Buying Guide: The best lenses for Fujifilm X-mount mirrorless cameras

Because it's extremely slow to autofocus. Most people suggest to get the 35 f2 instead.

Maybe you should create a "Nifty Fifty" category ;-)

It is not that they don't open the mount to 3rd party, but lens makers like sigma finds that the market is too small for them to make a profit.

Indeed, best primes are the old ones and they miss the new pricessors, the new motors, WR. Also, corner sharpness is not great with the entire lineup. There are colour fringes here and there. Instead of addressing these issues, Fuji is milking us with surging prices and cost cutting by exporting production lines to the Philippines. I have personally seen the IQ going down with the production line relocation and there are reports from other users as well, especially concerning 18-55.

Fujifilm is the new Canon!

Not having a choice shows great disrespect to the loyal user base. Sorry, but Fuji is not Leica.

just wait for them to upgrade all their 1.4 lenses to mk2. it will be much better.

In the bin where it belongs as its a hateful abomination of a lens.

The Sony 18-135mm is pretty good. I also like the Sony 35mm and 50mm primes. They're not as sharp or fast as the Sigma primes, but the stabilization is really useful.

Lenses are a big issue if you want any kind of good telezoom or fast zoom without spending ungodly dollars on heavy FF glass.

To me it is the 35mm 1.4

This lens is disaster even when stopped down. Even Russian lenses like Helios are much much better in terms of IQ.
I completely lost confidence in Fujinon when I got this lens, sadly I can't return it.

Sounds like you guys all (somehow) got lemons, or you're bent on repeating what others say.

The 23/2 is one of the sharpest, best lenses in Fuji's lineup. Check out Optical Limits.

My 23mm f2 is tack sharp on my X-T2, maybe you got very unlucky?

My copy is solid and a favorite

My 23mm F2 is also really sharp on the X-T2. One of my favorite lnses (and I have a few).

Yep, another one here whose 23mm f2 is a solid performer. It has no noticeable distortion and it rivals my X100F's 23mm lens with the advantage that it's a little sharper at the edges than the fixed lens. My only real complaint is that my copy only provides noticeably better resolution than my 18-55 zoom at the same focal length up to around f5.6. From f8 and beyond the zoom equals and perhaps even surpasses the prime. Oh well, I suppose I shouldn't complain and just be happy I have such a decent zoom.

This brilliant lens should definitely get a mention. A great wide for day time and astro work - assuming you can find its infinity point.

And the 27, which is just oodles of fun.

depending on what im up to the 56 1.2 or 14 mm.... but the 18-55 2.8-4 stays on my bodies a lot because its a lovely lens and versatile

i really enjoy them all..... prior to the 56mm the 35 1.4 got use cause its FAST

Similar to your suggestions, my travel kit is usually:
-- 35mm 1.4
-- 56mm 1.2 r
-- 10-24mm 4
-- 55-200 3.5-4.8 (less used - often in a pouch in my day pack)

I think those comments show lack of understanding about photography.
Each system has it's own advantages. Both the RP and A7ii are very slow operating cameras with the fastest barely reaching 5 FPS. Neither show decent video with the A7ii being the worst with no 4K option.

APS-C clearly has speed advantages in a given price segment and Fuji produces fascinating out-of-camera colours imho. These are two strong arguments for the X system.

But the lenses make this system expensive when comparing to full frame offerings. F/1.8 zooms (equivalent to full frame f/2.8) are missing completely, also because this is a closed system. The size advantage is more due to mirrorless than due to sensor size. And speed will eventually be good enough with any serious camera.

Of course the new full frame mirrorless offerings are incomplete yet. Nevertheless I think that 'buy a different camera' is the right advice for anyone who wants to build up a system with a couple of good lenses, no matter if coming from m43, apsc, or full frame DSLR.

To formulate it more optimistic: Fuji must not rest on their laurels of having the most complete APS-C system. The X system would benefit from const f/2.8 kit lenses to compete with new full frame f/4 kit options as well as 16-50 f/1.8 and 50-140 f/1.8 to compete with the popular full frame 24-70 f/2.8 and 70-200 f/2.8.

again i dont buy these equivilency claims a fuji 2.8 lens if projected on a ff sensor throws the same intensity of light, but of course would have a smaller image circle.
and that tiresome argument about sensors. I love how these, carved in stone, formulaic conclusions are abandoned when convenient

cell phone s even ones that fuse 3 sensors rarely have above 65 sq mm of size yet they certainly punch above their weight due to faster adoption of
technology for tiny sensors
similarly 44x33 sensors"mf" has a sensor 70% larger than "FF" does not matter cause the generally accepted idea is
the large sensors refresh slower tech wise than ff [ and it seems true]

i will tell you that fuji iq wise punches at the weight of many 24x 36 sensor cams 7 a fuji 2.8 lens is a 2.8 lens period

if apsc speeds are diminished compared to ff, then lets insist that cellphones w claimed apertures of 1.7 or f2 be described as the f11 to 16 equiv as the mentality seems to mandate.

cosinaphile - I'm afraid equivalence is provable, regardless of whether you buy into the idea.

F2.8 is F2.8 but a 16mm F2.8 is not equivalent to a full-frame 24mm F2.8.

It's just a way of understanding how much light a camera gets to operate with. Because most sensors of a given age perform broadly similarly, you can extrapolate a rough expectation of some performance metrics from that, but naturally you add some imprecision when doing so.

But the idea of equivalent apertures is no more or less shaky than the idea of equivalent focal lengths. It's just maths and it directly relates to whole-image outcomes.

But, Fujis have ISO performance as good as the A7ii and better dynamic range than the RP. And much better video.

You'd be provably wrong to compare something mathematically rigorous and demonstrably true to something that was disproved by the scientific method, but sure, there's nothing to stop you.

Where are you seeing APS-C fast zooms being described as slow? I'd generally avoid such phrasing, simply because 'fast' is generally used in an exposure-model (per-area) context, so it becomes messy if you try to apply it to equivalence without extensive explanation. The term 'fast' leads to all sorts of logical errors, but that's a separate issue.

You are again wrong in assuming any agenda. We don't regularly talk about equivalence of smartphones because they tend not to have interchangeable lenses, so it's not a comparison that's so likely to be made (you're rarely choosing between smartphones with radically different sensor size), but it's by no means verboten. Some brands are also a bit vague about their sensor size, focal length and f-number, which doesn't help. But there's no policy or agenda.

As I explain in this article, there will be situations in which multiple formats exceed a 'good enough' signal-to-noise ratio threshold, such that you don't necessarily appreciate any additional difference.

That doesn't mean the difference doesn't exist, though. It's much easier to understand the likely magnitude of difference (so that you can understand outcomes across a range of situations), rather than concluding that it doesn't matter in one situation, and therefore doesn't seem to exist, then start having to factor it back in again in other circumstances.

You see my note about different sensor sizes? That tends to be why we discuss equivalent aperture values in compact camera reviews: because there are often multiple sensor sizes at play and two 24-70mm equiv. F1.8-2.8 cameras may give radically different results if they use different sensor sizes.

In fact that was our primary motivation for including equivalence: because it's a powerful and relatively simple tool for understanding the impact of sensor size differences.

We include actual focal lengths and f-numbers as well as equivalent focal lengths and f-numbers. They are no more theoretical than equivalent focal lengths. They do not express the superiority of full-frame. The 135 film format just happens to have become the de facto industry standard baseline for focal length comparison.

There are differences in sensor performance, so it's much easier to be able to indicate how much light is available to the system (you know, just like f-numbers do) along with DoF implications.

Super relevant comment Ruekon . *thumbs up*

While it’s true that the A72 has seen a price drop, so has the X-T2.

As I say, I'd tend to avoid the term 'speed' or 'fast' because it's generally tied to a per-area, rather than per-image analysis. If you can point me to some examples where we say a lens is slow or fast, then I'll consider changing it.

However, it's worth considering what you mean by 'fast.' Historically it's meant 'can be used with a fast shutter speed.'

If you can shoot an image with an APS-C camera at 133mm and F2.8 and 1/500th of a second, then a FF camera with a 200mm lens set to F4.2 and 1/500th of a second will experience the same amount of light per image. It'll receive less light on any square mm of sensor, but the same amount of light overall. So, unless there's some rule I missed that says full-frame shooters have to view and print their images 2.36x larger than APS-C users, then the images will be broadly comparable.

At which point, in what way is the F2.8 lens 'faster'?

again by your own statement your using lens speed descriptions to explain differences in sensor prowess

which is my point

why not say when you use a 1 inch sensor at iso 800 its equiv to iso
100 or 200 on full frame .. iso 100 on ff is eqiiv to iso 400 on m \43?
would that not at least blame the culprit which is the sensor

....and which is least of all the lens the lens differences are described
best by thinking about the image circles ..... they do not change magically from 2.8 to f 4 just because they are in fromt of a less capable sensor .... but you metrics suggest they do

finally when you say things like a fuji 56mm 1.2 is the same as an 85 1.8
canon or nikon you do a diservice to the truth about those lenses simply to make a point about sensor differences ... fov everyone understands and in not important in this discussion

F2.8 is always F2.8, but this is only useful information if you are only allowed to consider images at a size proportionate to the format they were shot on. Which is not something I've ever seen digital photographers do.

The exposure model in per-unit-area based. Film photographers then used to consider how much they magnified their image (which relates the per-unit area approach to output size). Equivalence does a similar thing: relating capture format to output size.

It doesn't favour any format over another, beyond bigger being better, in terms of IQ potential (but that's just basic physics, which I wouldn't recommend arguing with). Equally, equivalence shows where a larger format doesn't give any advantage, despite the cost and weight it's likely to bring.

no that infois valueable when buying a lens for qualities of speed

calling apsc 2.8 constant zooms and m43 2.8 constant zooms the equiv of cheap
ff lenses of f4 is a disservice to the truth
as is calling a 56mm 1.2 functionally the same as a milktoast 85mm 1.8

when olympus or fuji make a fast 2.8 zoom , your metrics suggest its like a f4 zoom in ff or worse a kit zoom .... a very misleading thing imo

And, no, my APS-C vs FF example didn't explain differences in sensor prowess (though most APS-C chips from a given brand at a given time generally perform similarly, per unit area, to the corresponding FF chips).

I'm simply talking about how much light a system gets. What the specific sensor then does with it is a different matter. But understanding the per-image effect of changing format size in terms of light availability is crucial to avoiding errors such as 'this camera is better because the pixels are bigger.'

A m43 F2.8 lens is equivalent to an APS-C F3.7 lens, which is equivalent to a FF F5.6 lens with the same angle of view, if you plan to view or print the images at the same size.

That is the provable truth.

We will also state the absolute focal length and F number.

I'm afraid you're wrong.

Please read all three parts of this series. Format differences generally make more difference than sensor performance differences.

But thank you for illustrating exactly the error that I said could come from solely focusing on a per-unit-area perspective.

using lens speed alone to describe all the differences that are being discussed here , when its the least blamable culprit is misleading, i stand behind that

there are better ways to point out the real differences considered .

a new numerical metric [using ff as a reference due to the wide understanding ] calling the iq better or worse as it deviates away from
"one" [ full frame] calling it .82 or 1.23 of the "standard " or iso equiv

im still not clear why the lens itself must be the manipulated value.... with its incorrect suggestions of being a "slow lens " which i think is profoundly misleading

when you shoot an apsc camera like a fuji xpro 2 a f 1.2 you have the advantage of stopping power over a 1.8 lens

while i understand that a ff 864 sqmm sensor has an advantage over a 368 sqmm apsc sensor . i think using manipulated f stop numbers is the least helpful and most misleading way of describing the real world differences seen

[ as per unit as you say]

As I say, we tend to avoid using the terms 'speed' or 'fast', specifically because they are a bit too closely tied to a the per-unit-area/exposure model.

Please explain what makes you think an F1.2 lens on one system automatically has more 'stopping power' (?) than an F1.8 lens on a larger system.

In the simplest possible terms: amount of light is the biggest determinant of image quality (light being the fundamental property underpinning all photography), and equivalent values help describe the per-image light properties of two systems.

It's one thing you can be certain of, before delving into the details of each sensor's specific properties. Ultimately it's no more or less descriptive than actual F-numbers.

F-numbers tell you the (approximate: vs t-stosp) per-unit-area light transmission of a lens. Equivalent F-numbers tell you the per-whole-image light transmission of a lens.

Fujifilm 56mm F1.2 and, let's say, a Nikon 85mm F1.8 will project the same number of photons towards a sensor during the same exposure time (notwithstanding any transmission differences, which is just as much a problem as F-numbers). Both sensors see the same view of the world through the same sized hole (47mm diameter).

What they then do with the light will depend on the specific sensors, but in terms of total light per image, there is no distinction between those two lenses.

So in terms of the lens properties, it's reasonable to say they are equivalent. That's all equivalence lets you be certain about (again, this is the same amount of information the actual focal length and f-number give you)

Everything beyond that requires a degree of extrapolation.

If the two sensors offer comparable performance (which is a reasonable but not cast-iron assumption), then you can expect those two images to look very similar when viewed at the same size.

well i would claim the fuji 56 projects double the photons over its image circle that the nikon 1.8 does over its image circle .... because the fuji is one stop faster

the nikons are much more spread out and therefore less intenses as per

"unit of measure"

you are right about the vagueness of f stops i wish they all were forced to describe their lenses in T stops, a much more honest measure

Bear in mind that, in the way film photographers talked about enlargement, blowing up a 24MP APS-C image and a 24MP FF image to the same size are not the same degree of enlargement. And the error that it is, is precisely why we need a framework for understanding the underlying principles, rather than mis-translating the mechanism used when they printed from film using an enlarger.

Yes, you can try to find underperforming sensors and compare them to ones with noise reduction applied to the Raw and prove that sensors perform differently. No one is denying that.

Similarly, we did not say that 44x33 medium format does not have an advantage over full-frame. We said it should have a roughly 2/3EV advantage (and, over most full-frame cameras, it does). However, if a smaller format had the ability to accept 2/3EV more light by, say, having a lower base ISO, then that gap is reduced. A theory that is backed-up by real-world shooting.

You'd be demonstrably wrong in that claim.

The 56mm F1.2 projects more photons per square mm than the F1.8 (a little more than double: it's a 7/6EV difference), but the X-Pro's image circle is around 42% of the size. Which directly cancels itself out.

Please read my three articles. Please read the letter to the scientific journal on the underlying maths.

There are no gotchas. You're not going to find a flaw by repeating the same old arguments at me.

Equivalent apertures provide a per-whole-image idea of how much light a lens provides, just as the actual f-number tells you how much per-unit-area (per-sq-mm) light it projects. It's mathematically rigorous and can be (and has been) demonstrated in the real world.

FF is the reference. Any compact or smartphone camera specifies its focal lengths equivalent to full frame, like 28-56mm for iPhone X. This makes a lot of sense since sensor sizes vary.

cosinaphile and ecka84 - this is a beginners' guide and no place for personal attacks. If you want to discuss this issue further, please use the forums but criticise the other poster's point, not the poster.

With regards low ISO noise, I'm seeing similar levels of noise in ISO 200 shots from full-frame cameras, as I am in ISO 100 shots on APS-C ones.

Whether this noise level matters to you is a separate issue (it'll be worse in lower light), but it demonstrably exists.

However, while this is the outcome that equivalence leads you to predict, equivalence itself is just about how much light the sensors get, it doesn't guarantee how well they use it.

Trying to compare either of these or Mark them as alternatives to the Fuji crowd shows you have no clue what it’s like to shoot with the Fuji! It’s simply a better camera than either of those bodies!

Film sensitivity did not alter the amount of light your camera got to work with. Changing film formats did (but a more overt need to consider enlargement, and relate per-unit-area exposure and capture area to final image size acknowledged these differences).

Your attempted comparison makes no sense.

The 1.532x (Sony) APS-C percentage is 42.55%, which is 1:2.35 (the area ratio). So, if you want precision:
APS-C ISO 200 = FF ISO 470, because 200/0.4255=470
FF ISO 200 = APS-C ISO 85, because 200*0.4255=85.1

My thanks to the participants of this discussion. It was interesting yet not new. Amazing read for a comments section- that it was.
@Richard: let me buy you a beverage when I am in Seattle again :) thanks for the energy invested.
@Cosi: I cant make that offer to you as I don’t know where you are from.
Enjoy your weekends.. :)

@cosi: :) great place to be!
Offer of course stands when I get a chance to stop by :)
A sunny weekend to you

I dunno about far greater... M4/3 does have more low budget options, and it definitely has more zoom options, Fuji is pretty evenly matched as far as primes go tho and they have lower priced weather sealed primes which M4/3 is still lacking... OTOH all the pricier Pro/PL M4/3 primes (usually $1-1.2K to Fuji's $900) are really fast to focus and weather sealed, so things flip flop depending on priorities. I'd be happy with either lineup tbh.

If dpreview lens database is complete and correct, there are currently:
- 63 lenses with native X mount,
- 117 lenses with native u43 mount.
u43 has many lenses that are very similar and article with recommendations would be very useful!

It is a Dpreview buying guide. Thank you for sharing yours.

It's more that our lens buying guides only include a general walk-around prime (and, new for this one, a portrait prime), and also that we only cover autofocus lenses, since the guides are aimed at a general audience, not all of whom will necessarily want to start using manual focus lenses.

That's an ingenious solution, but 18mm-equiv primes still don't fit into our 'all-round prime' or 'portrait prime' categories.

But if you create a new comment (not in this discussion) explaining how you work and what you like about the 12mm, I can give that post recommended status, so that readers of the article are more likely to see it.

Done: https://www.dpreview.com/reviews/buying-guide-best-lenses-for-fujifilm-mirrorless-cameras?comment=8308935557

You will find some good suggestions in this link:
https://m.dpreview.com/reviews/buying-guide-best-lenses-for-fujifilm-x-mount-cameras?utm_source=self-mobile&utm_medium=marquee&utm_campaign=traffic_source

I got the 35 1.4 with my xe1 and still use it on my xt2. It focuses a little slowly but indeed - wonderful images.

35mm f/1.4 was the first lens I got for my X-Pro 1 (relatively) long time ago. Fast and compact and very sharp. 16mm f/1.4, 35 and 56 f/1.2 make a perfect old school prime set for X-series Fujifilm bodies.

There's some great articles over on the lens rentals blog. If you search for variance it will return articles about various lenses and brands.

https://www.lensrentals.com/blog/

@dansclic Sigma does have a fair number of APS-C lenses, so I could see it happen... But generally speaking, yes Sigma makes HEAVY lenses.

Fuji has not released their lens mount specs for the electronics and Sigma has stated they are not likely to produce them as the reverse engineering effort to get acceptable autofocus performance would be too great.