Gamma Cygnus

[andrewluck]

Has anyone any ideas for reducing the glow around the central star in this image using either Photoshop or PixInsight? Or, do I need to do something different when aquiring the images. This is a stack of 10x 10 minute frames using a QHY9C on a Zenithstar 70.

Andrew

[bogdog]

impressive picture,i downloaded and used a programme called paintsupreme it had an auto contrast option to apply to picture

results below,so maybe contrast is what you need to fiddle with in photoshop (thats if it looks like the result you want)

http://i44.tinypic.com/2jdqe87.jpg

[Cloudwatcher]

Has anyone any ideas for reducing the glow around the central star in this image using either Photoshop or PixInsight? O

Andrew

You could try using the Gamma Correction slider in the Exposure tool of Photoshop to reduce the 'glow' around the central star, on separate layer of course and then merge when happy with the level of reduction.

I also think bogdog is right with regard to having a fiddle with the contrast.........all a matter of personal preference of course.

[andrewluck]

Not an easy one. The star is quite nicely reduced but the overall effect is not really what I'm aiming for. The subtle shading in the darker regions has been cut by the contrast increase i.e. darks are now too dark, and the overall effect is just too harsh for my taste. Of course, a lot also depends upon the monitor used for display. I'm one of those still using a CRT which reproduces colour very differently from an LCD. However, having looked at your images I think I should go a little darker as the darkest areas on my image are about 60. I've used the levels command to darken the shadows and increased the mid-tones to compensate. The result has rather more impact.

[michael.h.f.wilkinson]

That last processing is the best so far, I think. Some compromise between cloudwatcher's and andrewluck's might be even better

[martin_h]

How's about this version?

In PS bring down the black points to get a better black point, make a copy and lasso (with a feather) the star and halo use curves to bring down the brightness blend the two layers...

[andrewluck]

After moving the blackpoint the midtones probably require lightening a bit to brighten the image as it looks a little dark on my monitor.

The star has gained a bluish halo with the curves processing. Maybe that could be reduced by a smaller lasso and curves again. The outer areas are improved though.

[ollypenrice]

Here's how I do bright stars.

1) Make the normal image as you have done.

2) Go back to the linear data (in LRGB shooting I'd use the RGB only to keep the bloat down.) Do a star curve stretch (AKA a Core Control Curve) in which the lower left of the curve rises steeply as in the normal image but which flattens off very early. The idea is to get the background sky similar to the original image while keeping the bright star down. Pay attention only to the bright star and its immediate background. Get these as good as you can, ignoring the rest of the image which will look awful! Here's an example using Alnitak.

3) Paste the full stretch onto the core control image and use a well feathered eraser a bit bigger than the halo to erase, partially, the top layer. Progressively reduce the size of the eraser and keep erasing, letting more and more of the bottom layer through as you near the core of the star.

4) Colour balance. Because even the best apos are not apo on very bright stars I go to the star charts to check the spectral class of the star in question. My own Sadr data wanted Sadr to be blue but I know it's not so I balanced the colour of the bottom layer until it looked reasonable. (The data would have it blue because blue suffers more dispersion in the optics and creates a blue halo.) By adjusting Saturation, colour and Curves in the bottom layer while seeing it through the top layer you can get a pretty seamless balance.

To do a decent job you have to go back to the linear data. You can pull the region of the bloat down a little inside a feathered marquee as below, but back to linear is best. If you try this method it is vital to pin the curve at the good background level closest to the halo.

I decided to leave Sadr quite large because it defines the region;

[andrewluck]

Thanks for posting that Olly. Hopefully I'll get some time tomorrow to give it a go. It'll give the new 6 core AMD PC something to work on . Andrew

[andrewluck]

The bit I'm having trouble with is matching the background level. This is resulting in a dark hole in the middle of the picture. More experimentation required I think. Anyway, this is as far as I've got. The image is stretched rather more aggressively but I've kept Sadr at about the same brightness as before at the expense of some ringing in the glow.

There's some residual light pollution that's very tricky to remove as there's not really much sky background in this image.

Andrew