Jump to content

Banner.jpg.b89429c566825f6ab32bcafbada449c9.jpg

Help With imaging Techniques please!


TimBarber

Recommended Posts

Hello 


 


Im very new to Astrophotography or any type of Astronomy for that matter and i could really use some help on the processing side of things.


 


 


I have a couple of questions if anyone could just take a few moments to explain i would very much appreciate it.


 


 


1) I have read people stacking 6 x 10 sec exposures.......is this a good technique?.....good enough to get details in M31 with just a zoom lens?


 


2)If i stack say 6 x 30 secs of different images....could i take one image in raw and stack the same image of a 30 sec exposure 6 times?


 


3)Is there a difference between 6 separate images of the same exposure and same point in the sky, and one picture of a point in the sky stacked 6 times? If there is a difference i really don't understand why.


 


 


I realise that these questions make me appear pretty stupid, which to be honest in the world of Astrophotography that is fairly accurate. 


 


 


Any help would be very appreciated.


 


 


Thanks in advance


 


 


Tim



Link to comment
Share on other sites

Hello 

 

Im very new to Astrophotography or any type of Astronomy for that matter and i could really use some help on the processing side of things.

 

 

I have a couple of questions if anyone could just take a few moments to explain i would very much appreciate it.

 

 

1) I have read people stacking 6 x 10 sec exposures.......is this a good technique?.....good enough to get details in M31 with just a zoom lens?

 

I can't comment on how good it will be with M31 as I've only attempted imagine widefield (I also don't have access to a telescope).

 

2)If i stack say 6 x 30 secs of different images....could i take one image in raw and stack the same image of a 30 sec exposure 6 times?

No. The reason is that (in my experience so I'm willing to be proved wrong) each exposure will pick up something different (different details in the sky)

 

3)Is there a difference between 6 separate images of the same exposure and same point in the sky, and one picture of a point in the sky stacked 6 times? If there is a difference i really don't understand why.

Each exposure may pick up more or less detail. The reason you stack the images is to add / bring out more detail so that the end result is far more detailed. If you imagine the amount of data you get from ten seconds worth of exposure, imagine exposing for sixty seconds; you'd get far more data. The reason I stack images is that with an eighteen millimetre lens on a crop sensor, I can only go to about fourteen seconds before the stars start trailing or I get fat blobby stars so to get sixty seconds worth of data, I'd need four and a bit exposures at fourteen seconds.

 

I realise that these questions make me appear pretty stupid, which to be honest in the world of Astrophotography that is fairly accurate. 

 

 

Any help would be very appreciated.

 

 

Thanks in advance

 

 

Tim

Hi Tim,

I've replied inline to your questions. Hope that helps a little bit. :)

James

Link to comment
Share on other sites

The same image is the same image so there is no point in stacking it.

The point of stacking is very simple; the unwanted camera 'noise' (the visual equivalent of the mushy hum and crackle on old sound recordings) is partly random. A pixel which is randomly too bright on the first image may be randomly too dark on the next one. This means they will cancel out. But the real signal fom the object is constant and will not vary between subs so it will not cancel out. This statistical neutralization of noise goes on working till you have maybe 40 images in a stack. To halve the noise you need four times as many exposures. Attaboy!  :grin:

Olly

http://ollypenrice.smugmug.com/Other/Best-of-Les-Granges/22435624_WLMPTM#!i=2266922474&k=Sc3kgzc

Link to comment
Share on other sites

The same image is the same image so there is no point in stacking it.

The point of stacking is very simple; the unwanted camera 'noise' (the visual equivalent of the mushy hum and crackle on old sound recordings) is partly random. A pixel which is randomly too bright on the first image may be randomly too dark on the next one. This means they will cancel out. But the real signal fom the object is constant and will not vary between subs so it will not cancel out. This statistical neutralization of noise goes on working till you have maybe 40 images in a stack. To halve the noise you need four times as many exposures. Attaboy!  :grin:

Olly

http://ollypenrice.smugmug.com/Other/Best-of-Les-Granges/22435624_WLMPTM#!i=2266922474&k=Sc3kgzc

Thanks so much for explaining that! makes sense

Link to comment
Share on other sites

A simple analogy might help you understand why stacking images helps.  If you were to toss a fair coin (i.e. on that is not loaded), what is the probability of getting heads? We all know it is 50%, so half the time you get heads, half the time you get tails.  We'd write that as a probability of 0.5 (i.e. half).  But what if you set out to measure that probability for yourself?  I've done that for you below - I have tossed a coin 1,000 times and after each toss I have counted the total number of heads so far and worked out the probability of getting one:

post-18840-0-95356700-1412238129_thumb.p

(Don't worry, I didn't really stay up all night tossing coins, I used a spreadsheet to simulate it, but the results are valid).

You can see that during the first few coin tosses, the probability of heads is zero, since all tails came up.  For the next 50 or so, the probability is nearer 0.6 than 0.5 since I got more heads than tails overall.  Eventually the line gravitates towards 0.5, but it never quite stays there.  This is what your maths teacher would call 'in the long run', which is a way of saying that no matter how long you toss the coin for, you will never be 100% certain of the probability of getting a head.  You'll get close to knowing though.

The same thing happens in imaging - photons don't arrive at your camera in a constant, steady stream.  They arrive randomly in fits and starts because that's the way that light is generated by atoms in stars and nebulae - randomly.  You have to measure lots of photons to become reasonably certain how bright an object is.  Unlike the coin, you don't have the benefit of prior knowledge of what the true value should be though.

There are also more sources of noise than just read noise - dark current noise and skyglow noise will also come in to it, but the upshot is that the longer you measure for the easier it becomes to distinguish the target from the noise.

In the ideal world, there would be no reason to take multiple exposures.  All that matters is how long the total exposure time is, so 1 x 60 second exposure should be as good as 6 x 10 second exposures, or 10 x 6 second exposures.  The longer the total exposure, the more certain you are of the brightness of the object.

In the real world there are limitations that mean multiple exposures become necessary or even desirable:

- You mount may not be able to track accurately for very long, limiting you to short exposures.

- Even with a good mount/guiding setup, you will eventually want to take many exposures over multiple nights.

- Your camera sensor will have limits as to how long it can expose for before it goes non-linear (stops measuring photons accurately).  How long that is depends on the camera sensor and the brightness of the target, but it can be measured so you can figure out the best exposure length.

- Multiple exposures allow you to do pixel rejection - so if you have an aeroplane trail, satellite trail, cosmic ray hit (bright spots/streaks) or hot/cold pixels in the sensor (broken basically) you can compensate for all of them as the unwanted information will not be in the other exposures.

The downside of multiple exposures is that each one of them has its own dose of read noise, but the benefits of dealing with the above always outweigh this once you get to reasonably long exposures (a few minutes or more).  With really short exposures of a few seconds, read noise is going to be much more of a problem and you would need way more total time for short exposures than you would for longer ones to compensate.

Link to comment
Share on other sites

A simple analogy might help you understand why stacking images helps.  If you were to toss a fair coin (i.e. on that is not loaded), what is the probability of getting heads? We all know it is 50%, so half the time you get heads, half the time you get tails.  We'd write that as a probability of 0.5 (i.e. half).  But what if you set out to measure that probability for yourself?  I've done that for you below - I have tossed a coin 1,000 times and after each toss I have counted the total number of heads so far and worked out the probability of getting one:

attachicon.gifcoins.png

(Don't worry, I didn't really stay up all night tossing coins, I used a spreadsheet to simulate it, but the results are valid).

You can see that during the first few coin tosses, the probability of heads is zero, since all tails came up.  For the next 50 or so, the probability is nearer 0.6 than 0.5 since I got more heads than tails overall.  Eventually the line gravitates towards 0.5, but it never quite stays there.  This is what your maths teacher would call 'in the long run', which is a way of saying that no matter how long you toss the coin for, you will never be 100% certain of the probability of getting a head.  You'll get close to knowing though.

The same thing happens in imaging - photons don't arrive at your camera in a constant, steady stream.  They arrive randomly in fits and starts because that's the way that light is generated by atoms in stars and nebulae - randomly.  You have to measure lots of photons to become reasonably certain how bright an object is.  Unlike the coin, you don't have the benefit of prior knowledge of what the true value should be though.

There are also more sources of noise than just read noise - dark current noise and skyglow noise will also come in to it, but the upshot is that the longer you measure for the easier it becomes to distinguish the target from the noise.

In the ideal world, there would be no reason to take multiple exposures.  All that matters is how long the total exposure time is, so 1 x 60 second exposure should be as good as 6 x 10 second exposures, or 10 x 6 second exposures.  The longer the total exposure, the more certain you are of the brightness of the object.

In the real world there are limitations that mean multiple exposures become necessary or even desirable:

- You mount may not be able to track accurately for very long, limiting you to short exposures.

- Even with a good mount/guiding setup, you will eventually want to take many exposures over multiple nights.

- Your camera sensor will have limits as to how long it can expose for before it goes non-linear (stops measuring photons accurately).  How long that is depends on the camera sensor and the brightness of the target, but it can be measured so you can figure out the best exposure length.

- Multiple exposures allow you to do pixel rejection - so if you have an aeroplane trail, satellite trail, cosmic ray hit (bright spots/streaks) or hot/cold pixels in the sensor (broken basically) you can compensate for all of them as the unwanted information will not be in the other exposures.

The downside of multiple exposures is that each one of them has its own dose of read noise, but the benefits of dealing with the above always outweigh this once you get to reasonably long exposures (a few minutes or more).  With really short exposures of a few seconds, read noise is going to be much more of a problem and you would need way more total time for short exposures than you would for longer ones to compensate.

This is so kind of you to take the time of explaining this to me so eloquently. I really really appreciate it, and even to a total newbie i understand the principle now thanks to you 

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.