes a damnation, sometimes a blessing. With a pancake lens can fit in a pocket
- mechanical quality of the body seems excellent. Time will tell
- looks gorgeously old-style
- some controls are just like where they should be, including shutter speed
- it has plenty of features
- mechanically speaking, Fujinon lenses are superb
I'd like to tell you more about the matter, but I'd rather move to the approach Fujifilm has taken with lenses,
and from this point of view the Fuji item I'm here comparing to legacy lenses is exemplary.
Their optics are intended for APS-C and strongly optimized for the digital world. The Fujinon 27mm is an
extremely compact pancake lens with a remarkable sharpness for the price and size, how did they achieve this?
Well, one fact is that sensors, comapared to films, don't like oblique light rays. So, a modern
wide-angle likely relies on an optic scheme that differs from those form the past.
But also, Fuji seems to me to have followed a pattern: they are ready to sacrifice what is too
hard to fix with optics if it can be easily fixed with software-processing.
Here we are: their 27mm does a great job sacrificing distorsion correction, because that is the
easiest issue that can be fixed in software. Combine this with the fact that in-camera software
knows about the specifics of each Fuji lens, et voilat, the image is... so good!
The contenders
These are the lenses that I'm putting against each other in this short review:
- Takumar-A 28mm f/2.8
- Vivitar MC 24mm f 2.8 (no. 9---*)
- Vivitar MC 28mm f 2.8 (no. 98105672*)
- Fujinon Super EBC XF 27mm f/2.8
Understand that 24mm optics are harder to design, so obviously such length is fighting with a gap.
* the Vivitar-branded lenses were produced by a number of firms; a lot of information can be found at
the Great Vivitar 28mm bestiary,
or
on this site
and
here,
where an attempt to account for most the variants has been made.
My two Vivitar's, then, were probably made by Cosina.
The verdict: sharpness
There is no story here: not perfect a lens, the modern Fujinon nevertheless wipes out the competition, though at f/5.6 the Takumar-A (aka Pentax) comes close and af f/8 even does better (look at the grid on the bottom-left corner).
Yes, you could pick the latter for a stroll and rely on the fact that with the X-trans sensor you can push the ISO up and stay at f/8, but without auto-focus you'd strive in low light so the only marginal scenario for the Pentax to win is when you know for sure that you're going to shoot outdoor. Is it worth the effort?
Unless you care more for other characteristics...
The Vivitar items are never a match, so I'm not discussing them here.
Just a final notice: the top performance is reached at f/5.6 by the Fujinon (
as you can also see here), at at f/8 by the Takumar.
The verdict: aberrations and distorsion
The story is totally different here: the Fujinon displays a marked barrel distorsion (photographer usually quantify this distorsion as minimal, I honestly don't like much to see curved windows!) and gets a beating from the other items, even from the 24mm.
As already mentioned, you can easily fix this via software and the camera does a perfect job, too: you won't see it in JPEGs coming straight out of it... not to mention the fact that this sort of distortion is mostly disturbing on architecture subjects, you probably wouldn't care much in other situations. Do you hate seeing slightly crooked people at margins in a group portrait?
On the other side, you see spherical aberration playing a major role, as it affects the sharpness of the legacy lenses in a perceivable way: here the Fujinon comes out first, again; also notice that it is the only lenas that shows no lateral CA (Chromatic Aberration). In the end, I think that the winner is clearly the Fujinon, much helped by featuring an aspherical design.
The verdict: colours
You can oly see this glancing at the image as a whole, but the Takumar-A 28mm here shows the qualities for which the good old Pentax lenses were much appreciated: rich and saturated colours, somewhat even better respecting the perception I had when shooting for this test. So the Pentax/Takumar wins here, but again, this is a characteristic you can tweak via software, or, for JPEG files from the Fujifilm camera, by choosing a "film simulation" that better suits your taste.
The verdict: overall
The Fujinon does not just best the other lenses from the optical point of view (asphericity is probably the most important factor here), but it also delivers auto-focus and is much, much compact. Being meant for APS-C, it can probably rely on a more proper design.
It is the lens to go for, even though sometimes I may want to pick the Takumar-A but mostly for the fun of shooting with a legacy lens.
Do not think that, just because three samples from the old film world are surpassed here, that legacy lenses are bad in general: as you can see
here (and more tests are coming), the situation is quite opposite for lenses that go from 35mm to 50mm. Also consider that the three lenses tested here are not considered by any means strong competitors. Alas, I do not possess a Zeiss Distagon, and I'm not going to buy it just for the sake of testing, but I'm sure that things would have been more interesting.
The crops
For each lens and each aperture, I took two pictures, re-focusing each time, manually for the three legacy lenses
but with focus-peaking helper on, while autofocus was used for the Fujinon.
For each aperture (and each lens), only the best shot has made to this page, unless I wanted to prove something (how spherical aberration affects the performance).
Obviously, the information automagically extracted from the EXIF data in the file, that is reported inside the image
itself, is only valid where logical; for instance, if aperture is reported for a legacy lens, you must ignore it,
beacuse the camera cannot communicate with the lens. Shutter speed is right, instead.
The center crop is placed in the top-right part of the image; the border crops come in the four
squares that occupy the left and the center columns.
Fujinon 27mm/2.8
Fujinon 27mm at f/2.8
Fujinon 27mm at f/4
Fujinon 27mm at f/5.6
Fujinon 27mm at f/8
Takumar-A 28mm/2.8
Takumar-A 28mm at f/2.8 (the discarded shot for this aperture, focused at infinity, had a much better left-bottom corner, but was definitely worse in the center). The winner was obtained focusing a bit before infinity
Takumar-A 28mm at f/4 (focus point at infinity, for this aperture I want to keep both the images, so you can see for yourself how non-asphericity makes center and border crops be better in one case or another)
Takumar-A 28mm at f/4 (focus point before infinity, see comment in the other f/4 shot)
Takumar-A 28mm at f/5.6
Takumar-A 28mm at f/8 (focus point at infinity)
Vivitar 28mm/2.8
Vivitar 28mm at f/2.8 (focus at infinity. The other shot had two better corners but a much worse center. Spherical aberration, again)
Vivitar 28mm at f/4
Vivitar 28mm at f/5.6
Vivitar 28mm at f/8
Vivitar 24mm/2.8
Vivitar 24mm at f/2.8. Focus at infinity: corners are not bad for this aperture, but the center is ugly.
Vivitar 24mm at f/2.8. Focus a bit before infinity: a much better center, but the corners are spoiled.
Vivitar 24mm at f/4. I keep this one, but the other shot had a better center, not much though (just slightly improved over f/2.8)
Vivitar 24mm at f/5.6 (and again, the other shot had a bit better center)
Vivitar 24mm at f/8. Only at f/8 the spherical aberration mitigates its effects
I wish that you, dear photo-friend, will find these tests useful. Long live photography!