ArcsignhStretch - Bright blue star halo's

Scott Iver

Hi Adam,

I think I'm going horribly wrong again...   I took some data on M31 this last weekend.  Everything seems to go pretty good until I get to the point of stretching the color (RGB) image that I earlier combined.   Maybe I'm applying way too much, but bright blue stars in my image get this strange, halo effect where the blue turns into a ring around the star, the core of the star turns more a milky white.   Orange stars seem unaffected.

I had this same thing happen to me with my Trifid data, but it was only really noticeable when zooming in.  Now with my M31 it's noticeable even zoomed out.

Could this be one of the instances you were talking about in the video when you mentioned that this proces doesn't always provide the desired result?  If so, what process or method should I fall back to?

Scott Iver

Link: https://www.astrobin.com/full/361627/0/?real=&mod=

BlockHead

ArcsinhStretch is very powerful. The stretch it is doing will mess with the luminance profile (of the color image)  and this can cause blending issues. There are some techniques for dealing with the effects of this process- but indeed you need to be careful. I would think that standard stretching techniques for M31 will do most of the work with a small amount of something like arcsin.. It really depends on the data. I believe that in Horizons I show some techniques for dealing with the issue. One relatively easy thing to do is to substitute your original RGB stars where there are problem stars. You might do this if you felt you like everything else about the color...but just not the problem stars. 

Something to think about.

-the Blockhead

Scott Iver

Ah, thanks Adam.