Inconsistent results between TIFF and EXR output in image positioning

[Matija Kogoj]

Hi.

TIFF looks normal, but when I opened the EXR in Luminance it showed the gaping black hole in the side of the road. I have also marked road edges - it seems as though the road was mapped differently in EXR.

There are also problems with stack alignment. In the "Align_image_stack can't add control points" I posted the following:

{

1. I have a different problem with aligning image stacks - it doesn't seem to do anything. I expected to see more control points, but all pictures showed 0.

Running basic CP-find found points among pictures normally.

2. In the masks tab I set a significant crop on the first picture and chose the option to apply the same crop to all images - does that crop have any effect on where Hugin looks for control points? It still found them outside of the crop marks - is there any way to limit the search area other than to digitally crop the images before putting them in hugin?

}

In case it is a different issue.

[T. Modes]

Am Montag, 23. Juli 2018 15:26:14 UTC+2 schrieb Matija Kogoj:

Hi.

TIFF looks normal, but when I opened the EXR in Luminance it showed the gaping black hole in the side of the road. I have also marked road edges - it seems as though the road was mapped differently in EXR.

 

Which Hugin version? (Also the operating system, from the screenshot I assume Windows?)

Which blender was used: enblend or the internal one.

You could try to change the blender (on the stitch tab). For enblend you could also add the parameter --primary-seam-generator=nft to see if this helps.

There are also problems with stack alignment. In the "Align_image_stack can't add control points" I posted the following:

First, the message was deleted. Second don't hijack other threads.

1. I have a different problem with aligning image stacks - it doesn't seem to do anything. I expected to see more control points, but all pictures showed 0.

Running basic CP-find found points among pictures normally.

align_image_stack works only with nearly fully overlapping images (with only a low amount of movement between the images).

So without seeing the images it is difficult to give more advice. (But don't paste the image in the message, instead provide a link (website, file hoster, ..)).

2. In the masks tab I set a significant crop on the first picture and chose the option to apply the same crop to all images - does that crop have any effect on where Hugin looks for control points? It still found them outside of the crop marks - is there any way to limit the search area other than to digitally crop the images before putting them in hugin?

Only cpfind can handle the crop and masks. Align_image_stacks ignores the masks/crop. But you can afterwards call remove cp from mask from the menu.

[Matija Kogoj]

I was under impression that it was better not to make new topics when a significantly similar thread was already open.

Operating System: Windows 7 SP 1, 64-bit edition
Architecture: 64 bit
Free memory: 13771976 kiB

Hugin
Version: 2018.0.0.5abfb4de7961 built by Thomas

Libraries
wxWidgets: wxWidgets 3.1.1
wxWidgets Library (wxMSW port)
Version 3.1.1 (Unicode: wchar_t, debug level: 1),
compiled at Nov 18 2017 10:28:15

Runtime version of toolkit used is 6.1.

libpano13: 2.9.19
Exiv2: 0.26
SQLite3: 3.18.0
Vigra: 1.11.0
LittleCMS2: 2.9

I only used tools which come within hugin and as default, having made no changes to settings.

The following images are about the worst I could find. There is one situation where the lens shifted, but I reason this should still result in some stacks finding at least some similarities, not a flat 0.

http://pasteall.org/pic/8383ef2de64b6647f3c35beddd01abdb

http://pasteall.org/pic/31331e0337f932013ca0bbb91868683d

Regardless, it doesn't change the fact that the TIFF file was correct. Though I should note I only wanted the .ext and don't remember asking for any other output files, so it is possible the TIFF was a leftover, temporary file of the only layer that had control points in the CP tab.

[T. Modes]

Am Dienstag, 24. Juli 2018 11:41:06 UTC+2 schrieb Matija Kogoj:

I only used tools which come within hugin and as default, having made no changes to settings.

The following images are about the worst I could find. There is one situation where the lens shifted, but I reason this should still result in some stacks finding at least some similarities, not a flat 0.

This is not reproducible. Align_image_stacks finds here 158 cp for the given example. Some of them are on the lens hood, but otherwise it find cp.

(Not so bad for this example: mostly featureless sky, then a big lens hood shadowing the borders. Then overexposed areas and lens flares. There are only minor areas which can be used for alignment - and these are mainly bushes and tree with possible foliage. This is not a good image for automatic processing, not sure what you are expecting.)

[Matija Kogoj]

If alignment is not the problem I can't imagine what would cause the resulting TIFF and EXR files to be so different.