Flare on top corner

[hschnee]

Originally Posted by T-Lama

It's possible that only a very small part of the sensor (CCD reading circuit, I guess) is heating up too fast to be cooled down, even in cold weather.
By the way, what camera model is that guy using to take 2 hour exposures?
E-1, for instance, will allow 8 minute exposure max, I think.

He uses a Canon 5D mkII, and previously used the original 5D (he's a great photographer, so I like him anyway ). It's hard to say how different sensors will respond to heat now that I think about it--the larger 5DII sensor generates more heat, but also has a greater surface area for heat dissipation. And the placement of other electronic circuits could affect where and when heat builds up on the sensor.

- Hal -

[photo_owl]

Originally Posted by hschnee

the placement of other electronic circuits could affect where and when heat builds up on the sensor.

- Hal -

bang on Hal - no could about it, you only have to look at the E400 for this with long (and or high ISO) exposures.

for reasons I don't understand it (heat) generally seems to result in a purple 'flare' though, but NR should pick it up.

[throt]

Tried it out again: about same place same time but with a bit longer exposure and NR turned on. Just remembered that when got the camera read the manual and thought naah his part can forget as probably never will use long exposures.

It might be CCD and MOS physical architecture that might make the difference.
Also there might be a possibility that in years there has came "some" dust inside the body that blocks the heat transfer.
With NR on you'll get the E1 with the full slowness - 1 min exposure + 1 min NR. But as the beginning of dSLRs you woudnt expect more

Thank you all for helping out

[T-Lama]

Originally Posted by photo_owl

..for reasons I don't understand it (heat) generally seems to result in a purple 'flare' though, but NR should pick it up.

There is an article about CCDs (quite complex, if you ask me), describing some stuff as dark current, etc.

In short (if I understand it right):
Long exposure -> plenty of electrons on the way to CCD reader -> reader heating up -> corner of CCD being warmed up -> altered conductivity of affected photocells -> false reading from affected photocells -> violet blob on the picture.