First post: Nikkor 10.5 fisheye

[OldRoy]

Hi to everyone on LL.
I have been looking at the site for a while but only just registered on the forums. I'm probably the sort of contributor who has far more questions than answers, but with time that could change!

I am more or less retired and have become very interested in panoramic photography over the last year. I will never be a really good "pure" photographer, so the combination with post-production/IT is really interesting to me. I have been trying to teach myself spherical panoramas using an NN3/D200/Nikkor 10.5 (and various bits of software of course, notably PTGUI).

The question I have is about this lens. I'm well aware of the inherent problems associated with lenses of this type but have nothing to compare it with, personally. My current workflow (to dignify my bumblings) involves NEF to TIFF via CR, with correction of ca and vignetting but no other processing.

Fot the ca I'm using +/- 30 (approx) but for the vignetting I'm a bit puzzled. I can't see any way of getting sufficiently even illumination to measure the vignetting, and thereby devise a correction factor, so I'm tweaking by eye. Sometimes this looks about right at  < +20 but sometimes it looks as though it needs to be off the scale.
To put the question briefly, what values seem appropriate to others who are using this lens?

And my goodness, doesn't it look horribly soft at the periphery!

[BernardLanguillier]

Fot the ca I'm using +/- 30 (approx) but for the vignetting I'm a bit puzzled. I can't see any way of getting sufficiently even illumination to measure the vignetting, and thereby devise a correction factor, so I'm tweaking by eye. Sometimes this looks about right at  < +20 but sometimes it looks as though it needs to be off the scale.
To put the question briefly, what values seem appropriate to others who are using this lens?
[a href=\"index.php?act=findpost&pid=178666\"][{POST_SNAPBACK}][/a]

Vignetting depends on aperture obviously, besides nothing says that it is linear between center and corners. You could try to shoot an evenly illuminated surface (a sunlit white wall for instance), and then try to find the best setting in CR, but chances are the correction will not be perfect. You'd have to do this anyway for each of the apertures that you would consider use for panorama work.

Have you considered DxO 4.5 to do these corrections? It is based on a local measurements of all variables and can therefore cope very well with non linear vignetting etc...

You'll probably want to avoid correcting the geometry of your fisheye, since PTgui should be able to do this pretty well while merging.

cheers,
Bernard

[OldRoy]

Vignetting depends on aperture obviously, besides nothing says that it is linear between center and corners. You could try to shoot an evenly illuminated surface (a sunlit white wall for instance), and then try to find the best setting in CR, but chances are the correction will not be perfect. You'd have to do this anyway for each of the apertures that you would consider use for panorama work.

Have you considered DxO 4.5 to do these corrections? It is based on a local measurements of all variables and can therefore cope very well with non linear vignetting etc...

You'll probably want to avoid correcting the geometry of your fisheye, since PTgui should be able to do this pretty well while merging.

cheers,
Bernard
[a href=\"index.php?act=findpost&pid=178726\"][{POST_SNAPBACK}][/a]

Thanks Bernard
I sometimes wonder whether a semi-transparent cap might be a partial solution to the illumination problem?