Yes, now I see... I've used a FITS directly extracted from a OSC camera, that's a mono one with the bayer pattern on it. The use case is first, debayering this image, and the pass the RGB image to stellarsolve. Understood. Now I'm able to see the differences clearly.