Reprocess of M101, just over 4h of data

[michael.h.f.wilkinson]

Not really enough, but I think I am getting better at teasing out a bit more detail.

Still need to control the colour noise in the background better

[wimvb]

Very nice!

29 minutes ago, michael.h.f.wilkinson said:

Still need to control the colour noise in the background better

What software do you use? Pixinsight allows one to target chroma noise aggressively without destroying detail. 

[vlaiv]

48 minutes ago, michael.h.f.wilkinson said:

Not really enough, but I think I am getting better at teasing out a bit more detail.

Here is a simple recipe that will help you tease out even more detail:

- bin your data 2x2 in software. Or even 3x3 - it will still be same size when fit to screen is done (FireFox says it shows it as 29% scaled when opened in new tab).

[michael.h.f.wilkinson]

1 hour ago, wimvb said:

Very nice!

What software do you use? Pixinsight allows one to target chroma noise aggressively without destroying detail. 

I just use Gimp and AstroImagej. I write my own software in C at work, and want to develop a so called morphological connected filter for this task. It will be based on the MTObjects/Sourcerer code base. This is a replacement for SExtractor, used in astronomy for detection of objects. Our method is way more sensitive, and uses statistical tests to sift noise from real structure effectively. Should be fun to develop this idea into a tool that enhances only detected structure, and suppresses the noise. 

[michael.h.f.wilkinson]

1 hour ago, vlaiv said:

Here is a simple recipe that will help you tease out even more detail:

- bin your data 2x2 in software. Or even 3x3 - it will still be same size when fit to screen is done (FireFox says it shows it as 29% scaled when opened in new tab).

Certainly, binning helps. 

[alan potts]

Nice image Michael, miserable here, cloud and rain.

Alan