Page 1 of 1

PixInsight MaskedStretch help

Posted: Wed Aug 04, 2021 10:04 am
by Jockinireland
Hi Folks,

Does anyone know of a way to stop pink cores in bright stars when doing MaskedStretch on a RGB/OSC image

Thanks

David

Re: PixInsight MaskedStretch help

Posted: Wed Aug 04, 2021 11:26 am
by Juno16
Hi David,

Being pretty new to PI, I only have a guess (untried).

You probably already thought about this, but would the PI CorrectMagentaStars script help? Maybe running it on just the stars (Starmask) to the non-stars part of the image is not affected?

Re: PixInsight MaskedStretch help

Posted: Wed Aug 04, 2021 11:50 am
by Jockinireland
Juno16 wrote: Wed Aug 04, 2021 11:26 am Hi David,

Being pretty new to PI, I only have a guess (untried).

You probably already thought about this, but would the PI CorrectMagentaStars script help? Maybe running it on just the stars (Starmask) to the non-stars part of the image is not affected?
Hi Jim,

Hope you are doing well.

Yes I tried that but it left the cores a nasty grey colour which was not much better to be honest. :cry:

Rgds

David.

Re: PixInsight MaskedStretch help

Posted: Thu Aug 05, 2021 8:10 am
by Graeme1858
Would a tweak to the Hue Curve with a star mask in place do the job?

I'm just guessing too!

Regards

Graeme

Re: PixInsight MaskedStretch help

Posted: Thu Aug 05, 2021 11:00 am
by Jockinireland
Graeme1858 wrote: Thu Aug 05, 2021 8:10 am Would a tweak to the Hue Curve with a star mask in place do the job?

I'm just guessing too!

Regards

Graeme
Thanks Graeme - same as with the correct magenta stars it left rather nasty aberrations.

I got a response on the PI Forum about using "repaired HSV seperation2 scrip which worked to an extent although it left pale blue cores instead. :roll:

Rgrds

David.

Re: PixInsight MaskedStretch help

Posted: Thu Aug 05, 2021 1:53 pm
by Juno16
Thanks for passing that information along David.

Good to know, but the results still sound disappointing.

I haven’t had that issue (yet), but I’ve only used the Masked Stretch a few times. I’ve had good success with dragging the STF instance to the Histogram Transformation, but my success is limited since I haven’t imaged in quite some time.

Thanks again for sharing the info and if you have better success with another method, please share with us less experienced folks (me).

Re: PixInsight MaskedStretch help

Posted: Thu Aug 05, 2021 3:45 pm
by STEVE333
Hi David - The real problem sounds like your stars are saturated, maybe even in the RAW frame. For the future it might help to use shorter exposures. Of course that means more images to handle.

I believe the HSV process needs to be applied when the image is still linear (before the MaskedStretch). If you did that then you have done everything correctly. Some bright stars are just hard to deal with!

Cheers - Steve

Re: PixInsight MaskedStretch help

Posted: Thu Aug 05, 2021 8:45 pm
by Jockinireland
STEVE333 wrote: Thu Aug 05, 2021 3:45 pm Hi David - The real problem sounds like your stars are saturated, maybe even in the RAW frame. For the future it might help to use shorter exposures. Of course that means more images to handle.

I believe the HSV process needs to be applied when the image is still linear (before the MaskedStretch). If you did that then you have done everything correctly. Some bright stars are just hard to deal with!

Cheers - Steve
Thanks for that Steve. You are absolutely right, the stars showing pink were indeed saturated right from the raw data. So that's a lesson learned! I had gone back to linear for the HSV.

Thanks again Steve.

Re: PixInsight MaskedStretch help

Posted: Thu Aug 05, 2021 9:23 pm
by STEVE333
Jockinireland wrote: Thu Aug 05, 2021 8:45 pm
STEVE333 wrote: Thu Aug 05, 2021 3:45 pm Hi David - The real problem sounds like your stars are saturated, maybe even in the RAW frame. For the future it might help to use shorter exposures. Of course that means more images to handle.

I believe the HSV process needs to be applied when the image is still linear (before the MaskedStretch). If you did that then you have done everything correctly. Some bright stars are just hard to deal with!

Cheers - Steve
Thanks for that Steve. You are absolutely right, the stars showing pink were indeed saturated right from the raw data. So that's a lesson learned! I had gone back to linear for the HSV.

Thanks again Steve.

You're welcome David. I had the exact same problem when working with my Canon camera. I never did find a good solution.

Now that I've transitioned to NB imaging I haven't run into the problem.

Steve