What would happen if...........

[raymo]

If one stacked a large number [ say 50, 75, or a hundred or more] of very short subs [say 8 secs] of M42 in order to retain the core of the nebula, what would happen if one added a single long sub that reveals more of the nebulosity and obviously blows out the core, and would the result vary according to how many very short subs were used? In other words, would the effect of the one long sub be diluted? Also, would the extra nebulosity shown by the long sub be diluted as well?
raymo

[glend]

DSS will warn you if your subs are of different lengths but not sure how it affects processing. Why not just use 'layering' in Photoshop to place the Trapezium short subs into longer gas cloud subs - is this not the standard way to process M42 - even though I have not done that myself -yet. I have a library of different length subs of M42 and you certainly need a short sub to grab the Trap stars without burn through. To get it perfect I think you would need multiple 'layers' as you work outward to the cloud detail. So it becomes a true processing creation as opposed to something you just shoot and stack. Back in my film printing past it was really easy to 'hold back' an area from over-exposure during printing by just waving a small paddle in the light stream from the print head to reduce light falling on the print paper. Something of an art but gave great results with over exposed negatives. Likewise you could burn in areas of under-exposure.

[raymo]

I too spent years dodging and burning, but regardless of all the help I have received from forum members, and video tutorials I have watched,
and books I have read, I just can't get my head around layers, masks, etc:
I wasted my money getting PS, I only use it for downsizing.
I use all sorts of different length subs in DSS, often three or four different lengths in the one stack; never had a problem doing that, and have never seen any kind of warning come up.
raymo

[RickS]

What happens will depend on your stacking software and how it normalises and weights the subs, but it won't be what you want :)

You need to stack your short subs and long subs separately then combine them afterwards. As Glen said, the "traditional" way to do that is with layers in Photoshop. Another option is to use HDR techniques to combine the short and long stacks and then compress the dynamic range back down to something you can represent on a screen or print.

Cheers,
Rick.

[raymo]

Thanks Rick; it would help if I knew what HDR techniques are, but I
suspect that they are no easier than PS layers etc:
raymo

[Somnium]

my thoughts on this is that there are 3 main ways that you combine data, median, mean and sum. assuming you had many short subs and one long sub wouldn't it just have the following effect

Median - the long data will be completely rejected
Mean - the darker areas will be brightened slightly, the core will not be impacted much and the overall effect will be that you wont see much of a difference because it is diluted by all other data
Sum - the overall image will look similar to your long sub with, you will have noise issues because the majority of the image data will be from a single source

i could be completely wrong, feel free to correct me.

[Atmos]

I personally think you'd be better off taking a lot of longer subs to get a higher SNR on the darker areas and then just take a dozen or so short subs because it will not take much to get a really high SNR on those very bright areas.

Lots of short subs will get a very clean core but a single long will be incredibly noisy. The short subs are only for restoring the blowout on the very easy to capture areas, everything else is for actually capturing your desired image.

[raymo]

Thank you all for your input, lots of food for thought.
raymo

[PRejto]

The tutorial on CCDStack by Adam Block explains a very neat way to do this right in CCDStack using rejection. I highly recommend those tutorials. In fact he uses Orion as the example. It's very easy and works extremely well!

Peter

[raymo]

Thanks Peter, I'll have a look.
raymo

[luka]

Raymo, may I suggest you pick a Photoshop tutorial from the net (this one for example (http://www.astropix.com/HTML/J_DIGIT/LAYMASK.HTM)) and go through it step by step. When you get stuck or something does not work as expected, post your steps and screenshots here and someone may be able to guide you a step further. Keep notes of everything you do.

Luka

[raymo]

Thank you Luka, when it is all boiled down, a chimp could do what I need to do, one step at a time. My fundamental problem is that there are so
many steps involved in much of image processing, that unless I carry out
a particular operation VERY frequently, or said operation requires no more than half a dozen mouse clicks, then I have a snowball's chance in hell of
remembering how to do it. If I had learnt all this stuff when I was young
[it didn't exist then, of course] I would probably still be able to do it now. My short term memory is now pretty ordinary.
raymo

[luka]

Just a suggestion, I often use pen and paper or take notes on the computer and then refresh my memory when needed. I am no expert by any means and my telescope was collecting dust for a few years thanks to small children. Now I am trying to get back into it and I am quite happy that I took notes and bookmarked processing webpages then.

Lots of stuff is quite repetitive, for example tweaking curves/levels in photoshop to bring out the faint details. But doing layer masks etc... I always go back to my notes or the original tutorials.

[raymo]

Thanks for the suggestion. It will go into storage in my brain, from whence I will retrieve it when I next do some processing.[ If I can
remember to].
raymo