AWARE SYSTEMS
TIFF and LibTiff Mail List Archive

Thread

2002.05.09 14:30 "Orientation and Tiled images?", by Michael O'Rourke
2002.05.09 15:24 "Re: Orientation and Tiled images?", by Jon Saxton
2002.05.09 16:10 "Re: Orientation and Tiled images?", by
2002.05.09 22:04 "Re: Orientation and Tiled images?", by Michael O'Rourke
2002.05.10 10:10 "Re: Orientation and Tiled images?", by Andrey Kiselev
2002.05.10 13:01 "Re: Orientation and Tiled images?", by
2002.05.10 13:29 "Re: Orientation and Tiled images?", by Michael O'Rourke
2002.05.10 15:21 "Re: Orientation and Tiled images?", by Andrey Kiselev
2002.05.13 19:12 "Re: Orientation and Tiled images?", by Antonio E. Scuri

2002.05.10 13:01 "Re: Orientation and Tiled images?", by

It is important to finally determine order of reading lines in bottom-left orientation. I think that the first line of the first tile of the bottom-left oriented image should be last line in the rendered image. TIFFGetRGBAImage() function in the latest libtiff works in such way (and above images rendered correctly). Is it right?

Yes, this sounds correct. And it is in accordance with the following:

What *really* matters is: THE FILE MUST RENDER CORRECTLY EVEN WHEN IGNORING THE ORIENTATION TAG. Of course, the image orientation (flip/mirror) will be wrong in this case, but that's all that may be wrong. If an application chooses not to process the tag, the it should still be able to load the image and let the user "manually" correct the orientation error simply by flipping and/or mirroring the image with any basic image manipulation software.

For instance, take the file "cramps-tile.tif". Use a tiff "tag edit" tool and change its orientation tag to 3 and the file should render upside down in a viewer that supports the orientation tag.

Best regards,
Peter Nielsen

http://www.pmview.com