MaskedStrechTransform

[Jordi Gallego]

Hola José Manuel,  (english below)
 
llevo unos dias muy desconectado, lamento no haberte respondido pero veo que ya te han orientado bien. Yo utilizo en algunas imágenes MST y dependiendo de la imagen va muy bien. Creo que es importante, como ya te han recomendado antes de aplicarlo recortar las sombras.

Para todos: este es el post donde Carlos Sonnestein explicaba su funcionamiento:

http://pixinsight.com/forum/index.php?topic=510.0

y otro ejemplo de aplicación se puede ver en las páginas 35 a 39 de esta presentación (ojo son 6 MB!):

http://astrosurf.com/jordigallego/articles/Tecnica_RGB_banda_ancha_JGallego.pdf

(Está en castellano pero las imágenes creo que se entienden perfectamente)

Un saludo
Jordi


Hi all,

Regarding MST, this is original post from Carlos Sonnestein explaining MST:

http://pixinsight.com/forum/index.php?topic=510.0

and you can see another example "before-after" MST aplication on pages 35 to 39 of this presentation (be careful it is 6MB!)

http://astrosurf.com/jordigallego/articles/Tecnica_RGB_banda_ancha_JGallego.pdf

(it is in spanish but I think that images are self-explanatory)

Best regards
Jordi

[José Manuel Pérez Redondo]

Gracias Jordi

Ya había leído tu documento, así cómo el de Carlos.

Thanks Jordi,

I have already read your document and the Carlos' too

José Manuel

[dhalliday]

I used this again recently...and I am still not sure it isn't just a nice easy way to do a stretch..
Not that I am complaining...(!!)
Someone expound briefly on how much better this is..?
Of course I often do a much cruder stretch myself ...mostly to darken my sky..but I am losing the fainter bits of signal every time I do a "step"...because it is not as well masked as the stronger data...

Does this process prevent that "creeping" loss of the fainter signal..and how ??

Dave

[Silvercup]

Hi Dave:

As Jordi said in his paper MST Script works really well and its main virtue is stellar profile improved by reducing stars size and with a single defect of decreasing the contrast of the image in comparation with a single Histogran Transfromation.

You can clearly see the difference on page 38 "Estirado Iterativo" with MaskedStrechTransform script and "Estirado Unico" as normal Histogram transformation.

On page 39 you can see the difference in the stellar profile. The stars are much smaller and not become saturated as well as with a single histogram strech.

With regard to background noise, whether you use the script or a single stretch you will have the same problem. If you do not have enough SNR the noise is going to "eat" faint details and cutting shadows with histogram is not the solution. The solution is to apply noise reduction algorithms where this exists and trying to protect faint details.

Best. Silvercup.

[dhalliday]

Silvercup
Sorry...do you have a link to that document ?
? English translation..?
So regarding NR...when do you use GreyCstaration...do you use it with a mask..?
It seems to eat data..
Before/after ACDNR..?
It depends..?

thanks!!
Dave

[Silvercup]

Hi Dave:

The document is http://astrosurf.com/jordigallego/articles/Tecnica_RGB_banda_ancha_JGallego.pdf. Images are self-explanatory.
I don't use GreyCstoration because is intended to diurnal images. When I have applied to deep sky images I get a lot of artifacts.

You must use ACDNR with is "built in" mask generation and checking Luminance Mask. And, of course, you can always add a standart mask to reduce ACDNR effects.

This is an application example:

http://www.pixinsight.com/examples/deconvolution/Gemini-NGC5189/en.html#A12

You must build a mask that applies to background noise and fainter and noiser parts of the object. It is cuestion of trial an error with previews till you correct fainter parts without loosing significant structures (or colors, It can happen) and the background whole corrected.

Depending on image you must be more agressive with Chrominace and less with luminance.
 
Best. Silvercup