p.2 #1 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
Didi88 wrote:
Nope, bought it second hand. No returns. How would you explain sharp center and soft sides when focused on the center, and sharp sides and soft center when focused on one side? (my last post)
If it is just the sides that differ from the center, your camera's sensor may not be perfectly parallel to your subject. (Is it sharp at top/ bottom edges?)
As pointed out elsewhere, it could also be a field curvature issues. One report that I looked at just now (http://www.photozone.de/fuji_x/783-fuji1855f284?start=1) says that there is a field curvature issue wide open at the 18mm end of the zoom range, but that it improves a great deal if you stop down. See the link for details.
Field curvature is not an unusual issue with lenses, and it is more likely to show up at the largest aperture since this narrows the DOF. I haven't looked (see above) at your examples in detail nor do I know the settings you used, but if you were shooting at the maximum aperture and at the wide angle end of the range, this could be in play.
p.2 #2 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
It could be field curvature, But I'm having it on all focal lengths. Also, field curvature shouldn't be visible in optical infinity, and that's exactly what i have. It could also be some kind of sensor misalgnment.
I updated the firmware today and it didn't make things any better.
I had a few misaligned lenses, and I could easily tell they are misaligned, but this... I have no idea what is this
p.2 #3 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
mdude85 wrote:
You could also just sell the lens and buy a new one.
Or you could just accept the lens for what it is and go out and enjoy shooting with it. I don't mean that to be rude, but you also have to think about how much time you're spending (wasting?) doing lens tests when you could be out there creating.
I don't want to sell bad gear to other people.
Accept the lens as it is and enjoy it, you say? The pictures I get are absolutely unusable. I have a moderately misaligned canon 85mm 1.8, and underperforming canon 70-200mm f4 IS. They are perfectly usable to me and I don't worry about them, but this fuji is just horrible
p.2 #4 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
Didi88 wrote:
I don't want to sell bad gear to other people.
I have a moderately misaligned canon 85mm 1.8, and underperforming canon 70-200mm f4 IS. They are perfectly usable to me and I don't worry about them, but this fuji is just horrible
Now you seem convinced that you have a bad lens (even if you don't), in which case you only have one option if you don't want to shoot with it -- send it to be repaired, and take a risk that you might spend $250 without any tangible result. There's only so much you can do in the camera (updating firmware, changing lens settings) or in post to enhance image quality.
But what you said about your bad 85 f1.8 and your bad 70-200 struck me. It seems rather unlikely that you would have 3 bad lenses (including the 18-55). To me, that reflects perhaps an expectation of the lens quality that does not match reality of what the lens can deliver. Or perhaps you're buying your gear heavily used. I personally do not know about the Croatian market. Here in the US, if you go with a reputable seller, even 1 bad lens is statistically rare, let alone 3 bad lenses.
The IQ in the first image is not that great, but the van is in focus, and the rocks and trees are out of the focal plane. So you would expect them to be a bit soft. In the second image, I can't tell a problem. I've been using the 18-55 for about 4 years now so I have some idea of what it can deliver (I also shot Canon FF with L lenses with many years before that -- I was using a 6D, which had a similar sensor to the 5DIII).
p.2 #5 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
Didi88 wrote:
[F]ield curvature shouldn't be visible in optical infinity.
It certainly could be. In fact, according to the review at the link I posted, IIRC, the corners may actually focus closer at 18mm wide open.
Some points.
1. No lens from any manufacturer is "perfect." All have "flaws" and compromises of various sorts. These may be performance issues at different apertures and focal lengths, field curvature, vignetting, corner sharpness, and the list goes on. Even a lens that looks "perfect" in some situation can have issues in a different situation. (I'm thinking of my Canon 70-200mm f/4L IS II, an excellent lens that can have some issues wide open at MFD.) So if you are testing to find "perfection..." you will always be disappointed.
2. Once you start looking very critically at lens performance and you try to understand the performance of a lens via "testing," you need to become very methodical and logical in your approach. It isn't enough to point that lens at something and make some pictures and then draw overly broad conclusions. If you are looking to understand focus issues, you'll want to eliminate all kinds of variables from your test: the camera's sensor must be perfectly parallel to your target, your target must have sufficient detail and contrast, your focus must be perfect, you need to try a range of apertures, your camera/lens system must be vibration free (use electronic shutter, a remote release, a tripod, and let the system "settle" before triggering and exposure), manually focus (using camera MF aids) to eliminate AF as a source of the problem. Once you understand the baseline performance of the lens you can introduce other variables such as handholding the camera, using AF, etc.
3. Results need to be calibrated to realistic expectations. For example a cropped sensor camera is not going to equal a FF camera with the same or greater number of photo sites. At apertures other than the supposed ideal aperture performance will be less. All lenses will vignette and resolution will not be constant across the frame. And on and on...
4. The real world implications of the test results are critical. For example, let's say you discover that a lens vignettes a little bit (as all do) and is sharper in the center than in the corners (also typical). Does this actually matter? When and how? Are we looking at something that is only a "problem" because we can detect it if we look closely at something on the test bench?
In the end, if you think you have a "bad copy" of a lens, there are really only a couple of ways to verify this. One is to try another copy of the same lens and see how it differs. Another is to send it in for inspection and adjustment. If, as you write, it is "utterly unusable," that would indicate a lens problem that is far from typical for this lens. I've used it in the past and its performance is actually quite good.
p.2 #6 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
mdude85 wrote:
Now you seem convinced that you have a bad lens (even if you don't), in which case you only have one option if you don't want to shoot with it -- send it to be repaired, and take a risk that you might spend $250 without any tangible result. There's only so much you can do in the camera (updating firmware, changing lens settings) or in post to enhance image quality.
But what you said about your bad 85 f1.8 and your bad 70-200 struck me. It seems rather unlikely that you would have 3 bad lenses (including the 18-55). To me, that reflects perhaps an expectation of the lens quality that does not match reality of what the lens can deliver. Or perhaps you're buying your gear heavily used. I personally do not know about the Croatian market. Here in the US, if you go with a reputable seller, even 1 bad lens is statistically rare, let alone 3 bad lenses.
The IQ in the first image is not that great, but the van is in focus, and the rocks and trees are out of the focal plane. So you would expect them to be a bit soft. In the second image, I can't tell a problem. I've been using the 18-55 for about 4 years now so I have some idea of what it can deliver (I also shot Canon FF with L lenses with many years before that -- I was using a 6D, which had a similar sensor to the 5DIII). ...Show more →
Tell me one thing... when shooting a distant scene, where everything is in infinity, everything should be in focus, right?
In this case it's not. And you are telling me that is a normal lens behaviour?
Yes it is normal lens behavior that some objects could be out of focus even if you focus to infinity.
All lenses have what is called a hyperfocal distance, defined as the distance from the lens (in feet or meters) to a specific point in the scene such that anywhere beyond that point should be at an 'acceptable' level of focus.
By contrast, any objects that fall within the hyperfocal distance may not be within an acceptable level of sharpness, in other words, may be out of focus.
Your image of the van was shot at 55mm @ f5.6, equivalent to 82mm @f8.4 on a FF sensor. Looking at a hyperfocal distance chart (which calculates the hyperfocal distance in feet for a given focal lengths at a given aperture), this image's hyperfocal distance is around 100 feet. This means that any objects within 100 feet of the lens might not be acceptably sharp.
So, if the rocks in the foreground of your image were less than 100 feet away from you, then they might be soft because they are out of focus.
A good rule of thumb is that hyperfocal distance increases with focal length. Thus, when you shoot with a longer lens, you will have to expect that some objects in your scene may be within the hyperfocal distance and thus not acceptably sharp. To ensure that the entire image looks sharp, you should compose your shot to eliminate objects that are within the hyperfocal distance.
p.2 #8 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
I wasn't talking about the rocks in the foreground, but about those trees in the distance. They look absolutely horrible as you look further away from the van
If you have focused to infinity, then objects that are very far away should all be in focus. Hopefully that answers your question.
I don't know what trees you're talking about specifically, as there are a lot of trees in the image. They all look, in my opinion, to be about right for the lens and the sensor. The hills look a bit soft, but I don't think they're out of focus, I think it's just a kit lens and a 16 MP sensor in an OOC photo without any adjustments applied to it.
Do you really need images sharper than this in a RAW file OOC, without any adjustments? If so then you might want to upgrade to a better camera or a better lens.
p.2 #11 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
gdanmitchell wrote:
Not sure if it came up yet, but ALL raw files will be soft until you sharpen them. That's normal — it is how raw files behave.
No they won't. A sharp lens will still render a sharp file without sharpening, but of course less sharp.
The lack of sharpness in the examples provided cannot be fixed with sharpening. Something else is wrong. I suspect x-trans + crappy lens. Or possibly not optimal focus.
If you have focused to infinity, then objects that are very far away should all be in focus. Hopefully that answers your question.
I don't know what trees you're talking about specifically, as there are a lot of trees in the image. They all look, in my opinion, to be about right for the lens and the sensor. The hills look a bit soft, but I don't think they're out of focus, I think it's just a kit lens and a 16 MP sensor in an OOC photo without any adjustments applied to it.
Do you really need images sharper than this in a RAW file OOC, without any adjustments? If so then you might want to upgrade to a better camera or a better lens. ...Show more →
Ok, you convinced me my lens is within what's considered normal for this lens. I also found some full size image samples which confirm that. Mine is maybe i bit weaker in the corners. And it has very strong field curvature. Time to start enjoy shooting
p.2 #13 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
Makten wrote:
No they won't. A sharp lens will still render a sharp file without sharpening, but of course less sharp.
That is one of the more bizarre posts I've seen around here.
I suggest that one reason that a raw file might look soft is that it hasn't been sharpened, which is a standard part of virtual anyone's post-processing workflow with raw files.
You reply that a) unsharpened raw files won't be softer, but that the will "of course" be less sharp."
OK. Got it. Sort of a Schrödinger's Photograph situation, eh? Where the file will be both sharper and not sharper? ;-)
p.2 #14 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
gdanmitchell wrote:
That is one of the more bizarre posts I've seen around here.
I suggest that one reason that a raw file might look soft is that it hasn't been sharpened, which is a standard part of virtual anyone's post-processing workflow with raw files.
You reply that a) unsharpened raw files won't be softer, but that the will "of course" be less sharp."
OK. Got it. Sort of a Schrödinger's Photograph situation, eh? Where the file will be both sharper and not sharper? ;-)
How about reading your own post that I was replying to instead of making things up?
"ALL raw files will be soft until you sharpen them"
No, they won't. "Less sharp" than sharpened doesn't mean they are soft. Certainly not as soft as the examples in this thread.
Anyone can try this for themselves. Just back off sharpening in the raw converter when processing a seemingly sharp image. It won't turn into a muddy mess. Well, maybe if it's from a x-trans sensor.
Edit: Here's a 100% crop with zero sharpening from a sharp lens... I'm not saying the 18-55 should be this sharp, it's only to prove that your generalizing claim is false.
p.2 #16 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
Nope, I withdraw my previous statement, it's definitely something wrong with my lens. In a brickwall test, when I focus on a left side, the right side is blurry. Focus on a right side and the left side is blurry. Or could it be sensor/mount misalignment...
p.2 #17 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
Didi88 wrote:
Nope, I withdraw my previous statement, it's definitely something wrong with my lens. In a brickwall test, when I focus on a left side, the right side is blurry. Focus on a right side and the left side is blurry. Or could it be sensor/mount misalignment...
It's so hard to get what's going on here.
Sounds like something is tilted, yes. Do you have any other lens to try the same test with?
You could also try the zoom at different focal lengths. Tilt doesn't necessarily show over the entire range.
p.2 #18 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
Makten wrote:
Sounds like something is tilted, yes. Do you have any other lens to try the same test with?
You could also try the zoom at different focal lengths. Tilt doesn't necessarily show over the entire range.
II don't have any other fuji lens, nor I know anyone who has fuji. I just sent both body and lens to service.
At 18 and 35mm, tilt effect is obvious. I would say it's moderately strong, but still visible when looking at fit to screen size on a monitor. At 55mm... I have no idea what's happening. 55mm gives me all kinds of strange results.
p.2 #19 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
Didi88 wrote:
II don't have any other fuji lens, nor I know anyone who has fuji. I just sent both body and lens to service.
At 18 and 35mm, tilt effect is obvious. I would say it's moderately strong, but still visible when looking at fit to screen size on a monitor. At 55mm... I have no idea what's happening. 55mm gives me all kinds of strange results.
This is reminding me of a Panasonic 14-140mm I got at a used gear shop some months ago. I tested it in the shop at wider angles and it looked fine on the camera screen, when I got out and took some pictures at 100mm or more the images started to get crazy blurry, the camera would even struggle to focus.
Fortunately I was able to return it to the shop, the technician at the store said it was indeed a lens element out of whack. So I bet that may also be the case with your lens.
Feb 04, 2022 at 07:00 AM
leonasj Offline • • • [X]
p.2 #20 · Do I have a bad copy of Fuji 18-55mm f2.8-4?
you have good copy,all is good.no wast time,just go and shoot.
