[Hugin-devs] [Bug 1402613] [NEW] TIFFReadDirectory: Incorrect count for BitsPerSample.

2014-12-15 Thread Peace-maker
Public bug reported: Can't render the below tiff file created by hugin using any program that uses the official libtiff: https://www.dropbox.com/s/4ieoef5nn1rfsaz/IMG_2322-2324.tif?dl=0 = $ tiffinfo broken.tif TIFFReadDirectory: Incorrect count for BitsPerSample. =

[Hugin-devs] [Bug 1402613] Re: TIFFReadDirectory: Incorrect count for BitsPerSample.

2014-12-15 Thread zarl
Opens without an error message in Photoshop CS6 (Mac) and GraphicConverter. Hugin however isn't able to decode that file. EXIF information of the image reveals that a software called Daminion (from daminion.net ) touched your image at some point. Have you seen that problem before using Hugin?

[Hugin-devs] [Bug 1402613] Re: TIFFReadDirectory: Incorrect count for BitsPerSample.

2014-12-15 Thread Peace-maker
Any software that uses libtiff fails when try to render this image. It seems that the Photoshop CS6 (Mac) and GraphicConverter use the proprietary tiff processing libraries. I've played with both Hugin and Daminion, but can't reproduced this issue with the latest program versions. I'm waiting for

[Hugin-devs] [Bug 1402613] Re: TIFFReadDirectory: Incorrect count for BitsPerSample.

2014-12-15 Thread zarl
I sometimes feed panoramas back into Hugin directly after stitching (e.g. to level a wavy horizon) but never had issues like the one described here. Can the person who has that problem be pointed to this discussion so the developers here can get more details? The image contains some information

[Hugin-devs] [Bug 1402613] Re: TIFFReadDirectory: Incorrect count for BitsPerSample.

2014-12-15 Thread tmodes
Hugin and enblend are using libtiff (via vigra) to write the images. And they write tiff files always with an alpha channel. So the shown image was not touched by Hugin at the end. Another software has removed the alpha channel without correctly updating the header. This is nothing which can be

[Hugin-devs] [Bug 1402347] Re: Segmentation autooptimiser

2014-12-15 Thread tmodes
There are several issues with this pto file: 1.) First it contains 1266 images, but the control points refers up to image 1670, which does not exists. 2.) Not all images are connected by control points. checkpto reports 410 image groups. 3.) Each image has it own lens: so there are 1266 lenses,