Trying to tame my nuclear stars - M63 as a test subject.

[graemlourens]

Hi Everybody.

I started this thread as a split-off of my previous thread:

The discussion drifted off towards my horrible lack of skill of taming my stars (i'm glad it did!)

I used this fresh knowledge gathered there and applied it as well as i could as a first-timer doing this in photoshop (i was purely a PI guy until now) to my very quickly gathered M63 data. (btw i lost a lot of color detail binning the color data, i shouldn't have done that)

I'd love your feedback if my efforts have paid off. I am WAY more happy with these kind of stars than the ones i had before, but would like some critical voices around here to judge this capture.
Most time i spent is to accurately do the star-control (with the starshrink plugin mentioned in the other thread, as well as lots and lots of layers and masks) not only on L but also RGB. Initially i had rather bad colored halo's (and i think i still have few of them lurking around)

Kind regards, Graem

[wimvb]

Very nice! You are quick to learn.

To pull out more light from the image try this:

Create a luminance layer from the rgb image.

Blur the colour image and use colour saturation on it

Apply the luminance to it as an LRGB-combination.

Easy to do in PixInsight.

 

[graemlourens]

Hi Wim.

Thank you for your tip. Tried it out, works well. I had to add a mask when raising saturation, or else my blue halos around the stars would pop up again. The difference is noticeable though. I didn't think about raising saturation like that before, always did it on the RGB image itsself. I guess your version reduces introduction of chromatic noise as we're blurring (convolution i guess you were talking about in PI?) the image first?
 

I lost a lot of blue, i think next time i'll go for slightly longer blue exposures compared to the rest.

Kind regards, Graem

[wimvb]

Blurring the easy way in PI is to use MultiscaleMedianTransform or MultiscaleLinearTransform and just double-click the first 3 - 4 waveletlayers. These are then excluded from the image. The lower the layer number, the more fine detail it represents. So, the more layers you exclude, the more you blur.