PDA

View Full Version : Question: 5D MkII image anomaly


Tony Whitaker
October 5th, 2009, 02:11 AM
I have found an anomaly occurring when I shoot raw+jpeg. This manifests itself as a small array of red pixels near the middle of the image. It is 5 pixels H by 9 pixels W with brightest red running horizontal thru the middle then fading in both directions, similar to a jpeg compression array. I have narrowed down that it appears to be a problem with the jpeg processing rather than dead pixels on the sensor as the raw images do not display any error even when viewed at 3200% or pixel level. It is independent of lens or whether flash is used or not. It does not occur when shooting jpegs only. It is in exactly the same position on every image. Anybody else experienced this problem or any ideas on what may be causing it? Thanks

Bart_van_der_Wolf
October 5th, 2009, 03:00 AM
I have found an anomaly occurring when I shoot raw+jpeg. This manifests itself as a small array of red pixels near the middle of the image. It is 5 pixels H by 9 pixels W with brightest red running horizontal thru the middle then fading in both directions, similar to a jpeg compression array. I have narrowed down that it appears to be a problem with the jpeg processing rather than dead pixels on the sensor as the raw images do not display any error even when viewed at 3200% or pixel level. It is independent of lens or whether flash is used or not. It does not occur when shooting jpegs only. It is in exactly the same position on every image. Anybody else experienced this problem or any ideas on what may be causing it? Thanks

Hi Tony,

When you say it's not in the Raw data, be aware that some Raw converters eliminate Dead/Hot pixels without asking. The only way to really be sure is by looking at the Raw data itself. Also, when the same phenomenon occurs at the same position in different images, it would be suspect.

Cheers,
Bart

Mike Shimwell
October 5th, 2009, 10:40 AM
Hi Tony,

When you say it's not in the Raw data, be aware that some Raw converters eliminate Dead/Hot pixels without asking. The only way to really be sure is by looking at the Raw data itself. Also, when the same phenomenon occurs at the same position in different images, it would be suspect.

Cheers,
Bart

I don't think DXo maps out hot pixels automatically, so that would be an alternative way to see if that's the problem. One or two hot pixels culd cause such a problem if they aren't mapped out.

Mike

Tony Whitaker
October 5th, 2009, 01:10 PM
Thanks, Mike & Bart ... we (my son is a professional photographer also running a 5D mkII, and he's been looking into it with me) had considered this but because this problem does not occur when shooting only jpegs we had dismissed the possibility of a dead pixel. One thing we have not looked at yet is whether the problem occurs at all jpeg conversion settings as so far I have only ever operated with the large files. We'll try a series of test shots later today and see if we can work out just which combinations of RAW/jpeg are affected. Tony.

Kenneth Wiley
November 29th, 2010, 02:31 AM
Just curious if you have the latest firmware installed?