Open Alex-Vasile opened 1 year ago
Thanks so much for this work! Would we expect the test image(s) to change as a result of this addition? If so, can you please also make a PR against x3f_test_images with the new image(s) converted with this fix? That way, we can make sure make check
continues to pass.
(as a side note, this is obviously an argument for a monorepo reorganization, but as you can see, we haven't really been active in this project for 6 years)
Thanks so much for this work! Would we expect the test image(s) to change as a result of this addition? If so, can you please also make a PR against x3f_test_images with the new image(s) converted with this fix? That way, we can make sure
make check
continues to pass.
I have tested visually:
(as a side note, this is obviously an argument for a monorepo reorganization, but as you can see, we haven't really been active in this project for 6 years)
Is this something you are still interested in supporting? I think there is a way to cancel out the green circular cast. Hasselblad has a way of dealing with it for specific lenses by taking an image of just a white piece of plastic in front of a lens, creating that as a "sample" to treat the color-cast. I can link the video if you are interested. I was wondering if we can add this feature to kalpanika. It will certainly make it more useable, since the color cast on merrills ruins it. @mmroden
The right "shielded" area is not really shielded and can't be used for calculating the black level. Removing this area completely fixes the green tint in the shadows. Closes #117.
Below is a screenshot of the shielded area from the _SDI8040.X3F from x3f_test_files while is a DP2M file. I'm showing the green channel as shown by RawDigger, the red and blue channel are similar (except the red is much noisier and the blue much less noisy).
(As an aside, that region is very conspicuous with its linear increasing exposure may be used for something).