hugin-devs team mailing list archive

[Ulrich-windl <1633352@xxxxxxxxxxxxxxxxxx>]

Public bug reported:

I have a panorama with 14 TIFF images with 16 bit per color (in summary about 200 megapixel). The images fit rather nicely and there is no warning before conversion.
But when converting these warnings appear:
enblend: info: loading next image: 2749-27620005.tif 1/1
enblend: warning: unable to run Dijkstra optimizer
enblend: note: seam-line end point outside of cost-image
enblend: note: contour #2 of 2, segment #1 of 1, vertex #2716 of 2717
enblend: info: loading next image: 2749-27620006.tif 1/1

It's hard to find out from these texts where the actual problem is and
how to fix it. I have a slight suspect that the 32-bit version might
have a memory problem with such a huge (28152x3284) panorama. Up to
recently I was running the 64-bit version of the 2015 edition, and I
never saw such problems before, but I use 16-bit TIFF as source rather
rarely. Anyway:

In the resulting panorama you can clearly see two artefacts; one at the to of the panorama, the other one at the bottom. I'm attaching the relevant part of the preview to relate the position of the artefact to the images. I extracted a part of the resulting panorama, scaled it down, and I'm attaching it as part of the data for this bug. Unfortunately the material is too huge for an attaching.
I tried to make a smaller test case using JPEG images, but the the problem did not exist.
When trying 16bit TIFFs that were downscaled, the problem even occurred twice (using a different set of control points).

** Affects: hugin
     Importance: Undecided
         Status: New


** Tags: enblend

** Attachment added: "ZIP archive with project file, artefact image, log file, preview"
   https://bugs.launchpad.net/bugs/1633352/+attachment/4760776/+files/2749-2762-bug.zip

-- 
You received this bug notification because you are a member of Hugin
Developers, which is subscribed to Hugin.
https://bugs.launchpad.net/bugs/1633352

Title:
  "enblend: warning: unable to run Dijkstra optimizer" for large
  panorama

Status in Hugin:
  New

Bug description:
  I have a panorama with 14 TIFF images with 16 bit per color (in summary about 200 megapixel). The images fit rather nicely and there is no warning before conversion.
  But when converting these warnings appear:
  enblend: info: loading next image: 2749-27620005.tif 1/1
  enblend: warning: unable to run Dijkstra optimizer
  enblend: note: seam-line end point outside of cost-image
  enblend: note: contour #2 of 2, segment #1 of 1, vertex #2716 of 2717
  enblend: info: loading next image: 2749-27620006.tif 1/1

  It's hard to find out from these texts where the actual problem is and
  how to fix it. I have a slight suspect that the 32-bit version might
  have a memory problem with such a huge (28152x3284) panorama. Up to
  recently I was running the 64-bit version of the 2015 edition, and I
  never saw such problems before, but I use 16-bit TIFF as source rather
  rarely. Anyway:

  In the resulting panorama you can clearly see two artefacts; one at the to of the panorama, the other one at the bottom. I'm attaching the relevant part of the preview to relate the position of the artefact to the images. I extracted a part of the resulting panorama, scaled it down, and I'm attaching it as part of the data for this bug. Unfortunately the material is too huge for an attaching.
  I tried to make a smaller test case using JPEG images, but the the problem did not exist.
  When trying 16bit TIFFs that were downscaled, the problem even occurred twice (using a different set of control points).

To manage notifications about this bug go to:
https://bugs.launchpad.net/hugin/+bug/1633352/+subscriptions