HDR to Control Blooming

[dgbarar]

I use an ST10 XME/AP 155 EDF and getting ready to begin my first HDR compositions to control blooming on brighter stars.  Does anyone have any guideliines they use in taking the non-bloomed light frames.  For example, answers to the following might be helpful:

1) If total imaging time were 8 hours for the bloomed light frames, how much imaging time should I devote to the non-bloomed frames so they blend well in the end?
2) Relatively speaking, how long should I expose non-bloomed frames, i.e.  75% of full well, 50% of full well.

If anyone can think of anything else I should consider before beginning this project that would very helpful.

Don

[avastro]

Hi Don
I've the same camera and an AP 130EDF,so I take most of my image with HDR in mind.
I take long exposure to capture the details and short exposure to replace the stars or other areas bloomed.
As the goal is just to capture the stars not bloomed I made some test to define the optimal exposure time to arrive at 20 to 30000 ADU around 30to 40% full well, it's usually in the range of 10 to 60 seconds, then I expose the 3 filters RGB ( considering an RGB image), total time is quite short 20 to 30mn by filter just enough to capture the stars with good colors.
I then do the individual color pane for short and long exposure ( calibrate register and integration), alignment short and long exposure, and then HDR for each filter. HDR work great and it's easy using binarizing threshold  to find the value to replace saturated pixels.
With the HDR pane you then process your image as usual.
You can made a short test before starting your project just exposing an area with some bright stars and compose and HDR image.

Antoine

[dgbarar]

Hi Antione,

Thank you so very much for your advice.  I was surprised to learn that your overall exposures time was so short.  I will use your information as a starting point for my first HDR exposures.

Thanks again so very much for your help.

Cheers,

Don

[mads0100]

Hi Don
I've the same camera and an AP 130EDF,so I take most of my image with HDR in mind.
I take long exposure to capture the details and short exposure to replace the stars or other areas bloomed.
As the goal is just to capture the stars not bloomed I made some test to define the optimal exposure time to arrive at 20 to 30000 ADU around 30to 40% full well, it's usually in the range of 10 to 60 seconds, then I expose the 3 filters RGB ( considering an RGB image), total time is quite short 20 to 30mn by filter just enough to capture the stars with good colors.
I then do the individual color pane for short and long exposure ( calibrate register and integration), alignment short and long exposure, and then HDR for each filter. HDR work great and it's easy using binarizing threshold  to find the value to replace saturated pixels.
With the HDR pane you then process your image as usual.
You can made a short test before starting your project just exposing an area with some bright stars and compose and HDR image.

Antoine

Antoine,

     Can you explain how you use the binarizing threshold to find the value to replace?  I'm having a horrible time trying to combine the two.  It just seems like I can't get rid of the blooms!

Chris

[dgbarar]

Hi Chris,

It can be difficult to optimize the parameters.  Have you attempted to use "Mask Smoothness" and "Mask Growth"?  I have found I need to make significant increases to the "Mask Growth" parameter when the blooms are large.

Of course, when I do this I can end up with noisy areas around those stars with blooms.  I have been fixing these in Photoshop by using the blur tool.

Hope this helps.  If anyone else has thoughts please comment.

Don

[mads0100]

I've found the same thing.  I was hoping Carlos would see this and post how he used it to fix his blooming images.  And, in some cases, I can't clean them up using the tool despite having no blooms around the stars.  It seems to 'ignore' the big ones.

[avastro]

Hi Don,
Removing bloom with HDR works sometimes depending of the severity of the blooms,finding the right setting of binarizing threshold,mask smoothness and mask growth is quite tedious.
I've not found a way who work fine everytime, I also tested different alternative solution but the result are not satisfactory.
sorry, 
Antoine

[mads0100]

yea, I was hoping Carlos would post in here about how he did it using his research photos. There were some significant blooms in the that he managed to remove. I haven't even remotely had that kind of success using HDRComposition.

Please Carlos! There seems to be enough of us now

[Carlos Milovic]

Hi guys
Sorry for the disappearance, but Vicent and I were dictating a PI workshop here in Chile.

About the HDR solution for blooming, Vicent performed that several times. I remember doing some experiments, but I don't know were to find those pictures at all. I'll tell Vicent to reply here.

[mads0100]

Carlos,

  Thank you!

[mads0100]

On some of my pictures I've had to reduce the binarize component to near .03 to make it remove the blooms. Is that how it's supposed to work?

[mads0100]

*bump*

[dgbarar]

Hi M,

I totally missed this.  What do you mean by *bump*?

Don Barar

[Andres.Pozo]

Hi M,

I totally missed this.  What do you mean by *bump*?

Don Barar

http://whirlpool.net.au/wiki/wp_threadbumping

[vicent_peris]

Hi,

It's quite easy to fix the blooms with HDRC. Just increase the mask growth parameter because usually the bloomings have non saturated borders (or even dark halos in some cases). After putting your short exposures on the longer ones, you'll get the bright stars sorrounded by areas of noise. Just activate the "Output composition masks" option and use these masks to reduce the noise in those areas just after the composition. I recommend you to use TGV for this purpouse.

We are introducing some improvements in this tool in the near future. For the moment it can be a problem if you have gradients in the images.


Regards,
Vicent.