Page 1 of 1

Horsehead Nebula

Posted: Thu Nov 02, 2017 5:11 am
by toriqo
With a ton of light pollution and the moon high in the sky in two consecutive nights (the first set was shot with a very poor drift alignment)

1st night
36 light frames of 60sec@iso1600
50 dark frames

2nd night
71 light frames of 120sec@iso1600
41 dark frames
50 bias frames

Total exposure time: 2hrs58min
horsehead.jpg
horsehead.jpg (448.52 KiB) Viewed 5928 times

Re: Horsehead Nebula

Posted: Thu Nov 02, 2017 8:37 pm
by Rowland
With that amount of data the image should blaze a hole in your screen and look fantastic. Just my opinion, but leave out the first night for now. Take some flats for the second night and do the preprocessing again. As it is, the image looks like it hasn't been wiped or stretched?

Tip: at first, don't try to cover up noise. AutoDev Wipe and have a good look over the data. With experience, this will tell you a lot about the quality of your acquisition and the effectiveness of your preprocessing. That way you can make changes to your work flow beginning with the imaging session.

Can you post or dropbox one of your subs from the second night...

Re: Horsehead Nebula

Posted: Fri Nov 03, 2017 6:35 am
by toriqo
i thought so as well (i've seen a 75min exposure which looked superb). i'll shoot some flat frames (i've got some dust on the sensor since the filter has been replaced anyway).
i've noticed i can't get any other colors to pop up since i've added the baader filter on my camera. is there something i should do (color balance on the camera or something like that)?
also, you're right about the autodev and wipe. i'm very afraid of those because once i do autodev for the second time (after the wipe) it adds a TON of noise and i rather use develop instead of autodev.
by the way, the image i posted is the wrong one. it's only a deepsky stack processed in photoshop a little bit.

this is the one processed in startools.
horsehead_nebula_st_f.jpg
horsehead_nebula_st_f.jpg (152.44 KiB) Viewed 5915 times
i'll upload the stacked 120seconds frames (with flats) and send it to you

thank you very much!

Re: Horsehead Nebula

Posted: Fri Nov 03, 2017 11:47 am
by Rowland
At this stage it's good to see the noise, as it tells us a lot about the image. We may need to change our preprocessing workflow to improve noise reduction or change our imaging parameters. Longer exposure for instance.

Re: Horsehead Nebula

Posted: Sat Nov 04, 2017 8:14 pm
by Rowland
I'm sure that one of the more advanced ST users could do a better job of post processing.

Hopefully, this is helpful and a starting point...

First up, is a very basic approach to get a look at what's going on in the image. The idea being, that if the image bounces off the screen, cleanly, with just a little basic processing, it's gunna be a pearler with a bit more work. It also tells us that acquisition was pretty good, quite apart from the total amount of data acquired.

We are just taking a peek at what we've got.

What's OK with the image?

Tracking and focus look OK...

Image calibration looks OK, but we know the flats were taken a little later, so flat division is not perfect - no big deal for this exercise.

There is evidence of fine detail, but this is drowned in the horizontal lines drifting across the image.

What's irrelevant?

Distortion around the edges and in the corners - optical and meaningless for the purpose of this post.

What's a challenge?

Light pollution - possibly, inadequate sampling - more data or longer exposures, whichever is appropriate for the conditions and gear in use.

Though calibration looks reasonable, it's difficult to say exactly because the images have not been dithered (adequately).

Without dithering, any calibration errors will be accentuated, giving the appearance of lines across the image, in the general direction of drift, between the images (very basic explanation). In this state, it's just too difficult to get a good result

All said and done dithering will clean the image up in a number of key areas and it will be easier to process.

Bin 50% and used Dev followed by HDR Reveal All, Color, and Tracking off with noise reduction - it has been despeckled and spiffed up in ImageMagick. I didn't push it hard to keep noise under control.
Autosave.jpg
Autosave.jpg (248.86 KiB) Viewed 5888 times
Horizontal lines - dithering will fix this
AutosaveAD.jpg
AutosaveAD.jpg (146.64 KiB) Viewed 5888 times

Re: Horsehead Nebula

Posted: Sat Nov 04, 2017 8:35 pm
by toriqo
Thank you very much, Rowland!
About dithering, I didn't even know what it was until a week ago. I'm going to start using it, especially now that I've also got auto guiding equipment (orion 80mm short tube, orion star shooter auto guider and an orion x-y guide star finder) and I can use PHD2 to do the dithering. As well as APT, actually.

Anyway, thanks for all the help!

Re: Horsehead Nebula

Posted: Mon Dec 04, 2017 8:30 pm
by 1074j
Results from dithering are great! I wished I did it sooner as now I can get away with not using darks as long as I have at least 20 subs and use DSS median or sigma stacking.