Second attempt at an astrophoto

[aparker]

So I have finally had a clear night without interference from work, travel, horrible seeing, etc, in which to try to apply a little of my learnings from EAA to a more systematic picture-taking exercise.  Given that it runs about 2x per year that I can do this, I foresee a protracted learning curve....

 

Posting what I was able to throw together here, and very interested in any hints or suggestions.

 

What I used:

My C8 with Optec focal reducer - operating at ~1330mm EFL.  I realize this is long, but hey it's what I have.  New 60/240mm guide scope, with cheapo orion guide cam.  I have this VERY firmly attached to the C8 OTA, so hopefully zero flexure.  Bhatinov mask for focusing. CGEM.  PHD2.  I followed the EXCELLENT guidance in this fantastic thread: 

To really start to understand how guiding works.  Among other things I now know what periodic error looks like, because its obvious in my guiding graph...  I used my Ultrastar (uncooled ICX825 CCD) as the imaging camera.  Someday I may be good enough at this to merit a cooled camera, but the EAA cam will do for now.

My subject was M27 - sticking with bright objects for now, esp given the long FL.  I was doing this from my city home in Boston, so red/white zone LP.  I had a fairly short session, as it's only truly dark here from 11-2 in midsummer, and I go to work at 6AM...

I was using narrowband filters, initially did:

40 x 30 sec Ha

40 x 30 sec OIII

12 x 30 sec dark

13 x 30 sec Lum (through CLS-CCD LP filter

 

I was stacking these live with Starlight Live while saving the FITS files for later processing, so was able to preview how much information was piling up.  I then decided to see if I could do longer subs given that my guiding was pretty stable, and shot 6 x 180 sec through the OIII filter, and 6 x 180 sec dark to match.  The live stacks looked just as tight as 30 sec, which I was pretty happy about.  ( On my list is to see how long I can guide at this long focal length before the star size/shape degrades versus the 30 sec benchmark.  I suspect it is quite a bit more than 180 sec, and I may run into limitations of an uncooled camera first. )

At this point it was after 1 AM and sleep beckoned, so I resolved to try again the next night (also forecast to be clear).  Unfortunately that night rolled around and the seeing was atrocious, so no more subs were collected.  But I spent some time playing with what I had.

What I did was done entirely with Nebulosity 4; I have not yet explored any of the fancy processing programs like Pixinsight, Maxim, etc.  I have GIMP but did not use it here.  I first dark-subtracted and normalized all the frames for each filter/exposure combination, then stacked using drizzle.  I then took the four resulting stacks, and normalized and aligned them with each other.  I then experimented with various combinations of the Ha with the two flavors of OIII (either 40 x 30 or 6 x 180) to compose an LRGB image.  I also tried some RGB without the L frames; in this case I liked the way the nebula looked but had few, horribly color distorted stars.  Doing the LRGB, I found that the 6 x 180 OIII frames gave a much nicer result, and ended up using them for both the B and G channels in my best attempt.  This despite the fact that the total integration time is about the same.  I also employed the curves function in Nebulosity, three times in sequence, to achieve a nice stretch, and used Nebulosity's default auto color balance.  The below is the best of several tries I made.  There are still some odd color halos on some of the brighter stars, and the mapping of pure color channels to the narrowband data definitely produces a slightly "artificial" look versus real RGB photos I've seen, but still pleasing to my eye.

My main takeaways from this exercise are that I should try to find time to do a larger number of 180 sec (or even longer if I can keep the guiding good) exposures in OIII, and redo the Ha at 180 sec.  Given that I used the same frames for G and B, the above image is a total of 44 minutes of exposures, the vast majority through narrowband filters.  Ideally I think I would try to add a couple hours worth to this. Also collect more Lum exposures, but not sure if I really need to go longer than 30 sec on those - I worry that the brighter stars may get bloated.  If possible this coming weekend I will try to do this from my dark sky site.  Even though this is narrowband, it can't hurt.

Any other thoughts on how to go forward with this would be greatly appreciated.

 

 

[aparker]

For comparison, here's the same frames, but not normalizing intensities on the L, Ha, OIII before doing LRGB synthesis

Not too different but I thought the star colors were perhaps a bit worse.

Here it is doing just RGB, no L frames:

I kind of like the slightly smokier and more 3-D look of the nebula, but the starfield loses out horribly without the L.

[swag72]

M27 has a wonderful outer shell, but you are going to need longer subs to get it convincingly I would suggest. Your guiding looks like it's fine at this point - so why not try a longer sub? Do some more darks and also dither and this should help with the noise as well. This is looking a very good result so far

[uhb1966]

Beautiful shot !

[laudropb]

Very nice image for the amount of data collected.

[toxic]

yes very nice indeed  and as Sara says go longer if you can, do a minimum 300 secs of OIII and HA for the outer stuff (i found most of it was OIII)

[Maximidius]

Wonderful image, i'd be well proud of it. Keep up the great work 

[aparker]

Thanks all - I will see if I can bulk up the image with 180-300 sec exposures this weekend.  Fingers crossed for good seeing. which is often my limiter for doing stuff at this long FL.

[ultranova]

Great result, loving the field of view, stars look great

well done

Paul

[aparker]

Here's a quick stack/stretch of 12 x 180 I was able to grab in Ha last night.  Seeing was less than perfect, but this definitely has more of the dimmer parts of the nebula captured.

 

 

[PatrickGilliland]

Well let's start where we should - if that is your second attempt at an Astro pic then great work.. i would have been over the moon with that.  Suspect with some (often painful and frustrating dedication) you have a great future ahead of you.

As Sara and others have said - longer duration will start to release the outer shell - your HA is only just starting to get the brighter elements of this.  Increasing sub length does look viable as stars are just about holding shape but only just a little drift is apparent.  

I would lock everything down at 300 seconds before i pushed any further and really get getting bang on.  You may be surprised even at 300 seconds sufficient subs will start to realise more data, not as much as 600 second for sure but it will start to add up.  What you will also find (consider i cant see the histogram that goes with the pic) is that data is being cropped or overlooked towards the left/dark side of histogram, with more subs you will be able to render this better and the integration process should realise its there!

Great start - keep at it.  DO the basics lock them down and then move it up.  Get it right now at 300 then move to 600 is half as hard!

Paddy