Question about Darks

[gnomus]

If I understand correctly the purpose of taking Darks is to get a measure of the sensor noise produced by your camera at a given ISO, for a given length of exposure, and at the same ambient temperature at which the Lights are taken.  The standard advice on taking Darks is that one should put the cover on the telescope and fire off 20+ images.

The downside of this, of course, is that the telescope is effectively "tied up", quite possibly, for several hours.  This is time that I could be spending doing some visual observing - especially important given the relatively few clear nights that we have in the UK.  

I use a DSLR.  These come with body caps - a 'plate' that covers the internals when there is no lens on the camera.  Is there any reason why I should not be able to take my Darks by: removing the camera from the scope; attaching the body cap to the camera; placing the camera somewhere near the scope (so that it is at the same temperature); and letting my intervalometer take the Darks while I use the scope for some visual observing?

[D4N]

The camera doesn't need to be attached to the scope to take dark or bias frames, just for flats and of course lights

Make sure that the cap doesn't have any light leaks.

A lot of mine are taken in the fridge!

TSED70Q, iOptron Smart EQ pro, ASI-120MM, Finepix S5 pro.

[ollypenrice]

The telescope is actually a bad place to take darks. On a Newt it can be positively disastrous because light ingress is common, notably from the bottom of the tube. It is far better to use a cover over the sensor, ideally metal but plastic ones will do. Stray light however slight must be scrupulously avoided.

Olly

[Knight of Clear Skies]

if the sky is clear you may be better off taking more lights instead of any dark frames, until you have about 25-30 lights. The stacking process itself is very effective at removing noise. As I understand it, darks become more important when long subs are used and at higher temperatures. After reading this interesting discussion I've simply given up on taking darks for the time being. I'd much rather image the night sky than the inside of my lens cap.

[gnomus]

The telescope is actually a bad place to take darks. On a Newt it can be positively disastrous because light ingress is common, notably from the bottom of the tube. It is far better to use a cover over the sensor, ideally metal but plastic ones will do. Stray light however slight must be scrupulously avoided.

Olly

Thank you.  This is extremely helpful.  I had clearly misunderstood what I had been reading.

[ollypenrice]

Thank you.  This is extremely helpful.  I had clearly misunderstood what I had been reading.

Darks are all about the camera (and the temperature and maybe the cables etc etc) but they have nothing to do with the telescope.

I take a slightly radical view of flats as well but that's for another discussion.

Olly

[gnomus]

Darks are all about the camera (and the temperature and maybe the cables etc etc) but they have nothing to do with the telescope.

I take a slightly radical view of flats as well but that's for another discussion.

Olly

Thanks again.

I'd be interested to learn more about things like How many Darks, Biases, Flats. So far I have set my intervalometer to take 25 Lights. I then just set it off taking the same number of Darks. I finish by taking the same number of Biases and Flats (when I take Flats). I wonder, however, if I could get away with fewer Darks, Biases and Flats.

Regarding Flats, I have a (I think) Lumiquest 'softbox' that fits over my flashgun. For extra diffusion I hang a white shirt over the end of the dewshield, and I fire off with the softbox maybe a foot or so from the shirt. I have my doubts about this arrangement, so if there is a better way I would be interested to hear it.

[lensman57]

If I understand correctly the purpose of taking Darks is to get a measure of the sensor noise produced by your camera at a given ISO, for a given length of exposure, and at the same ambient temperature at which the Lights are taken.  The standard advice on taking Darks is that one should put the cover on the telescope and fire off 20+ images.

The downside of this, of course, is that the telescope is effectively "tied up", quite possibly, for several hours.  This is time that I could be spending doing some visual observing - especially important given the relatively few clear nights that we have in the UK.  

I use a DSLR.  These come with body caps - a 'plate' that covers the internals when there is no lens on the camera.  Is there any reason why I should not be able to take my Darks by: removing the camera from the scope; attaching the body cap to the camera; placing the camera somewhere near the scope (so that it is at the same temperature); and letting my intervalometer take the Darks while I use the scope for some visual observing?

Do not take the dark on the scope, just remove the camera off the scope put the camera cap on the opening and set the pc to take the appropriate darks. I normally leave the laptop and the camera in the garage after a session if need be and let it do the darks, this way it is more likely to take the darks at a temperature close to the lights as the software scaling algorithm will not be able to correct for large temp difference.

A.G

PS: I usually take about 200 bias, as many darks as time allows and between 50~70 flats. Believe it or not these calibration frames can actually introduce unwanted noise into the lights therefore a large number of them is needed to yield as noiseless a calibration frame as possible.

[Bizibilder]

I use a DSLR and have completely given up using darks.  They only seem to add noise to the final image.  Far better to take lots of subs!  I usually take 5min subs and stack at least 30 if I can.  To get the noise right down I find that 60+ is better.  Unfortunately to to the scarcity of clear nights I often end up with around 30 and a noisy image!!  I do however always use flats and bias.

[gnomus]

I use a DSLR and have completely given up using darks.  They only seem to add noise to the final image.  Far better to take lots of subs!  I usually take 5min subs and stack at least 30 if I can.  To get the noise right down I find that 60+ is better.  Unfortunately to to the scarcity of clear nights I often end up with around 30 and a noisy image!!  I do however always use flats and bias.

OK ... Well that's a little surprising. I thought Darks were the most important of the calibration frames! Are you cloning out the hot pixels then if you aren't using Darks?

I am certainly going to try 5 minute subs next time, and I'll aim for 30+ of them.

[Bizibilder]

Yes - I only get a few hot pixels and it takes just a minute or two to get rid of them if they are a nuisance. 

[gnomus]

Yes - I only get a few hot pixels and it takes just a minute or two to get rid of them if they are a nuisance.

Thanks again. I'm trying to photograph M33, as I type. Unfortunately, clouds keep passing overheard - rather more than the 0% that Clear Outside predicted!