I shot a lot of photos during my summer vacation. Traveling during daytime, with little time to edit or review photos I often copied and erased my memory cards at nights so I could be ready for next day. When I took photos they looked perfect on my camera.

But after my vacation I began to see some ruined images like the one below. Horizontal black lines in different places of frames were in some of my images. Some were not even visible in Adobe Bridge thumbnails and I saw that when I opened those files.

My problem is why these files got broken? Is there a problem with my memory cards or the sensor of my camera? Or is it because I might have used cut/paste instead of copy/paste or maybe I shot too fast that memory card's writing speed couldn't catch up with me?

I'm using D800. Thanks a lot.

corrupt raw file

corrupt raw file

  • 1
    \$\begingroup\$ It looks like something went bad during the copy process. That could have happened in a few places along the line. It's won't be a speed issue though, files will either write or they won't. The only way you'll be able to find out what happened is to replicate what you did and check carefully at each step to find out where the corruption is coming in. \$\endgroup\$ Commented Sep 1, 2013 at 12:03
  • 5
    \$\begingroup\$ possible duplicate of What causes this sort of RAW file corruption? \$\endgroup\$
    – AJ Henderson
    Commented Sep 1, 2013 at 18:55
  • 3
    \$\begingroup\$ Don't know why this happened I'm afraid, but you might be able to fix the images in Photoshop, you can use the magic wand tool to select the balck parts and then apply content aware fill. \$\endgroup\$
    – Matt Grum
    Commented Nov 9, 2013 at 10:26
  • 1
    \$\begingroup\$ At a guess I would suggest a memory card issue - but it could be anything. Is the issue repeatable? Have you tried different memory cards? (Some troubleshooting to isolate the issue would possibly help.) A broken shutter definitely looks different - so it isn't that. \$\endgroup\$
    – DetlevCM
    Commented Dec 9, 2013 at 10:49

2 Answers 2


A column or row of inoperative pixels will appear in the same position throughout all images.

Other than getting the camera repaired, taking several images of the same subject and shifting the camera by a few pixels between images will mask the column or row defect behind the working pixels, when the images are stacked.

Random lines are usually due to electromagnetic interference. Which may emanate from the camera but usually from an external unshielded or poorly shielded electronic source. The same technique can be applied as for row or column defects.

EDIT: But then the camera could be faulty.

PWM switching in close proximity to the camera - bias frame 1/4000sec.

PWM switching in close proximity to the camera - bias frame 1/4000sec


I would say that this is a sensor issue because sensor is read line by line.

If NEF files are losslessly compressed (this may be switched off IIRC) card corruption will not cause these simple artifacts and will rather cause box artifacts or whole image corruption.

The ADC converter is a bottleneck which gets all the throughput. There are not much more than two unreliable places - the ADC and the flash card, everything else is very complicated and would produce more more complicated errors (RAM is another place which can fail but it is usually broken in non-moving places). If NEF compression is turned on flash card failure cannot produce linear artefacts because compression is usually done in tiles. Therefore, the problem lies in ADC.

  • \$\begingroup\$ downvoted - errors like that would be consistently placed. These errors are not. \$\endgroup\$ Commented Apr 5, 2016 at 20:58
  • \$\begingroup\$ @james-snell: "errors like that would be consistently placed." - can you substantiate your statement? \$\endgroup\$ Commented Apr 6, 2016 at 7:44
  • 1
    \$\begingroup\$ My family ran an electronics repair business doing component level repairs to surface mount gear and pro AV kit when I was a kid, I've been around plenty of dead/poorly IC's to know that the failure of the sensor or a line wouldn't be nearly as inconsistent. Like everyone else I was waiting on more clarification from the OP, this question should be closed really \$\endgroup\$ Commented Apr 6, 2016 at 21:08
  • \$\begingroup\$ @james-snell: well, I did not feel like it was required to say so, but what I really wanted to say that there is a failure somewhere in ADC process, be it sensor or the ADC itself. \$\endgroup\$ Commented Apr 7, 2016 at 6:23
  • 1
    \$\begingroup\$ But the point is that even with that being the case the failure mode still doesn't support the cause in this answer, hence the downvote. The more likely scenario is the one linked by AJ but as the numerous requests for clarification have gone unanswered any answer would be guesswork. \$\endgroup\$ Commented Apr 8, 2016 at 8:44

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.