Over Saturated Stars

[jerryyyyy]

I know this must seem like a basic question, but with my Takahashi 180ED I have cut down the exposure times to between 100-130s for RGB images.  But I am getting what looks like over saturation.  Should I cut exposure down more, or is this a processing issue?  Here are some cropped parts of M42.  I do not want to lose signal in the dark areas either....

[gvanhau]

Hello Jerryyyy

It is difficult to say, if it is a processing issue since you dont tell anything about the steps followed.
You should inspect your frames while they are still linear and see if the pixel values are near the highest value admited by you sensor/CCD or near 1 if they are normalized to the 0-1 range.

The way to go if you want to preserve/show high and low signals together is doing HDR.  This mean in some cases taking long exposures for the low signal areas and short exposures for the high signal areas and then combine them using the HDR tool.

Geert

[jkmorse]

Even saturated, your stars should not be coming out with those kinds of artifacts unless there is a problem in color balance.  Just from appearances, it looks like you are missing red signal in those areas to offset the saturated blue and green.  I do not think it is a timing issue.  I have shot 5min+ images with fast f2 lenses in the past without this problem and I have seen images shot with a180ED as long as 15 minutes without these types of artifacts.  Longer exposures will lead to more saturation on the brightest stars but shouldn't create these types of artifacts. 

Did you try using LinearFit and/or Background neutralization and ColorCalibration to make sure your colors are balanced?

Jim

[RickS]

Hi Jerry,

You might find the 3dplot script gives you a useful view of the star profiles in your data: make a small preview with a few of the bright stars, click on the preview to display it, then run the script.  Start with the linear data and you'll see how badly saturated your stars are.

Cheers,
Rick.

[jerryyyyy]

Thanks for all the comments. 

I looked at the histogram on the blues in linear form and they are not completely saturated.  Enclosed is a screenshot of the target (M42) with the initial stretch.... you can see all the colors.

The workflow if pretty simple:

SubframeSelector
BatchProcessing (about 20 images apiece)
Crop
DBE
Linear Fit
LRGB to combine the RGB
Then STF->HT stretch

[topboxman]

I am not exactly sure that LinearFIT is a good tool for color calibration. I do not see Background Neutralization and Color Calibration in your work flow after RGB combine. Your histogram does not appear to show well calibrated or balanced colors (RGB do not line up).

Also, I never apply STF to HT. I like to do HT in iterative process (mid slider and then black-point (left) slider and repeat) until the image looks good. In my case, most of the time STF is too aggressively over stretched.

Peter

[Alejandro Tombolini]

Hi Jerry, check your image before and after Deconvolution.

Also see this example that may be useful

Saludos, Alejandro. 

[vicent_peris]

Hi,

You need to increase the smoothnes parameter in HDRComposition.

Best regards,
Vicent.

[jerryyyyy]

Hi,

You need to increase the smoothness parameter in HDRComposition.

Best regards,
Vicent.

Thanks, this seems to be the key.  Basically I think I was right that in some of the stars to cores are over-saturated.  The idea of the HDRComposition is to replace the overexposed areas with data from less exposed shots of the same target.... just like HDR in a DSLR with bracketed shots. 

This is showing up in the Blues for me I think because I am shooting them at 130s and the Red and Green at 100s.  Somewhere along the line I read that Blues had to be exposed more because their signal is weaker than the other two.  So, maybe this does not really apply to my new fast F/2.8 scope and I better just crank them all back to 100s, or I can use these "longer" exposures to bring out colors in the faint nebula but shoot some shorter Blues and use the HDRComposition tool to correct the stars. 

I found this thread:

http://pixinsight.com/forum/index.php?topic=4850.msg43913#msg43913

I am very happy that the target I found this problem in is exactly the same one used in the tutorials (M42).  The Blue Stars are very bright here and accent the problem, but I have seen it in other images with Red stars too. 

Seems like the PI tutorials still lack audio but Harry has a basic one with audio.

Probably will take me a week to understand all this...

Thanks again...

[jkmorse]

Jerryyyy,

It also depends on the type of filter you use.  If you happen to be using Astrodon Gen2 e series, those are very close to 1:1:1 in RGB so no need to do offsets.

Best,

Jim

[jerryyyyy]

Have made progress... not sure everyone will like the h-alpha enhancement:

[jkmorse]

Nice!