[ptx] Getting EXIF data from coolpix 950 jpgs using hugin

pablo.dangelo at informatik.uni-ulm.de pablo.dangelo at informatik.uni-ulm.de
Wed Dec 7 12:23:45 GMT 2005


Zitat von Thomas Steiner <e9826064 at stud4.tuwien.ac.at>:

> > I think it's easier to get it from the ratio between the focal length
> >  of the lens and the 35mm equivalent of it. Not many catalogues do
> > not  state the CCD size anyway.
>
> The crop factor of my Canon EOS 350D should be exactly 1.6, but hugin
> overestimates (some percents) it a bit (and missestimates the facal
> lenths then
> as well). Why?

The 350D exif data contains the sensor resolution (~size of one pixel) and I
just use this data to calculate the crop factor.

How did you arrive at 1.6?

The caculation goes like this:

from the exif data:
fxr = 3954.23/25.4;   % x resolution in pixel / mm
fyr = 3958.76/25.4;   % y resolution in pixel / mm
w=3456; h=2304; % image size

calculation of crop factor:

sw = w/fxr = 22.1996 mm    % sensor width (active pixels)
sh = h/fyr = 14.7828 mm    % sensor height (active pixels)

d = sqrt(sw.^2+sh.^2) = 26.6712 mm % sensor diagonal
d35mmfilm = sqrt(24.^2+36^2)
cropf = d35mmfilm /d = 1.6222  % crop factor

> What can i do against this?

Nothing.

The values in the exif will never be perfect and optimizing the v paramter will
take care of it.

I guess the focal length reported by the lens actually contain a higher error.

One really has to optimize the hfov (v parameter), because the exif data will
never be 100% right.

ciao
  Pablo

----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.


More information about the ptx mailing list