VDB 152 and weird stars

[Jules]

Hi

I have not posted for a while, I have had problems with guiding on my telescope which now looks like I am on my way to resolving. VDB 152 was a test for my guiding and so some of the stars a bit ovaloid still.

My processing routine in PI:

Combine RGB

Extract Lab from RGB recombine with new Lum, this is where I see the blue artifacts around the larger stars. No amount of clone stamping gets rid of this artifact to my satisfaction.

Julian

[Juan Conejero]

Two questions:

Why don't you use the LRGBCombination tool?

Are you working with linear (unstretched) data?

[Jules]

Hi Juan

Yes I am working with unstretched linear data.

I think I get smaller stars with this method along with using masked stretch?

[Juan Conejero]

Hi Jules,

The CIE Lab color space is nonlinear, so it makes no sense for linear images. Those artifacts around bright image features denote numerical overflow in the Lab -> RGB conversion, and happen because the chroma and lightness components are not coherent.     

I think I get smaller stars with this method along with using masked stretch?

Not at all. The correct procedure for LRGB combination is:

- Stretch the RGB and L images. You can use any combination of tools to perform this task, including HistogramTransformation and the MaskedStretch script.

- Use the LRGBCombination tool to replace the implicit lightness in the RGB image with the separate L image.

The RGB and L images must be well equalized before LRGB combination. If you haven't already done so, I strongly recommend you watch these videos:

http://pixinsight.com/videos/NGC1808LRGB-vperis/en.html

[pfile]

actually i see those artifacts on bright stars as the result of using the masked stretch script. gets worse if deconvolution was applied without masking the cores of bright stars.

[Jules]

Thanks Juan for correcting me, I can see your point with respect to CIE Lab colour space in linear images.

Pfile, I see that as well, I have been trying to implement masked stretch in my processing routine as it does keep the stars form becoming behemoths.

Julian

[Juan Conejero]

actually i see those artifacts on bright stars as the result of using the masked stretch script.

One of the priority tasks that we'll have in September is releasing a tool version of the MaskedStretch script, since we consider it a fundamental processing algorithm. These problems should be fixed in the new implementation.

[Jules]

Juan

Fantastic, I look forward to using it.

Julian

[pfile]

actually i see those artifacts on bright stars as the result of using the masked stretch script.

One of the priority tasks that we'll have in September is releasing a tool version of the MaskedStretch script, since we consider it a fundamental processing algorithm. These problems should be fixed in the new implementation.

excellent, thanks.