My copy of the Fujinon XF 55-200 mm - decentered or within acceptable range?

As you suspect, your copy sounds like it's de-centered some. There's always some variability in lenses, especially "consumer" grade. Unless you might have reason to think the lens might have been damaged you might go back to the seller and trade it for another copy and see if your luck improves.

I have a copy of this lens and have been quite satisfied with it.

Sheyingshi said:

As you suspect, your copy sounds like it's de-centered some. There's always some variability in lenses, especially "consumer" grade. Unless you might have reason to think the lens might have been damaged you might go back to the seller and trade it for another copy and see if your luck improves.

I have a copy of this lens and have been quite satisfied with it.

Click to expand...

Thanks Sheyinghi.

Yes. It's not super bad but seems slightly decentered. I already sent it to Fuji's repair center. The lens is still within the warranty period (24 months in Germany).

My learning is that one really should do a few but significant quick tests before/after purchase.

Hence, I opened a thread on quick (not scientific) testing of lenses.

Allemania also opened a similar thread.

Cheers,

300m away? Are you sure there was no wind at all?

If you can, when the lens returns, test it on something more stationary, like a distant roof, billboard, chimney, statue, even a light pole would be more steady/solid than some small branches on a tree-top. Also due to the contrast, I'm seeing quite a bit of CA which probably makes it look worse than it is.

Hi

After about 4 weeks (home-home), I got my slightly decentered lens back from Fuji's repair center in Germany.

I did not get a report, they just stated that a "focus matching" was done and that the lens is (now) "fully functional". Very brief standard statements.

In the meantime, I have concerned myself with the application of a test chart. I will also report on this lens but need more time for this. I am convinced that one should really test every lens within the return period.

What I first did is: I repeated my simple "tree test" that I still consider not bad with long focal length. Why? - because alignment plays no role (in contrast to test charts)! And: because it's really fast! The tree was in a distance of about 200-300 m. The day was windless. I shot series at 200 mm and at 100 mm.

Main source of error with such tests is in my opinion the AF. According to my experience, it does not always really exactly meet the point. AF was applied only once in each series for the center, then I switched the camera to MF so that the focus was kept constant while panning to the four corners (of course with a tripod). I used the ES in order to avoid any shock, and I used my telephoto lens support. OIS was switched off. The sharpest in the 200 m series is at f/8 but this is most likely due to the accidentally best AF accuracy in this series. From theory, f/5.6 should deliver the best resolution. The sharpest in the 100 mm series was at f/5.6.

Here's the result: I think the lens is a tad better than it was before. But still, bottom left is the weakest corner at 200 mm. The remaining decentration is not really visible at 100 mm. I will make my decision to keep / send the lens back a 2nd time dependent on the chart tests.

Other opinions welcome!


200 mm, center (400 px x 400 px), f/8. Great sharpness.


200 mm, top left corner (400 px x 400 px), f/8. Little weaker than the center.


200 mm, top right corner (400 px x 400 px), f/8. Virtually as good as the center.


200 mm, bottom right corner (400 px x 400 px), f/8. Little weaker than the center.


200 mm, bottom left corner (400 px x 400 px), f/8. Still, the weakest corner!


100 mm, center (400 px x 400 px), f/5.6. Great sharpness.


100 mm, top left corner (400 px x 400 px), f/5.6. Little weaker than the center.


100 mm, top right corner (400 px x 400 px), f/5.6. Little weaker than the center.


100 mm, bottom right corner (400 px x 400 px), f/5.6. Little weaker than the center.


100 mm, bottom left corner (400 px x 400 px), f/5.6. Not bad!

BR,

Martin

--
https://500px.com/bachrocks

JDsnaps said:

300m away? Are you sure there was no wind at all?

If you can, when the lens returns, test it on something more stationary, like a distant roof, billboard, chimney, statue, even a light pole would be more steady/solid than some small branches on a tree-top. Also due to the contrast, I'm seeing quite a bit of CA which probably makes it look worse than it is.

Click to expand...

JD you're right.

There was no wind but of course you never know. I am fully aware of the fact that this is not a scientific test. It is a simple and quick test. I wanted something filigree. On the other hand, I did a lot of repetitions, and the result was consistent in all cases. Influence of wind would definitely have averaged out.

I believe that the AF accuracy is the main source of error in such a test. But again, you're right. Maybe I can find a better subject for such tests.

Actually, I've started to use a test chart. But this really requires a lot of time, main reason being that the alignment is not so easy, particularly at greater focal length (with wide angle, you can immediately recognize even the slightest misalignment but not so with greater FLs). And, also with a test chart, really exact focusing is a potential cause of error, at least according to my experience. That's why I always do multiple series.

Actually, number one, you cannot create lab conditions at home, and number two, you can only test a system (camera + lens).

BR,

Martin

--
https://500px.com/bachrocks

I went through 3 samples of the 55-200 and 3 samples of the 10-24 before I found well centered lens. Remember decentering is not just L or R, it can be top or bottom as well.

Dr88man said:

I went through 3 samples of the 55-200 and 3 samples of the 10-24 before I found well centered lens. Remember decentering is not just L or R, it can be top or bottom as well.

Click to expand...

How did you test?
Assume you went back to the shop and requested a new copy each time?
Or did you test in the shop before purchase?

Rightsaidfred said:

Dr88man said:

I went through 3 samples of the 55-200 and 3 samples of the 10-24 before I found well centered lens. Remember decentering is not just L or R, it can be top or bottom as well.

Click to expand...

How did you test?
Assume you went back to the shop and requested a new copy each time?
Or did you test in the shop before purchase?

Click to expand...

I took photos of brick walls from 10ft to 30ft. Blades of grass from 10ft high. Trees from 200ft away. Center sharpness was good, it was always be the sides, top, or bottom that would be decentered in each case. This has been the norm for Canon and Nikon DSLRs too.

Dr88man said:

Rightsaidfred said:

Dr88man said:

I went through 3 samples of the 55-200 and 3 samples of the 10-24 before I found well centered lens. Remember decentering is not just L or R, it can be top or bottom as well.

Click to expand...

How did you test?
Assume you went back to the shop and requested a new copy each time?
Or did you test in the shop before purchase?

Click to expand...

I took photos of brick walls from 10ft to 30ft. Blades of grass from 10ft high. Trees from 200ft away. Center sharpness was good, it was always be the sides, top, or bottom that would be decentered in each case. This has been the norm for Canon and Nikon DSLRs too.

Click to expand...

This confirms my experience.

So, you always requested new copies, you did not go for repair? Is there a reason?

Rightsaidfred said:

Dr88man said:

Rightsaidfred said:

Dr88man said:

I went through 3 samples of the 55-200 and 3 samples of the 10-24 before I found well centered lens. Remember decentering is not just L or R, it can be top or bottom as well.

Click to expand...

How did you test?
Assume you went back to the shop and requested a new copy each time?
Or did you test in the shop before purchase?

Click to expand...

I took photos of brick walls from 10ft to 30ft. Blades of grass from 10ft high. Trees from 200ft away. Center sharpness was good, it was always be the sides, top, or bottom that would be decentered in each case. This has been the norm for Canon and Nikon DSLRs too.

Click to expand...

This confirms my experience.

So, you always requested new copies, you did not go for repair? Is there a reason?

Click to expand...

It's not repairable.

Here is a link to the famous Fred Miranda's easy test for de centered lens.

https://www.fredmiranda.com/forum/topic/1534737/

By the way, de centered lens repair is available at Fuji repair centers.

Foxjet said:

Here is a link to the famous Fred Miranda's easy test for de centered lens.

https://www.fredmiranda.com/forum/topic/1534737/

By the way, de centered lens repair is available at Fuji repair centers.

Click to expand...

So, Foxjet, this link was really helpful. Many thanks!

Actually, what Fred describes in his thread is nothing else than what I've started to call the 'tree test'. The tree doesn't have to be a tree of course.

But number one, Fred's procedure is interesting (turning the camera upside down for 2 f the 4 corners), number two, a great deal of the discussion in that thread is interesting, too, and the best, I found a cool browser tool on page 5 of that thread written by someone called keepcoding that greatly facilitates the afterwork. Thanks keepcoding!

I've added this test to the collection of 'SOPs' how to self test lenses after purchase.

Thanks again.

Will also report more on my assessment of my repaired 55-200 mm. Just give me a little time.

Btw, are there other opinions on the principal reparability of decentration?

BR,

Martin

57even said:

Rightsaidfred said:

Dr88man said:

Rightsaidfred said:

Dr88man said:

I went through 3 samples of the 55-200 and 3 samples of the 10-24 before I found well centered lens. Remember decentering is not just L or R, it can be top or bottom as well.

Click to expand...

How did you test?
Assume you went back to the shop and requested a new copy each time?
Or did you test in the shop before purchase?

Click to expand...

I took photos of brick walls from 10ft to 30ft. Blades of grass from 10ft high. Trees from 200ft away. Center sharpness was good, it was always be the sides, top, or bottom that would be decentered in each case. This has been the norm for Canon and Nikon DSLRs too.

Click to expand...

This confirms my experience.

So, you always requested new copies, you did not go for repair? Is there a reason?

Click to expand...

It's not repairable.

Click to expand...

It absolutely is repairable. There are actually shims inside every high quality lens that are used to adjust the lens to specification before it goes out the factory door.

BTW I like using what I call the "Polo mint" test for decentering, although Roger has a much more techy version here:
https://www.lensrentals.com/blog/20...ntered-lens-an-old-technique-gets-a-makeover/

Or a full test suite here:
https://www.lensrentals.com/blog/2010/11/how-to-test-a-lens/

Iuvenis said:

57even said:

Rightsaidfred said:

Dr88man said:

Rightsaidfred said:

Dr88man said:

I went through 3 samples of the 55-200 and 3 samples of the 10-24 before I found well centered lens. Remember decentering is not just L or R, it can be top or bottom as well.

Click to expand...

How did you test?
Assume you went back to the shop and requested a new copy each time?
Or did you test in the shop before purchase?

Click to expand...

I took photos of brick walls from 10ft to 30ft. Blades of grass from 10ft high. Trees from 200ft away. Center sharpness was good, it was always be the sides, top, or bottom that would be decentered in each case. This has been the norm for Canon and Nikon DSLRs too.

Click to expand...

This confirms my experience.

So, you always requested new copies, you did not go for repair? Is there a reason?

Click to expand...

It's not repairable.

Click to expand...

It absolutely is repairable. There are actually shims inside every high quality lens that are used to adjust the lens to specification before it goes out the factory door.

Click to expand...

Some of the compound lens groups are bonded sealed units and cannot be adjusted. Only the whole group can be adjusted, but this should be done at the factory.

It only takes a very small assembly misalignment or simply one element not to spec in the group and you have a non-adjustable decentering issue.

Not all consumer lenses are usually adjustable anyway, and some are impossible to tear down and rebuild beyond a point. Don't know about the 55-200.

In other words, it's much better to check the lens before or immediately after purchase and exchange it if there is a problem. Nikon never managed to repair any of my decentered lenses successfully.

My...my...my...we have ourselves another Pixel Peeper! Don't be silly...get out there and take some photos!

Dr_Jon said:

BTW I like using what I call the "Polo mint" test for decentering, although Roger has a much more techy version here:
https://www.lensrentals.com/blog/20...ntered-lens-an-old-technique-gets-a-makeover/

Or a full test suite here:
https://www.lensrentals.com/blog/2010/11/how-to-test-a-lens/

Click to expand...

Thanks for this.

We've discussed such tests in the parent thread on 'SOPs' how to test lenses. tokumeino made some contributions there.

In my opinion, the Zeiss Siemens start test is not very sensitive.

The 'Fred Miranda test' or 'tree test' or whatever you call it is really super quick and it is probably the best test when it comes to decentration. I have found no better test so far. It is so quick because you don't need no care for camera/subject alignment and interpretation is super fast with keepcoding's browser tool.

Next to this, I use test charts. But this really takes time: time for camera vs chart aligmnent and also a lot of time for interpretation.

BR,

Martin

pseudobreccia said:

My...my...my...we have ourselves another Pixel Peeper! Don't be silly...get out there and take some photos!

Click to expand...

Again:

This is not about pixel peeping.

This is about testing lenses within the return period. You're better off to do so. Like an incoming goods inspection you also do as a standard procedure in any company. Is my purchase within spec or not? There are so many examples here in this forum!

Look at the parent thread where I collect 'SOPs' how to best test lenses. It is my goal to collect quick and meaningful test procedures. Once you're done, you will either return your purchase or happily go out and take pictures. Big question: what is acceptable, what not? See here.

Cheers,

Martin

Rightsaidfred said:

Dr_Jon said:

BTW I like using what I call the "Polo mint" test for decentering, although Roger has a much more techy version here:
https://www.lensrentals.com/blog/20...ntered-lens-an-old-technique-gets-a-makeover/

Or a full test suite here:
https://www.lensrentals.com/blog/2010/11/how-to-test-a-lens/

Click to expand...

Thanks for this.

We've discussed such tests in the parent thread on 'SOPs' how to test lenses. tokumeino made some contributions there.

In my opinion, the Zeiss Siemens start test is not very sensitive.

The 'Fred Miranda test' or 'tree test' or whatever you call it is really super quick and it is probably the best test when it comes to decentration. I have found no better test so far. It is so quick because you don't need no care for camera/subject alignment and interpretation is super fast with keepcoding's browser tool.

Next to this, I use test charts. But this really takes time: time for camera vs chart aligmnent and also a lot of time for interpretation.

BR,

Martin

Click to expand...

No problem, although photographing a Polo mint on both sides of in-focus I find pretty instant, provided I haven't eaten them all already
(White Ring Binder hole reinforcers are less appetising...)