M106 Reprocess.

[MARS1960]

I have been reading Inside PI so wanted to implement a few things i have learned, also my other image was too blue and somewhat over saturated.

This image is not as colourful but i think it is a better process.

Any help/critique/thoughts much appreciated.

Hope you like it.

 

 

[wimvb]

Looks good to me.

If you create a luminance mask and a star mask, then use PixelMath to subtract the star mask from the luminance mask ('$T - star_mask' applied to the luminance mask), you will get a starless mask that will allow you to work on the galaxy without affecting the stars or background. You can use this to boost the colours in the galaxy.

[MARS1960]

32 minutes ago, wimvb said:

Looks good to me.

If you create a luminance mask and a star mask, then use PixelMath to subtract the star mask from the luminance mask ('$T - star_mask' applied to the luminance mask), you will get a starless mask that will allow you to work on the galaxy without affecting the stars or background. You can use this to boost the colours in the galaxy.

Ahhhh, that's great thanks Wim, i was having a tough time getting any colour out of it, now i have a solution, many thanks for the tip.

If you get the time a screen shot of pixelmath showing me would be very much appreciated, that's the one thing i haven't got into yet.

[wimvb]

You're welcome

From left: the original luminance data, the starless mask and the star mask. You'll note that the formula isn't exactly as I wrote earlier. When I tested it on this image, I found that this formula works even better. The ~ sign means that the inverse of star_mask is used (i.e. 1 - starmask). The Luminance is multiplied by the inverse of the star mask. This means that where there are stars (star_mask has bright pixels with value 1), the luminance mask is multiplied by 0, and where there are no stars (pixel value = 0), the luminance mask is multiplied by 1.

To get best results, you'll probably have to adjust brightness levels in the star mask and in the luminance image from which to subtract.

If you do this process again, but now with a star mask for the smallest stars, you can even get rid of those. Just create a new star mask from the "starless mask". You can also just eliminate the first 3 layers and blur the entire starless mask. Or you can use this starless mask to create a range mask for the nebula. Just experiment.

Good luck

[MARS1960]

Thank you very much Wim.