6 replies to this topic

[Wildetelescope]

New to PI and have been playing around with the software, in particular the Masked Stretch.   Below is an image of Alnitak where I have done the following

 

1. Calibration, Registration and Integration

2. Cosmetic correction, DBE, Background Neutralization, and Color correction

3. Using EZ suite, Deconvolution, Denoise

4. And finally Masked Stretch

 

As you can see, all my saturated stars have little (or not so little) pink centers.   I understand this is due to the mismatch in the peak values of the RGB channels, but not sure how to fix it.  I did try the HSV repair script as a remedy which I saw was suggested on the PI forum, but that just gave me blue centers. I did the HSV repair scripted between step 3 and 4 above, perhaps that needs to be done earlier?   Appreciate any suggestions that the group can provide.  I am excited about the rest of the result!  

 

 

Cheers!

 

JMD

[maxsid]

Use StarNet to remove the stars and process the starless image separately and then add stars back.

[the Elf]

Two solutions to this.

a) Use the highlight slider in HistorgramTransformation to saturate all three channels equally and the center will turn out white

b) Use a range mask, blown up by morphologic dilation and smoothed by convolution, desaturate the bright star cores using curves "S"

 

If you are using a DSLR bright parts of the image often saturate but only in one or two channels, casing wrong color. That happens in daylight photography with clouds. Raw converters detect this saturation of one or two channels and correct the colors to white. If you naively process a raw image with bright clouds they turn out pink.

[Bretw01]

Two solutions to this.

a) Use the highlight slider in HistorgramTransformation to saturate all three channels equally and the center will turn out white

Yes, do this before applying the masked stretch. Use the preview before applying and watch for the stars to turn white.

[Wildetelescope]

Thanks Everyone!   I will try these approaches.  Having fun learning what this program can do, and learning what each step does to the signal.   

 

cheers!

 

JMD

[Wildetelescope]

Two solutions to this.

a) Use the highlight slider in HistorgramTransformation to saturate all three channels equally and the center will turn out white

b) Use a range mask, blown up by morphologic dilation and smoothed by convolution, desaturate the bright star cores using curves "S"

 

If you are using a DSLR bright parts of the image often saturate but only in one or two channels, casing wrong color. That happens in daylight photography with clouds. Raw converters detect this saturation of one or two channels and correct the colors to white. If you naively process a raw image with bright clouds they turn out pink.

Approach 1 has worked well.  I am using an QHY 268 C but I assume something similar is happening to shift the saturation levels of the different channels.  Interestingly it seems to occur, or become evident, only after I do the color calibration in PI. After CC, sure enough the bright saturated stars are pink.  I have not seen this effect before with other software I have used, which may be because I have never tried masked stretching before.  Linear stretching just pushes everything past saturation.  It is fascinating to see how the signal is mashed and twisted by each of the processes.  

 

Thanks!

 

JMD 

[Psychlist1972]

Processing without stars has been a bit of a revelation. I love being able to stretch and adjust the nebula without messing up the stars.

 

Starnet requires you do a basic stretch first, but you can do just enough, remove the stars, and then continue processing the nebua.

 

Pete