Vignette problem

[Paul Joy]

Originally Posted by Steven Thomas

Not that it should matter, have you tried turning off image stabilization?

Yes, nothing like that makes any difference. I feel quite sure that it's nothing to do with processing or image enhancement, it feels to me like there's some kind of obstacle inside the lens.

Paul.

[Steven Thomas]

Originally, I thought the same thing. It's odd that its window is 10-25mm.

[Alex Leith]

Originally Posted by Steven Thomas

???
It depends what your thinking reasonable?
With the stock lens and no attachment, I'd say losing the top corner of your picture is unreasonable.

I would agree totally. Even on a $500 camera I would not expect to see any weird vignetting of the nature seen here.

[Paul Joy]

I think it's pretty safe to say that this is a one off problem now, with more camera's appearing there isn't any other mention of it which is great news.

The only downside is that the camera's going back tomorrow and I'm now back on the list with most of you waiting for another one.

You won't be disappointed though.

After 2 days of playing with the camera and posting here, I'd better get my head down and get some proper work done!

Paul.

[Dave Elston]

Does anyone else see a very brief vignette in one of the original tennis clips?

In the first second or so (00:01:05) of Clip "X02_0036_01.MP4" as the camera zooms back out and whip-pans across the net, there is briefly a shadow/vignette popping up in the top-left (I think), although not nearly as severe as Paul's example!

It seems to be triggered by the fast camera movement causing the OIS to steer left momentarily - perhaps this is indeed where Paul's unit is faulty (OIS stuck in a far-left position?).

It would make sense that a 'bug' like this could very well be the reason for the slow release of shipments (QC filtering out the duds). It is also something I think everyone receiving the early shipments would be wise to test for straight out of the box - long before taking the camera out on any critical jobs.

[Piotr Wozniacki]

You mean this?

Difficult to say whether it's the "vignetting" artifact, or just one of those ribs under the hall roof... Judging from their spacing, though, it shouldn't be there! On the other hand, considering fast pan - due to the rolling shutter everything is possible. Paul's were static examples.

Edit I posted two grabs on purpose; even if it may look like a part of the roof structure in the left (earlier) one, it shouldn't be there in the right (later) one! Seems like it's moving to the right with camera.

[Dave Elston]

That's exactly it Piotr, thanks for posting the grabs (I'm stuck on a work machine that can't frame-by-frame easily).

I don't believe it to be part of the background structure (see earlier frames), this example is only a slight intrusion compared to the problem Paul has experienced. Just seems to indicate an OIS issue is to blame as has previously been suggested.

[Paul Joy]

On the first image I would have said it's just the building, but in the second there is a dark area where the same area was not dark in the first image.

[Edit] you know, the more i look at this the more it looks very similar![/ edit]

Only the camera owner could answer for sure though.

Paul.

[Steven Thomas]

IMO, that does not look like vignetting. With all the quick pan motion blur, this is probably part of the background.
Also, I asked the camera owner to do a test a day ago. He said he would.
He has not mentioned any vignetting issues.

[Piotr Wozniacki]

Originally Posted by Paul Joy

On the first image I would have said it's just the building, but in the second there is a dark area where the same area was not dark in the first image.

[Edit] you know, the more i look at this the more it looks very similar![/ edit]

Only the camera owner could answer for sure though.

Paul.

Exactly Paul; this is why I posted both grabs!

Well, this would support my little theory earlier in this thread... Some design feature might cause all the cameras to be on the verge of vignetting (or whatever you call this artifact); most do not so the QC never noticed. But some do, and now Sony is inspecting and returning those faulty ones - hence the delay and shortages...

But I will ask again: did anybody try to shoot without the lens hood?

[Steven Thomas]

It possible, but at least looking at these two images which of course are not one for one. One is wider and at a slightly different angle. IMO, you can not tell anything from these images based on the FOV is different.

[Dave Elston]

Perhaps the decision to go with 1/2" sensors in a 1/3" form factor has forced all the optical elements/devices into a very narrow margin of the critical optical path - PLUS this is a first in lens design with the integration of "Full-Auto-Full-Manual" controls.

Hopefully just teething trouble that Sony will iron out completely in future production runs.

[Paul Joy]

I see what you're saying Steven, the second shot is wider and that could cause the roofline object to be staying in view. If you look at the two light objects leaving a trail each side of the net line at the back of the room though, the second image is definitely panned further to the right than the first one. I don't have the source file, does it show the shadow moving with the camera?

[Sami Sanpakkila]

Just a thought. Could hardly make a difference but does something happen inside the lens when you change from autofocus to true manual focus ie. pull the focus ring back?

Those who have it have you tried to see if the vignetting appears both on manual and auto focus setting?

Sami

[Paul Joy]

Originally Posted by Dave Elston

Perhaps the decision to go with 1/2" sensors in a 1/3" form factor has forced all the optical elements/devices into a very narrow margin of the critical optical path

Lets hope the fix isn't to install a set of 1/3" sensors :)