Unable to get focal length, object distance and minimum object distace (MOD) from Quattro files

[erikrk]

Quattro files lack the entire PROP section and the ObjectDistance and LensInformation entries in CAMF. This makes the functions get_focal_length, get_object_distance, and get_MOD in x3f_spatial_gain.c fail with a warning and return arbitrarily chosen default values.

An example:
http://www.dpreview.com/forums/post/56207560

I suppose that, in principle, this data should be possible to obtain from the EXIF data in the embedded JPEG, but should that really be necessary? Shouldn't it really be somewhere in X3F?

[rolkar]

You cannot find this data in CAMF.

There are some data in the beginning of the file that lacks
the formal definitions as in PROPS and CAMF. It is just data.
Some data is just known by the Sigma reader and some is in
a list of extra data with integers as keys.

It may be found somewhere there. If we had an SD Quatro camera
we could test it :) The focal lengths of the DP1, DP2 and DP3 are 19, 30
and 50 mm. Searching for floating point numbers that equals those values
might find the position for the focal length data.

Focussing and comparing the data might find us the object distance data.

Of course, reading the EXIF data in the JPEG is probably much easier, except
that some data is found in the MakerNote tag. Then I hope EXIF Tool migh
have cracked Sigma EXIF data. Hmmmm .... but, most tools gave up to be compatible when
Quattro arrived. So, maybe not.

[rolkar]

BTW - why is it important? Do we need this data?

[erikrk]

It is needed for spatial gain correction. There are several different corrections depending on aperture and focusing. We need to know this data to be able to select the right corrections and interpolate between them.

As a side note, the way to select and interpolate between corrections for Merrill and Quattro is a bit sketchy. There are comments about that in x3f_spatial_gain.c. It's all guesswork more or less.

[mmroden]

We also need it for later programs that use this information. I tried to
build my sunset panorama using x3f tools, but hugin uses that information
to determine the field of view for the lens and provide a more accurate
stitch. I ended up guessing on the numbers for field of view, etc, but it
didn't work so well.

The crop factor may not be a direct 1.5, for instance; it may be 1.49 or
1.51, and that can make all the difference in a good stitch.

On Sun, Jul 26, 2015 at 8:43 AM, Erik Karlsson [email protected]
wrote:

It is needed for spatial gain correction. There are several different
correction depending on aperture and focusing. We need to know this data to
be able to select the right correction and interpolate between them.

As a side note, the way to select and interpolate between corrections for
Merrill and Quattro is a bit sketchy. There are comments about that in
x3f_spatial_gain.c. It's all guesswork more or less.

—
Reply to this email directly or view it on GitHub
#59 (comment).