MaskedStretch and hard edge stars

[joelshort]

I've been playing around with the MaskedStretch process and have noticed that on bright stars the edges of the stars are distinct and "hard".  See attached (master synthetic LUM image).  Granted this is with a very bright star that is saturated (K=.96).  Is there a way to adjust the MaskedStretch settings so this hard edge doesn't happen?

I've been able to mitigate the effect somewhat by reducing the amount of MaskedStretch (lowering the target background to .09000) and then using CurvesTransformation to stretch the image a little more.

[Dimitris Platis]

U can chck this out
http://pixinsight.com/forum/index.php?topic=7348.msg49626#msg49626

or perform a convolution with a star mask

[joelshort]

That looks promising Dimitris.  Thanks for the link.

[lucchett]

Hi,
one of the benefits of MS should be a smoother star profile compared to something like HT.
my personal experience is that the choice of the background value is very important.

I follow the advice suggested by Vicent in one of the old processing examples.

I do a normal HT until I reach the stretch I am looking for.
I take note of the statistical values of the background and use this as a parameter for the MS on the linear image.

hope it helps
Andrea

[joelshort]

Thanks Andrea.  One follow up question, do you use the MS just once?  Or do you apply MS to the same image more than one time?  And is the MS stretching sufficient or is it necessary to follow it up with a light HT?

[Alejandro Tombolini]

Hi Joel,

That depend on your image. With the previous Masked stretch script that allowed only 200 iterations I used to use several application for faint nebulaes. With the new Masked Stretch Module you can set until 1000 iteration that is enough in most of the cases.
You can use MS alone or together with other intensity tranformations to reach the desired level of stretch.

Saludos, Alejandro.

[joelshort]

OK. I'm finding this script particularly helpful on widefield images taken with camera lenses in which the stars tend to easily bloat. A few iterations of the script seems to work better in those cases.

[jkmorse]

Joel,

Also take a look at Harry's astroshed.  He has a great process of getting the most out of combining both masked stretch and HT.

Best,

Jim