Re: mild vignetting = blurry corner(s)?
- From: David Littlewood <david@xxxxxxxxxxxxxxxxxx>
- Date: Wed, 5 Oct 2005 10:31:32 +0100
In article <llm6k15sb27klnscf3v9gbsge6pkl9sena@xxxxxxx>, Alturas <username@xxxxxxx> writes
On Mon, 3 Oct 2005 21:17:01 +0100, David Littlewood <david@xxxxxxxxxxxxxxxxxx> wrote:
I don't know the camera, so apologies if this is necessarily general.
I think it is unlikely that vignetting would cause loss of sharpness. Indeed, if there is no darkening of the image, there is by definition no vignetting at all. Even where there is vignetting it should not cause loss of sharpness.
I tend to agree but I was thinking of a boundary effect like when you hold your fingers very close to your eye and look through the blur at their edges. In that case there is a fringe where things are blurry but not noticeably darker, but I can't mimic that in the camera display. I am also thinking this is something that happens the moment the picture is snapped, like the OIS lens element being canted in motion. If only the display could be zoomed to full 5MP detail before the shot was taken, I could rule that out.
Interesting. I just tried it with my computer screen. I put myself too close to focus (regrettably easy these days!) and then put my finger into the centre of the field of vision. There was indeed a fringe visible around the finger, but its apparent effect was to increase sharpness, not decrease it - I could clearly read words in the fringe area whereas I could not read the rest. So I am inclined to think your idea is not the answer, unless someone can come up with a more scientific study than mine!
I would guess at three possible causes. Slight decentring of one of the lens elements could cause loss of resolution in one extremity. Slight misalignment or lack of flatness of the sensor (as you say) would do so also. Thirdly, a small defect or smear on the periphery of the lens might cause it. All three would probably become less noticeable of stopping down.
To test this may be difficult. In a fixed lens camera, you obviously cannot change the lens to see if the problem disappears, nor can you easily examine the sensor. You can of course examine the outer surface of the lens, which is the one most likely to become smeared or scratched. A smear should be easy to clean; a scratch is more of a problem.
The glass looks pristine in all elements that I see. Everything seems lined up well; at least no obvious off-center elements. Running the camera in aperture-priority mode and limiting it to F5.2 (when practical) usually sharpens the affected corner.
OK, scratch that one then.
Give it a try and let us know if it works, would be interesting. Of course, you may find one of the upper (as used normally) corners is also affected.
One off-the-wall suggestion: you may be able to use the camera upside down. Since you have the problem mostly with landscapes, and landscapes usually contain sky at the top, and since detail in sky is not usually an issue, the results might be fine. Don't know how practical this would be though.
I never though of that but it wouldn't be easy to pull off and I'd have to ignore some stares! Lossless JPEG rotation would indeed make it feasible. If anyone has owned any camera (including film) with just one corner blurry, please chime in.
David -- David Littlewood .
- References:
- Re: mild vignetting = blurry corner(s)?
- From: David Littlewood
- Re: mild vignetting = blurry corner(s)?
- Prev by Date: Dutch riverscapes
- Next by Date: Re: Dogs ate the New Orleands Dead (photos)
- Previous by thread: Re: mild vignetting = blurry corner(s)?
- Next by thread: Re: mild vignetting = blurry corner(s)?
- Index(es):
Relevant Pages
|
