Is this a stupid question?

[pyrasanth]

I saw an amazing photograph of the M31 galaxy shot with just a DSLR camera & a moderately wide lens. It was full of the most amazing detail.

It used a technique called stacking where loads of 20-30 second exposures are taken then stacked in software.

There is something I don't understand- how is such a fantastic image created from dozens of images of the same exposure?

If I take one 30 second picture it has no more detail than the other 100 30 second pictures- so why does stacking bring out all the miraculous fine detail- it looks like the detail literally came from no where if you compare with just a single frame. 

[ronin]

It is to do with the way a camera collects light, the whole thing operates differently to the eye.

Also yes the images are stacked, but they are also processed. You can identify from the wavelengths the regions that are say Red, then increase the degree of reddness to make it more prominent.

In easy terms any not so good images can be thrown and you can stack just to good ones, then ask the software to sharpen the result.

The idea is simple: You collect lots of light but instead of one big exposure you use many shorter ones. Really just a case of if a 10 minute (600sec) exposure goes wrong you have nothing, if 2 exposures of 30 seconds from a set of 20 go wrong you still have 18 good ones.

The other aspect is to forget "picture" and think "data". In multiple images you have collected more data, and it is this data that is available to be worked on, and information extracted from.

Just be aware that the final image you see will be the final processed image, and the processing could have taken 2 days of work.

[romx]

Because each image will very slightly differ from each other. by combining the images it can bring out the detail from multiple exposures into a single image therefore theoretically much more detail will be present in the stacked images also. just how i expect it would work as ive not looked into imaging very much but also the process could help removing digital noise i would assume because if a specific bit of digital noise is in 1 or 2 frames only out of many it could be removed perhaps cleaning up the image. I am sure some of the advanced members will give a much better and very accurate explanation any minute though

[Leveye]

The results are much much better stacking long exposures of say 300-600 seconds than many 30 second exposure for faint objects,nebula and galaxies and the like than say star clusters which are better at shorter exposures. It really depends on the DSO your imaging.

[Bizibilder]

The main reason for stacking is to eliminate noise from the data.  If you stack frames the "image" data will add up as it appears in every frame - thus you can "add" the times of each of your frames (called "subs").  The noise that you also collect will be random and therefore when you stack your frames it will average out (in a perfect world to zero!) therefore as you stack frames the signal to noise ratio improves and you get a much better image overall.

[brantuk]

Also bear in mind the central core is brighter than the outer dust lanes. So longer exposures are required to capture the detail of the dust lanes and some not so long exposures of the central core will be more than enough. If they were all the same length then long exposures will over expose the core and blow the picture out, too short and you get good core but wouldn't capture all the detail in the dust lanes. Good question btw.

[leemanley]

The only stupid question is the unasked one !

Its all about the signal to noise ratio, this is what we're after, more signal, less noise, we gain this via multiple exposures and stack them together with dark frames- bias frames -flat frames then put through untold amounts of processing programs !

Its a large topic to cover, I'll have a look for a related topic and re-post

Cheers
Lee

[lensman57]

Hi,

Have a read of the articles in the attached link, it will explain a lot of what u need to know.

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

Regards,

A.G