sorry - stupid question - stacking of multiple sessions with different exposure lengths

[Russe]

Guys,

this is probably a very stupid questions, but I've only just done this for the first time:

I've got 11 odd hours data on M74 as 300s (ISO800) and thought, I wanted to add 600s exposures (ISO800).

What do I do now - stack all of the data (DSS) together with their respective flats/darks/bias (as you can add different groups, I'm sure you follow).

Or do I do a separate stack now for this new data. But if I have a new stack with this data, I'll have to work on it on it's own in Photoshop and then insert it as a layer with differing degrees of opacity or smth like that?

I'm confused.

Please help.

[Alveprinsen]

If you look at the bottom of the screen of DSS, you will see there are some tabs. Main group, and Group 1.

Usually you would just smack everything into Main group, and that would be it, but with different sessions,  you put the files from each session in its own group.

Note however, that all calibration frames that are put in Main group - by default applies to ALL other groups.

So to get around this you simply put one frame in the Main group, and de-select it. Then you just put all your 300sec files in Group 1, and your 600sec files in Group 2.

Just select the first frame in Group 1 as reference frame, or whatever. Often my frames from different sessions are ill aligned. So I simply use a frame from whatever session I am currently working on as reference frame.

When you put files in Group 1, a new tab called Group 2 will automatically appear. For M42 Orion nebula I got like 5 tabs from different sessions or something.

One GOOD question though, which I have been meaning to ask is: When stacking different sessions at different exposure times and ISO, will the final image be overexposed? does on need to stack the 30 sec frames, 2 minute frames, 5 minute frames and 10 minute frames separately for high dynamic range targets, or can one stack all these together in different tabs like I explained above, and still be able to start at a "low exposure" when stretching the image in Photoshop, or WILL the 5 and 10 minute exposures overwrite the 30 sec and 2 minute exposures in the final stacked image? Lets say M42... which has a core that needs low exposures to be brought out, but long exposures to bring out all the HA surrounding it....

Alveprinsen.

[glowingturnip]

oops, think i just gave my opinion on this in another thread, my money's on keeping them separate until you blend in PS.  To be honest I woulldn't trust DSS to get it right even if it looks like it's doing it.

[jgs001]

I've done just this with different length exposures from different nights, but the differences were much less. It worked fine for me. Given the number of subs you have, you could try the Entropy Weighted Average (HDR stack)

[Russe]

Blending seems difficult and arbitrary - and where would the benefit come in from getting a smoother background?!

I've stacked it last night, exactly as you described it today - I do this routinely for all my multiple sessions (but with the same exposure). The end result now allowed for more stretching. Not bad. But not blown away either yet... I wonder how much more data needs adding to make a big difference.

As for 2min and shorter exposures - I don't think that's possible. It should simply get "rid" of the underexposed data?!

[Russe]

oops, think i just gave my opinion on this in another thread, my money's on keeping them separate until you blend in PS. To be honest I woulldn't trust DSS to get it right even if it looks like it's doing it.

Do you have a link to the other post??

[Russe]

Lol - found it!! That was my post too...

[Russe]

I've been thinking that it will average things out too. I think that it probably should be stretched and added separately... Will try later.

[Russe]

the problem with a second stack is, that there is slight rotation in it that prevents it from being layered nicely in PS...

[glowingturnip]

not sure if this would work, but if you run them all together in DSS and use the settings to save intermediate registered files, then those intermediate files will be rotated ?  Then you can split them up again, stack the longer exposures together, stack the shorter exposures together, and when you come to PS, hopefully they're all pre-aligned...

You'd have to be careful which darks/flats you use in DSS though - maybe run it twice, once with the shorter exposure darks to get the short exposure aligned files, then again with the longer exposure darks to get the longer exposure aligned files.

Remember when you do the stacks for the second time after aligning them, you shouldn't use the darks/flats again, because they've already been applied.

hope that makes sense... ?  probably won't work...

[Alveprinsen]

the problem with a second stack is, that there is slight rotation in it that prevents it from being layered nicely in PS...

Just throw in a single frame from the first stack with the second, and right klick - use as reference frame. It will be included in the stack, but one single frame give or take isnt going to make much of a difference.

If you have a 300sec session and a 600 sec session, just use a single frame from the 300sec ones together with the 600sec ones... Choose the 300sec as reference frame. All the other frames will then align on that one. Its what I do when I dont stack'em all together....

[Russe]

Great idea!! Hadn't thought of that ,..

[jgs001]

Once read that you can use the stacked result from the first stack as a reference frame but uncheck it. It'll align the stack in the unselected reference.

Sent from my GT-I9300 using Tapatalk

[glowingturnip]

both much better suggestions than mine  

[ollypenrice]

Firstly, why take different exposure lengths? They can't both be optimal and the manner of combining them becoems tricky.

I would only take different sub lengths if I had a good reason for doing so. This would normally be to cover a high dynamic range target like M42. Using a DSLR you are likely to get more star colour from short exposres, too. Another reason might be to do with guiding. Slightly trailed images can still contribute to the background sky and the faint detail. The bright stuff with sharp detail would be added as a top layer and the background and faint stuff partially erased (after being selected using Colour Select) to let the harmlessly trailed faint stuff contribute to the final image.

What I would do is stack the sets separately and then give each of them a initial stretch by moving the grey point to the left and clipping the background back to a value like 25. I'd not go so far as to get a final stretch but to around 60 or 70 percent of what the data will allow. Critically both stretches need to be by the same amount, which is easy using the grey point in levels. Just look at the number by which you've stretched.

Next paste one onto the other and see what each does best by blinking on and off. You could play with the opacity slider to get an idea of where each image is best. Then use colour select and an eraser with an opacity chosen to take off the right amount of the worst parts of the top layer. This might be all of some parts, or half of some parts, or whatever. Such an approach really isn't complicated and gives yu the chance to decide for yourself which blend is best, fifty fifty, sixty forty, etc on a zone by zone basis rather than globally.

Olly

[Russe]

thanks for your input

I wasn't planning on using other exposures but was thinking - why not let's try and add some double exposure subs. Rubbish idea, I can see that now. Much better to add more data in exactly the same way I started out... well, next time