various mask bugs

[clepsdyrae]

I have 23 images in the list which I am aligning with hugin. I made a mask on the first image, and set it to "exclude region from all images of this lens".

I made CPs, and they were excluded from the mask (thanks for the recent fix).

Then I stitched (generating remapped images, no exposure correction, low DR), but I forgot the mask was there, and so the region was not present in the images -- this seems correct (not a bug).

So I changed the mask type to "include region" and ran the stitcher again.

The first image was normal: the whole image was generated. The other 22 images, however, lacked the region, as if they all had the same mask but it said "exclude". I'm not sure if this is a bug, but it surprised me: with just one mask, on one image, set to "include", it excluded that region from all the other images. Maybe it's normal -- just reporting it in case.

Second issue:

So I removed the mask entirely -- now there were no masks at all on any image. I ran the stitcher again, and the result is obviously a bug: many images with strange, random masks, as if the masking data is corrupted. A few examples are attached. I note that between running the stitches I was recalculating the geometric optimizations.

This is with hugin via the nightlies ppa, commit b55e2c48db83, on Kubuntu 16.10.

System info:

Operating System: Linux 4.8.0-46-generic x86_64
Architecture: 64 bit
Free memory: 2167732 kiB

Hugin
Version: 2016.3.0.b55e2c48db83
Path to resources: /usr/share/hugin/xrc/
Path to data: /usr/share/hugin/data/
Hugins camera and lens database: /home/casey/.hugindata/camlens.db
Multi-threading using C++11 std::thread and OpenMP

Libraries
wxWidgets: wxWidgets 3.0.2
wxWidgets Library (wxGTK port)
Version 3.0.2 (Unicode: wchar_t, debug level: 1),
Runtime version of toolkit used is 2.24.
Compile-time GTK+ version is 2.24.30.

libpano13: 2.9.19
Boost: 1.61.0
Exiv2: 0.25
SQLite3: 3.14.1
Vigra: 1.11.0
LittleCMS2: 2.7

[tmodes (tmodes)]

First issue: This is the way how include masks work in stitching. With an include mask you want to force to include the area from one image. To achieve this this area is removed from all other images.

Second issue: Just looking at the further processed images does not help. Please provide the pto file which shows the issue. But I assume it is not an issue with mask, it is probably related to the translations parameters. see FAQ http://wiki.panotools.org/Hugin_FAQ#Why_does_my_output_covers_only_180.C2.B0.3F

[clepsdyrae (myopenid)]

Ok - thanks, as usual, for the helpful reply. I will make sure to save a pto next time and try to make something replicable. I just assumed it was a transient error so it didn't seem like a pto would be helpful. I'll know better next time.