Gradient in a dark frame ??

[Skipper Billy]

Just dusting everything off ready for when darkness returns in a few weeks.

Making a new library of darks with an Atik 460EX mono - metal dust cap screwed on to the camera.

Decided to massively stretch a dark frame to have a look and its got a shocking top to bottom gradient!

Is that normal ?? It was taken in a brightly lit room but with a metal dust cap on it.

Thanks in advance.

 

[sloz1664]

The Master bias frames on my Atik 460EXm like yours but my darks are quite even and no gradient no matter how much I stretch it. At what exposure and temperature did you produce your darks?

Steve

[Skipper Billy]

Steve - 600 secs at minus 10º

 

[steppenwolf]

Can also post up a BIAS from the same camera also at -10º ?

Just as a check for light leakage, also take another 10 minute -10º dark but in a dark room and compare the new dark with the old, applying exactly the same stretch.

[Uranium235]

Hmmm.... try subtracting a bias frame from it, then do the stretch again. Bit surprised youre doing darks with a 460, personally I wouldnt have bothered with it being as its quite a low noise chip in the first place.

[RichLD]

FWIW I always do my darks and bias in a dark room with the camera in a large box plus dustcap plus a foil cover to block IR. This is possibly overkill though!

[ollypenrice]

If you owed someone ten quid you wouldn't call that a 'shocking debt.' If you multiplied the ten quid by a thousand then you might be shocked to find yourself owning them ten grand! I think massively stretching a dark is analogous with that. Try reading the pixel values of the linear dark to see what the gradient really amounts to before the stretch.

That said, my own 460 dark shows considerably less of a gradient. What about temperature? The 460 doesn't have the best cooler so was the room on the warm side? Your gradient looks a bit like my test images before the cooler has got going. The FITS header will tell you the temperature.

Like Uranium235 I don't use darks with our 460. I tried them, comparing stacks calibrated with bias and stacks calibrated with darks. I could find no difference between the two so bias it is, for me.

As an aside, the real noise I find wben processing 460 data is to be found in the regions of low signal (background sky). It takes the form of a considerable speckling of over-dark pixels with values below the 23 I'm aiming for in a background. I deal with this by pinning the curve at 23 and above and giving it a lift below 23 to compress the dark sky histogram. This gives a more natural look than an NR routine.

Olly

[Adam J]

The only thing that really matters is can you see it in the final image. If you cant then why worry about it. I remember my old 1000D used to have a very similar gradient to that across the sensor but so long as I did not try any crazy short exposures it was never visible in the final image. Like olly says it could all amount to next to nothing that you have stretched so as so make it look like a big issue. A good test is to take an an actual image and stretch it roughly as required then apply the same stretch to the dark, You will likely find that with that level of stretch the gradient is not visible. 

[Skipper Billy]

Thanks guys

The actual temp was -10º according to the FITS header.

I think I am worrying about nothing.

This forthcoming season I am going to have a go at not using darks - but wanted to be prepared just in case!

Thanks everyone.

[Jessun]

That's what my 460s used to look like before they cooled down. I could watch the frames improve until they looked normal. 15 minute wait sometimes on a warm day.

/Jesper

[ollypenrice]

Yup, if I were to worry about anything here it might be the cooler. I don't know if the FITS header value can be taken as gospel but somebody at Atik would know.

Olly

[Skipper Billy]

The cooler can only do about 23º delta - the room temp was about 15º and it was struggling to hit -10º

I think I will pop it in a cool box and see if that helps.

[Adam J]

1 hour ago, Skipper Billy said:

The cooler can only do about 23º delta - the room temp was about 15º and it was struggling to hit -10º

I think I will pop it in a cool box and see if that helps.

Really? wow that's actually less cooling performance than my home made DSLR cooler if that is true! But looking on FLO it seems that it really can only achieve -25c.....