This is just over 4 hours with the hyperstar and C11 (121 x 120s with 9 subs lost due to oval stars), again the regreased NEQ6 with SKF worm bearings did the business.
I'm having a great deal of difficulty colour balancing this one so I've posted the substandard one for now - please ignore the colour! It shows a lot of promise so hopefully I'll update it when I sort myself out.
Edit - old image removed
Yes - it certainly shows a lot of promise.
Looking forward to the final result ;)
As I said before, it'll be a very interesting comparison: Hyperstar vs Bananascope
Will you be adding diffraction spikes? :twisted:
Yep looking good Robert. Need to colour in those stars... :-). Look forward to final vers.
Now mostly sorted but there are tooooo many stars !!!!
(http://farm9.staticflickr.com/8449/8024581859_36036f17fd_z.jpg)
Max res: http://farm9.staticflickr.com/8449/8024581859_80da66517b_o.jpg (http://farm9.staticflickr.com/8449/8024581859_80da66517b_o.jpg)
Beginning to look really good now!
The colour balance certainly looks better.
Mark
Completely reprocessed. I've cut out any subs that had a colour cast because it was too near dawn and got something I'm more happy with.
(http://farm9.staticflickr.com/8030/8027935510_952f46f188_b.jpg)
Link to full sized version: http://farm9.staticflickr.com/8030/8027935510_4e759450fb_o.jpg (http://farm9.staticflickr.com/8030/8027935510_4e759450fb_o.jpg)
Looking even better now!
Quote from: RobertM
I've cut out any subs that had a colour cast because it was too near dawn
This causes a problem for the sigma stacking as well - especially when combining data from different nights with different background gradients - it leads to a "posterised" effect in parts of the image.
I now have a solution to this - it enabled me to throw all the different nights data from Makarians Chain into one stack without colour casts, posterisation or severe gradients. Essentially, the idea is to select a reference sub with the least gradients and then remove the differential gradients from all the others after registration but
before stacking. I'm still experimenting with the best way of doing this. But the technique is looking very valuable.