Whenever I take a picture with my dslr, there is a strange line that shows up on the left of the picture all the time. Can this be fixed my myself, or do I have to send this one in? example of problem


1 Answer 1


CMOS sensors are read a line at a time. This appears like it may be damage to the sensor causing that line to not be read (though I thought they were normally horizontal rather than vertical). My guess is that this will probably require servicing and possibly a new sensor to fix. There may also be a way they can treat them all as dead pixels and have it still work, but I'm not sure if they have that level of calibration or not.

Other things you might be able to try is doing dust delete for it and/or seeing if there is a dead pixel remap function available for your camera. While the dust delete will simply subtract the line from the image, the dead pixel map will alter the demosaicing as described in Michael Clark's comment so that some of the information can be recovered. That I don't know off the top of my head, though personally, with that many pixels being a problem, I'd personally send it in.

  • \$\begingroup\$ A dead scan line was my thought too, but it's strange that the line is blurry in the out-of-focus parts of the frame. \$\endgroup\$
    – Caleb
    Commented Sep 26, 2013 at 18:51
  • \$\begingroup\$ @Caleb - yeah, it's hard to tell with the resolution and color of the sample. I think there may just be a greyish area in the image there. I think I might still be able to see the sharp line passing through that area, it's just really hard to pick out from the actual image I think. Also, JPEG compression could destroy the line in some blocks if the colors are close enough. \$\endgroup\$
    – AJ Henderson
    Commented Sep 26, 2013 at 18:53
  • 1
    \$\begingroup\$ If the dead scan line is only one pixel wide, then demosaicing could interpolate some values to the dead pixels that are in the proximity of other high intensity pixels under the same color filter in the Bayer mask. With a conventional Bayer pattern, the line would either be alternating G and B or G and R. For areas that have high intensities of both G & R such as skin tones, the interpolation would be stronger than in areas with less intensity of the two colors that make up the line of dead pixels. \$\endgroup\$
    – Michael C
    Commented Sep 27, 2013 at 6:41
  • \$\begingroup\$ @MichaelClark - thanks for the extra details on the demosaicing. That's what I was alluding to with the dead pixel remap, but I wasn't confident enough to attempt a technical explanation. I'll make note of your comment in my answer. \$\endgroup\$
    – AJ Henderson
    Commented Sep 27, 2013 at 13:45
  • \$\begingroup\$ I guess that the picture is rotated, so it is horizontal... \$\endgroup\$
    – fortran
    Commented Sep 27, 2013 at 15:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.