wwp@yahoogroups.com:
Re: fisheye lenses : Nikkor 10.5 mm vs Sigma 15
pedro_silva58 2005-Apr-01 09:19:00
--- In #removed#, "Erik Krause" <#removed#> wrote:
> On 31 Mar 2005 at 21:00, pedro_silva58 wrote:
> > i find the sigma 8mm not too bad (considering) near the center,
but
> > much worse in the periphery. cropping a 15mm might give a nice
> > quality image all around (and require more shots: there's no free
lunch!)
> > also, is a "rectilinear" 14mm or zoom necessarily better for panos
> > than a 15mm fish? why?
>
> Rectilinear images are compressed in the outer regions (resulting
in
> more pixels and higher sharpness) whereas fisheye images are
> stretched. See Helmut Derschs article:
> http://www.einem.net/~dersch/heliarVR/heliarVR0.html
erik, thanx for helping me understand this. i'm getting there, i
hope... one thing still eludes me. helmut writes "In effect,
rectilinear lenses exhibit an almost constant angular resolution while
fisheye lenses degrade in the corners." so, i don't understand why
cropping the corners of a fish gives a worse result than cropping the
corners of a rectilinear. i'm not being argumentative, just don't
get it.
feel free to move to email, if this is geting too ot for this list.
> > if its distortion is moustache-like (quite
> > common, i hear), wouldn't that be harder to correct than a good
fish?
> No. As long as the distortion curve can be modelled by a fourth
> degree polinomial (this is what panotools does) there is not
problem
> to correct it. This wavy type of distortion is the reason why we
have
> three lens correction parameters and not only one or two...
> Erik Krause
thanx, i understand this.
cheers,
pedro