leifs wrote:

I have done the handiwork, moved some pictures and deleted some pictures with no links from the panorama. Then I have rendered it with "no antighost".
The resulting panorama is 19 gigapixel and the PSB-file is 103 GB.
So what is the result ? Better than expected, but still room for refinement !

As shown in the picture:
- 400mm give best details (surprise?)
- the APG blended result is somewhere between 100 and 400

What I would like to see:
in overlapping area: no blending! use the best image (=400mm)
per now I can do it in Photoshop, but I would like APG to do it, auto :-)

Leif

This brings up an interesting point--if you have one image that is sharper but darker than another, should APP prefer sharper image or one that is better exposed?  Or should APP take the contrast from the sharper image and the luminance/color from the better exposed one?

I have uploaded the images.
3 files (sorry about that): readme.txt , 400mm.zip 6.8GB, 100mm.zip 1.6GB

there are some issues with this set:
how to make a group ?
- simple group adding all pictures does not work
- gigapan import for all pictures does not work
- gigapan import for 100mm and thereafter gigapan import for 400mm into the same group is not possible
- gigapan import for 400mm adding 100mm's crash
- gigapan import for 100mm adding 400mm's worked (I am not able to repeat it !)

In the "merged" result there is a pattern outside the 400mm area, that makes the picture unusable (its not in the pure 100mm result)
It is shown in the pictures below

On my wishlist is:
some way to make a group of two (or more) gigapan sets of pictures at different focal lengths and
APG "merge" them using only the longest focal length in each area
(why not use the 400mm for the whole picture? : because of the problems stitching water and sky)

I have now published the 400mm pano (3.2 gigapixel) at http://www.rundskuer.no/krpano/voldafra … rnet2.html

Leif

Last edited by leifs (2011-05-12 21:05:37)

Here are the results. In the fisrt image, cutting priority is set to anti-ghost, in the second one, the cutting priority is set to 'long focal'.

Thomas

ThomasV wrote:

ThomasV wrote:

I will try to post 2 screens illustrating the difference on your datas.

On the first image, the cutting priority is set to 'anti-ghost", and on the second image, the priority is given to long focal.

Very fine! I realize again: there are options in APG for which to explore we need years - because they are hardly understandable documeted

Btw.: the second image looks definitely sharper . . .

best, Klaus

Last edited by klausesser (2012-08-30 15:41:09)

gkaefer wrote:

ThomasV wrote:

ThomasV wrote:

I will try to post 2 screens illustrating the difference on your datas.

On the first image, the cutting priority is set to 'anti-ghost", and on the second image, the priority is given to long focal.

... Klaus... If you read the lines from Thomas in detail ;-)

Here are the results. In the fisrt image, cutting priority is set to anti-ghost, in the second one, the cutting priority is set to 'long focal'.

if I look at image 1 (where the ship ghost is visible) and read Thomas explanation this means:

setting cutting priority to anti-ghost = the ghost remains in pano.
setting cutting priority to long focal = the ghost is removed (and by accident the resulting pano gets sharper).

and if you repeat this test by random multiplied with pi you may get a perfect HDR night view of the scene.



Georg

Hi,

In the first image, it is not a ghost that you see !
It is just that the 400mm pictures were taken after the ship has gone... So, setting priority to long focal makes the 100mm pictures to be overwritten by 400mm ones, and therefore the ship disappears.

This part of the panorama was choosen because the difference is obvious around the boat.

Thanks,
Thomas