Decent lens, but far from perfect...

When I moved from Nikon to Fuji I needed 35mm fast prime I am used to, so I decided to buy Fujifilm XF 35mm F2 R WR. I previously had AF-S DX NIKKOR
35mm f/1.8G, and Sigma 30mm f/1.4 DC HSM Art so it is inevitable to compare the new Fujifilm 35mm f2 lens with those two.

Fujifilm XF 35mm F2 R WR is a little bit soft wide open, close up picture quality is not very good, but it has decent contrast and chromatic aberrations are not that problematic. Compared to Nikkor 35mm f1.8G it is not as sharp, but it has less CA, while compared to Sigma 30mm f1.4 DC HSM Art it has less ghosting, more contrast, and about the same sharpness wide open while it better controls longitudinal CA.

This lens has very good build quality, but the picture quality is just about average. I have only 3 Fujinon lenses, but this 35mm F2 R WR and 18-55mmF2.8-4 R LM OIS I would call underdeveloped. Fuji decided to correct CA and distortion in camera, so those two lenses have real optical problems without being very sharp no matter how stopped down they are. Both lenses have extensive levels of distortion, which is maybe not a problem for end user, but that is something I am not used to.

For my, this is all about compromises. I needed something smaller than my Nikon, and Fuji X-T20 with those 3 lenses I have is really what I needed, and I really like my gear.I am not saying that Nikon gear I sold was perfect, oh it was not, it was far from perfect. I am saying that there is some hype about Fujinon lenses, people often brag about how good and sharp they are, also their price is noticeably higher than Nikon and Canon crop sensor lenses, I was expecting them to be sharp and optically superior, but that is just not the case.

This is your second thread where you find IQ average with lenses that are very good to excellent according to all reviews and experience here

So if one sample can be average (abd should be returned) the probability that several Xf lenses are average is very very very low indeed

I suggest you post samples in high res in order we can make up our mind about the issue you get

The 35mm f2 is very sharp

http://www.photozone.de/fuji_x/985-fuji35f2?start=2

Again what you find is against the general observations of users

Last but not least the different settings and options of the AF system deserve consideration to get the best of the system

Thank you very much. As with 18-55 f2.8-4, I tested my 35mm f2 with few more in the store, also I compared my RAF files with some other I found online, and all of them were the same. I suppose many Fuji owners shoot JPEG and the look at their photos with in camera sharpening and noise reduction, which is what I see in photos from the link you shared, but my tests include RAF files at 100% zoom, without any sharpening and noise reduction. If processed, my files look exactly the same as those on the site you shared.

I would be grateful if you could look at my RAF files, please tell my what kind of photos do you want me to make, and where to send the files.

Well, I am talking about lens sharpness without developing, with sharpness setting to zero in RAW developing software, just looking at the RAF file showing just the lens. As for developing, I can extrude some details from the files and then my photos look relatively normal.

alfaholic wrote:

I would be grateful if you could look at my RAF files, please tell my what kind of photos do you want me to make, and where to send the files.

Although I agree that the 18-55 is not ultra sharp, I feel it is as good or better than the best of its type. I'm surprised that you mention distortion and especially the lack of software correction in which you favour instead of optical correction. Surely this runs counter to what every right minded photographer would prefer.

I have not owned the 35mm f2, I prefer the 35mm f1.4 (which has its own peculiarities). However, from recent posts I've viewed, it has good sharpness even at fairly close quarters (food photography). https://www.dpreview.com/forums/post/60509549

When you say close focus - were both Nikon and Fuji 35mm lenses shot at the same distance - or simply the closest focus for each lens?

Have you posted some Raw files we can look at and process? Try Dropbox maybe leaving a link here and also post some some full size jpgs of raws which displease you.

Vic

alfaholic wrote:

Well, I am talking about lens sharpness without developing, with sharpness setting to zero in RAW developing software, just looking at the RAF file showing just the lens. As for developing, I can extrude some details from the files and then my photos look relatively normal.

Raw files should be flat with lacklustre colour and little sharpness - that's why they are Raw and need developing. I find that once tonal contrast is returned to the raw file the image immediately takes on sharpness.

Don't expect to process Fuji RAF files as you did Nikon NEF.

Vic

Thx for detailed answer

I do not have the 35mm f2 but the 1.4

I have the 18-55mm

Landscape photos are usually not a good subject for test

It is important to have all tests of the 18-55mm at f4 which is the best sweet point

Turn OIS off in case it could perturb ssharpness

Use 18 35 55 mm FL

Different targets : slates on a roof is a good example (a day without haze)

Easier pages of a book or a magazine

Most important for quick test is to choose flat targets

A wall with bricks

I use dropbox for transmission but I am sure you can get also service from wetransfer or similar Just post the links in PM from here

Best

Bob

That's a flawed measure I'm afraid. The human eye is not good at detecting resolution in isolation. It is easily confused between 'sharpness' and resolution (hence the use of high acutance films in analogue days).

In the digital era, you can have as much sharpness as you want. You can't have as much resolution as you want. If you have resolution, you can add sharpness later, but not the other way round.

If you're testing the lens scientifically, there are ways to do that (look round the web, it's not something I bother doing myself). If you care about how the lens looks in real life, experiment with different workflows until you get the results you want and don't worry about whether that's scientific.

It's possible you have bad copies of the 18-55 and 35, but the more likely issue is workflow.

Getting a final image from a RAW file is like baking a cake from raw ingredients, with the main difference being that you can do it again and again until you get the result you want. Having the right ingredients is essential, but you can still produce a horrible cake from good ingredients.

I never said I prefer software over optical corrections, it is just the opposite. This lens is not optically corrected, nor the 18-55 f2.8-4 and that is why I personally find them underdeveloped.

I know RAW files should be developed, and I know how to develop my photos, I am talking about the lens and how sharp it is. Also, I know how to process Fuji RAF files, again it is not about processing, it is about the lens, it is not very sharp by my standards.

Here you have some quick examples, shot on a tripod, around 55-60cm from the subject, look at the center of the frame where all lenses should be the sharpest. I included 18-55 as well, if this is sharp for you, than we do not have the same standards.

https://www.dropbox.com/s/gt20mlukfkozvwh/18-55%20vs%2035.zip?dl=0

Well, there is no need to play with words, but use the one you like the most. Call it resolution, I agree it is maybe a better choice.

It is very simple really, in my universe this is not sharp:

And no, my lens is not bad, I tested few of them in the store, and all of them were the same.

Also this has nothing to do with editing and processing, the lens should have good resolution in the first place, then you have something to work with.

Maybe you were responding to me. I'm not saying your terminology is wrong. You are probably right to say sharpness, and complain about sharpness. What is wrong is to assume that this is solely a lens issue, or tells you anything about the resolution of the lens.

Resolution is a lens issue.

Sharpness is a processing issue.

Your image is not sharp. However, I can't tell if the lens has adequate resolution, as this isn't a good subject to test resolution. You need to use a test chart of something with fine detail. That can include text from a book or magazine, but only from very far away. This image is too close and there is no fine detail. Also, I don't know the size of the text.

Brick walls are popular as bricks are a standard size. However, you have to shoot them from far away, as you are trying to resolve the bricks, not the texture on the bricks (which can vary considerably between bricks!)

Incidentally, if you are new to the Fuji system, I would test based on JPEGs using standard sharpening settings. That way you can isolate lens issues from workflow issues. JPEGs may have disadvantages, but they are a useful benchmark.

You are right, I am talking about lens resolution, not about sharpness in post processing stage.

The size of this text is 12, it is a regular size for most books. The book was around 60 centimeters from the lens, but I can make another example if you want to see.

What I am trying to say is that this lens does not have good resolution no matter what you have in front of you while testing and how far from it you place the camera. JPEGs are relatively good because they are processed and sharpened, and if you ask me whether this lens resolution can be fixed by sharpening in the software, then yes it can to some extent, but still it is not what would I call sharp/detailed/good resolution, and if you ask me it is always better to have a lens with good resolution to start with.

And it is very simple, Sigma 17-50 f2.8 has much better resolution wide open than this lens, it can catch details that simply can not be seen on this lens. Maybe that sounds harsh to you, and maybe you find it offensive, but my tests showed that.

I'm sorry, but we're talking about cross purposes.

60cm is far too close to measure resolution for size 12 text. Even a pinhole camera could resolve that text. What you are complaining about is that it doesn't have a sufficient acutance. I recommend you consider the article below (others are available online):

http://www.cambridgeincolour.com/tutorials/sharpness.htm

Have a look at the high resolution low acutance photograph example, and see how 'unsharp' it looks.

I'm not offended, I'm not a Fuji employee. I just want you to enjoy and get the best out of your equipment.

So you want to say that this soft image you can see in post 22 shows the lens with very good resolution but low acutance?

Greg Johnson, please report to the OR.  Code blue.

I can't tell if it's high resolution or low resolution because the the subject has no fine detail (as I said, the text is too close to be a real test of the lens and sensor). It is definitely low acutance.

Look, ultimately this stuff may not matter that much to you. I know I get accused of being too theoretical sometimes. If you get the results you want, none of it may matter that much. However, if you aren't satisfied, it's worth isolating exactly why.

The complaint you make (that the image is not sharp, but does improve when processed) suggests decent resolution and a lack of acutance. If you have a lens that does not resolve well, you can't reclaim it in post-processing; the detail from the scene is simply not recorded. That's why high resolution lenses are sought after, particularly in the digital era.

There have been many, many tests of the 18-55 and the 35 f2, done by a variety of reputable sources. Neither lens is perfect (not even the best lenses are) and if you had complaints about corner softness on the 35 f2, of field curvature on the 18-55 wide open, or that the 18-55 was not so good at the long end, these would all be things that professional testers have found.

What professional testers have not found is that these lenses lack resolution, or are inferior to the Sigma 17-50 f2.8. That suggests you either have unusually poor copies, or a workflow issue.

alfaholic wrote:

For my, this is all about compromises.

Heh.    That's photography in a nutshell.

I had 4 copies of both lenses in the store, tested them all together, there is no difference. Those lenses are the way they are. This was the first I checked, my friend owns the store so he gave me all lenses he had at the moment there, and we tested all of them.

There is no workflow problem because there is no workflow at the first place. Put the camera on a tripod, turn on countdown timer, turn off OIS, focus, do not breath, repeat the same two more times if maybe something wen wrong with the tripod legs, and that is it. In all my tests those lenses were softer than Sigma 17-50 f2.8.

This is taken from 3 meter distance, both cameras on a tripod, OIS off, countdown timer 2 seconds. The only difference was metering so Nikon measured the whole scene while Fuji measured only center of the frame.

Am I crazy, or there is some hidden details in this RAF file?