Image alignment... HELP!

[Full-Frame]

I really want to be able to create an image of the milky way. However, I can't afford a tracking mount and have to stick with short exposures of 20sec or less.

Firstly, does anybody know if there is software that exists which is able to align the images?

2nd, can anyone direct me to an article which explains the use of "dark frames" to reduce noise? I keep hearing people refer to subtracting dark frames and I have no idea what it means and can't find any reference to it!

Many thanks in advance!

[ollyb]

so a noob and looking forward to trying this once I get home from travelling. I was under the impressions that DSS (DeepSkyStacker) does frame alignment for you, but this leads to another question I was going to ask (not trying to hijack!). If DSS does do alignment, is it still preferable to track if possible? and if so, would using a motorised eq mount cause unwelcome vibrations in the image due to the movement?

Cheers

[Jeff]

See Troy's thread on software for links:
http://photography-on-the.net/forum/showthread.php?t=770788

For your questions:

A tracking mount isn't required for widefield shot like you want to do but would make it slightly easier on the software that aligns & stacks them.

Keep the exposures short like you say 20-30 sec depending on how wide you are shooting. ISO 800-1600 and wide open f/stop. Take lots of frames (these are your "lights"). 25-50-100? Whatever your time and patience will allow.

Then take some "darks"; same settings as lights with lens cap on so no light reaches the sensor.
I've read about 3:1 light to dark ratio but who knows...take at least 10 I'd guess.

See Baz's tutorial on stacking, it's great.
http://photography-on-the.net/forum/showthread.php?t=763083

re: motorized mounts & movement: No, they're smooth enough that you won't get any vibrations from the mount. Walking around, wind, etc may cause it to jiggle though.
Yes, tracking over non-tracking is preferable, but not required until you get to medium/long focal lengths.

[Sorarse]

My basic understanding of dark frames is that they help to reduce noise and hot pixels.

The noise pattern produced by your sensor will vary depending on exposure time and temperature. This is why it is important to take your dark frames at the same time as your light frames to maintain a comparable temperature, and also why it's important that your dark frames should be taken at the same shutter speed as your light frames.

The stacking software can then see where the noise and hot pixels are in the dark frames, and use that information to subtract them from your light frames, improving your light frames in the process.

That's my understanding anyway.

[spit]

DSS or any stacking software can align the images- but without tracking, youll have trailing

[hollis_f]

DSS or any stacking software can align the images- but without tracking, youll have trailing

Not that I noticed -

http://www.frankhollis.com/temp/2009-07-04_110249%20Stacked%20Small.jpg

This was my very first attempt at such a shot. DSS did a great job.

[ollyb]

Nice image! Was the tree only exposed in one of the frames (by flash?) of is that the result of a build up of many images without any particular lighting? How many frames?

[troypiggo]

From the Deep Sky Stacker website (http://deepskystacker.free.fr/english/index.html):

FAQ: Light, Dark, Flat, Bias... What are they and how to create them? (http://deepskystacker.free.fr/english/faq.htm#lightdarkflatoffset)

How to create better images

(http://deepskystacker.free.fr/english/theory.htm)

[MintMark]

The only problem I have had with using deep sky stacker to align images taken from a fixed tripod is that if the view is very wide, then deep sky stacker can't compensate for the perspective projection of the lens. The result is that alignment fails in those parts of the image where the apparent movement is towards or away from the centre of the frame.

I saw this problem with a 10mm focal length and frames taken over a 5 minute period. The longer the focal length the less pronounced the problem will be.