Combining data from 2 sessions

[TrevorW]

I'm taking another set about 2 hours data of M42 too add to my previous effort

How accurate does the alignment need to be for stacking purposes with DSS

I doubt that I'll be able to align everything perfectly from the other night but it'll be fairly close :help:

[jjjnettie]

I guess there's only one way to find out. LOL
Over winter I was collecting subs of m20 over different sessions. And DSS coped ok with it.
But I was using a refractor and had no diffraction spikes to contend with.
Good luck.

[Tandum]

DSS will adjust it ok but if it's miles off like this one you'll end up cropping the rubbish out.
This is straight from DSS, just stretched it a bit to see it.
The lines on this are from the overscan areas on a qhy8 image.

[TrevorW]

Well tried it in DSS and didn't like the final result a total of 51 frames 4 hrs data basically 2 from last night in 10 minute and 330s subs would not combine satisfactoryly with my data from the 16/12 unfortuantely I don't have Registar and can't justify $150US for an application I cannot test accurately from the trial version nor do I know whether it would remove the overlapping lines that the trial version showed in the final image combine

Anyone want 2 hrs M42 data for free

[multiweb]

Registar will definitely align everything. Now your blend lines will always be there. That's got to be worked out independently. Such as gradient removal, masking in PS. Normalising the raw frames. There are a couple of ways to do it. No easy "fit all" solution though.

[TrevorW]

Thanks Marc

the other thing I noticed with RegiStar is that it doesn't handle RAW files like DSS does

This might sound dumb but how do you normalise the raw frames

I would assume in a program such as Canon Digital Photo Professional you would apply the same levels historgram balance etc to all the subs

[multiweb]

I do all normalisation in CCD Stack. Normalisation is important to do data rejection prior to stacking. That's how you'll define hot pixel, outliers, etc... It will not get rid of gradients or difference in luminosity. Scaling, DDP and blending operations have to be done manually unfotunately. I don't know of any program who does automatic gradient removal or accurate automated masking. Doing a Sum sometime works better than doing a Mean stacking when differences exist in Luminosity and the frames don't match with evident seams.

[TrevorW]

Well thats beyond my software

I don't scrap anything normally darks and flats aren't captured on the same night time constraints with having to set up and dismantle the rig after every session etc

so everything good and bad is thrown into one pot and hopefully like a stew something good comes out of it

my version of pot luck

Cheers

[Octane]

IRIS is free and does everything. It has a normalization command (I run it prior to stacking). Its gradient removal is amazing.

Trevor, not taking darks/flats at the same time (after/prior) to imaging is going to cause you to present sub-optimal results.

It all depends on how critical you want to be, I guess, but, I think your images could improve greatly if you did go the extra mile.

Cheers.

Regards,
Humayun

[cfranks]

Darks you can take just about anytime providing you use the same exposure time and temperature as were for the Lights. Flats have to be taken with exactly the same focus so it is a bit more complicated doing at another time. A Light Box works fine for most people. Use it when you finish imaging, the exposures are fairly short, about 0.02 secs for my setup.

Charles

[TrevorW]

Thanks H I don't disagree (where is Nebulosity free I checked out Stark Labs and they want $45 for V1 and the download trial puts lines through the image)

.02 secs for flats, I usually capture 30-60 second flats ???

[Octane]

IRIS is free, not Nebulosity. :)

30-60 second flats -- are you sure?

Flats are taken at the same f/ratio as the scope, at ISO-100. Shouldn't be longer than a second, at most (depending on intensity of flat-light source).

Regards,
Humayun

[TrevorW]

Sorry H read it wrong too early in the morning

I've used Iris good program but long winded too do stuff

flats I normally capture at twilight so your saying don't need to be that long then

Cheers

[Octane]

Trevor,

Yeah, just make sure the histogram peaks around 1/2 to 2/3rd of the histogram on the back of the LCD.

Regards,
Humayun

[AlexN]

Trevor.. A few thoughts..

1 - if your field of view is not flat, and your target is not in exactly the same position in the field of view (both location and camera rotation) any stacking/registering program will have troubles... Even registar. a curved field of view will always result in warping of the star positions, this will result in odd looking stars when stacked..

2 - Flats are critical..


What I do (and I've done 4~5 night imaging runs from different locations etc, sometimes weeks/months apart.) is firstly, get your camera orientation as close to the same as possible.. a degree or two wont matter, but definitely, the closer to perfect you can get the better... Open up one of the previous nights images in your capture program (ie. Maxim, Nebulosity etc) and then take a new exposure, switch between the new exposure and the previous nights exposure to see any difference in composure and or camera rotation... from there, figure out what needs to be done to get them as close to the same as possible. Once you've got it all sorted, take your images...

That should ensure that your night to night exposures line up well enough that you only need to crop out a small section of the field of view.