Lobster claw, Bubble and M52 - H(HO)O

[geeklee]

Hi everyone - this image has taken a month or so to acquire and unfortunately not because it's got so much integration time 😁  I eventually got the OIII during the (very) occasional break in clouds during October, with the Ha obtained in September.

This is using a RedCat 51, Atik 460EX, filter drawer and Baader Ha and OII filters.  My OIII is a shocker for halos, although I think I've hidden them well...

Captured with Voyager, calibrated and stacked in APP then processed in PixInsight.  I drizzled the data - just x1 - as I was starting to see square stars.  The shapes came out better but at a small cost in size.  Image scale is ~3.8"PP.

Ha - 31 x 600s (~5hr)
OIII - 30 x 600s (5hr)

I like this colour blend although it won't be to everyone's (or anyones?!) taste.  I've got some challenges with stars and a very soft look but it's done for now.

As always, thanks for looking and any feedback for improvement is always welcome.

 

There's plenty of really interesting stuff in the image, here's the annotation with more details:

Here's the raw stacks, just Auto STF in PI

Edited December 14, 2020 by geeklee
Updated version of image

[tooth_dr]

Brilliant job and excellent processing!  How does drizzling at x1 work?  I’m on the border of square stars too so am very interested. 

[geeklee]

On 29/10/2020 at 12:53, tooth_dr said:

Brilliant job and excellent processing!  How does drizzling at x1 work?  I’m on the border of square stars too so am very interested. 

Thanks Adam.  I've read and tried to absorb a lot of the discussion on SGL about drizzle and always come to the conclusion I don't require it but in this instance when I saw the square stars I decided to look at it within APP.  I read some APP forum posts and the tool tips and just tried out the options.  

I put the droplet size back to 1.0 and the scale back to 1.0 and stacked - I got an image with original dimensions and rounder stars than no drizzle. 

Running them through APP and PI for star sizes, I did gain ~10-20% in size.  Say 1.5 pixels to 1.6 or 1.7 pixels.  

I have gone back to revisit this while working on a 10 hour Ha integration of the Heart nebula.  When doing the above, I notice a very slight loss of detail - hard to see at 1:1 but it's there. SNR was very close though - almost indistinguishable - a little better if the reported values were accurate.  Reducing droplet size to 0.75 and then 0.5 (keeping scale at 1.0) did seem to recover a lot of that detail while keeping the SNR similar.

Having droplet size at 0.5 and scale at 2.0 created a noisier image (as expected) with double the resolution.  With 10 hours of data this still looked great, but SNR was a little lower. 

Worth investigating if you have some time.

Here's the section in APP in case you haven't come across it.  Changing mode from interpolation to drizzle and then working with your droplet size and scale - tool tips provide a little help.  Requirements are under sampled image and well dithered.

Edited November 8, 2020 by geeklee
Updated after more investigation with drizzle