Re: [APML] New way to undsharp mask.

[Wei-Hao Wang <whwang@asiaa.sinica.edu.tw>]

Hi Jason,

This is a smart method!  Traditional USM enhances high frequency 
components in an image.  The high frequence components include
stars, noise, and details in the nebula.  In your method, before
the mask was applied to the original image, the second Gaussian
blur filtered out the highest frequency ones, stars and noise.
So eventually only intermediate frequencies are enhanced.

I have another version of USM that picks up only intermediate
frequencies too.  What I do is:
(1) duplicate the image
(2) high-pass filter the duplicate. this filters out low frequencies
    and only stars and noise are left.
(3) change the L channel curve of (2) into this shape "_/"
    this remove the negative halos around the stars.
(4) subtract the results of (3) from the original, so the original 
    doesn't contain stars and noise any more.
(5) change curves or do normal USM on the starless original to enhance 
    nebulas.
(6) add (3) back to the original.

These will give results very similar to yours.  Except for my step (3),
our methods are mathematically equivalent but yours seems easier to do.

http://www.ifa.hawaii.edu/~wang/gallery/picutres/m8_m20-2003.htm
This is a picture that I applied this method.  In step (5), I didn't
do USM but just applied curve chance.  Without this method, the stars
will look too bright and saturated after the nebula is enhanced.

Cheers,

Wei-Hao



_______________________________________________
Astro-Photo mailing list
Astro-Photo@seds.org
http://seds.org/mailman/listinfo/astro-photo