HDR to Control Blooming

[mads0100]

Vincent,

    Interesting, I think I stumbled through what you said.  However, there are cases where it doesn't work.  For instance, I was working on a copy of the Horsehead Nebula (with Alnitak), and I couldn't increase the mask significantly enough to reduce it's effects on the image.  My other option was to reduce the detail provided by my longer exposures to an almost unusable level.

Could you do a quick Juan style tutorial on how to do it?  I'd like to see how you're manipulating the settings to make it happen.

Thanks,
Chris

[mads0100]

For the mask output and then TGV, do I apply the mask to the image and then use TGV to remove the 'noise' around the stars?

Thanks,
Chris

[mads0100]

Bump.

[mads0100]

Id still like to see a tutorial demonstrating HDRComposition and deblooming. It's your purported method of correction, and it's not somethung I can figure out. I'd gladly give you some data to play with if thats what you need.

Chris

[dgbarar]

Hi Chris,

I just gave Vicent's method today.  Works well.  Here is what I did.

1) Apply the HDR mask to the HDR image that has had the blooms removed.
2) Open TGVDN.
3) Set the Edge Protection to the standard deviation of a section of the HDR image that is background.
4) For the strength I found I needed to use a fairly high value.  The default value of 5 worked well for me.

In most instances, I had removal of noise where the blooms used to be.  In one instance, I had a little noise left around the periphery.  I had to go back and redo the HDR image with larger mask growth.

Hope this helps.

Cheers,

Don

[mads0100]

Thanks Don, I guess I'm still having trouble getting rid of all my blooms using HDRComposition.  At some points I'm nearly at 0.01 for my binerizing threshold.  What settings are you using?

[dgbarar]

Hi Chris,

In my most recent deblooming effort  I used the following parameters:

Binarizing Threshold = 0.6
Mask Smoothness = 5
Mask Growth = 7

These settings were for very small blooms.  I am working on a new image where the blooms are significantly larger.  Have yet to capture all the data.  I'll let you know how it works out and what settings I use for more severe blooming.

Cheers,

Don

[mads0100]

Yea, I can get around the small blooms pretty well.  It's the big ones, like Alnitak in a flame nebula picture, that I'd like to get rid of. 

Have you seen their video tutorials in the processing examples sub-forum?  They have one video that applies to de-blooming.  It doesn't help the 'big blooms' but it works pretty good for the small stuff.

[dgbarar]

Hi Chris,

Are you the video tutorials found here:http://pixinsight.com/forum/index.php?topic=4850.0

If so yes.  But without the audio it is difficult to determine why some actions are being taken and how they are being applied.  There was a concept of using Pixel Math to do some blending that I am still uncertain as to what is going on.  Maybe I should take another go at it.

Cheers,

Don

[mads0100]

Yea, I was using the videos there.  It works pretty well as long as the blooming is light.  What I really want to see is some examples like their original 'tutorial' that was a PPT on how PI could be used to remove blooming.  It had blooms half way across the image and they showed how it could remove them. 

But, I don't seem to get any response from them anymore.  I guess they feel the problem solved.

[dgbarar]

Hi Chris,

I think they have other "fish to fry" at the moment, i.e. a new TGV tool.  That said, I wish some audio could be added to the tutorials.  Or, better still, a detailed step-by-step example that shows how they accomplished large bloom correction.

Don

[mads0100]

Maybe, they've been telling people you could do this for ages.  Maybe Warren and his crew would step in to make a video demonstrating significant blooming removal.