NGC6946 & 6939 with PixInsight SE

[enzosantin]

Hi all,
I have processed another image with PI SE beta 3: NGC6946 + 6939.
Image taken at Piano Battaglia, Madonie Mountains, Sicilia, Italia.
Tec data: Pentax apo refractor SDHF 75 f/6,7 + SBIG ST8E + SBIG CFW8 on a Takahashi EM10 mount. Exposure time was L:R:G:B = 80:20:20:40
The link:

www.enzosantin-astro.com/pagine_immagini/HiRes/NGC6946_6939.htm

I'm worked with ATrousWaveletTransform and DBE

Best regards,

Enzo Santin

[Juan Conejero]

Hi Enzo,

What a beautiful image. The composition is fantastic, with the two main objects creating tension on a diagonal.

The only aspect in which I think this image could be improved is the stars. I refer to large stars especially. I think they should have more Gaussian-like profiles, and more color, which would be a great improvement in my opinion.

There are a number of reasons why those stars look flat. How are you doing the initial stretch of raw data? This step is extremely important. If you are doing some DDP processing, this may be the cause.

Also, how are you doing the LRGB combination? I'm sure that if you try to perform these critical steps in PixInsight, using HistogramTransform, CurvesTransform and LRGBCombination, you'll be surprised  

Of course this is just a small note on what is definitely a fine image. Thanks for sharing it with us.

[enzosantin]

Thank you Juan,
you are right! I've made a DDP processing and I'm not very happy with these stars too flat. I will try to follow your sggestions.  I try to perform initial stretching only with HistogramTransform (no DDP) and a LRGB combination. Can I post the results for comparison?

Thank you again,

Enzo

[Juan Conejero]

Hi Enzo,

Absolutely yes. I think such a comparison could be very interesting.

It's no secret that we are firm detractors of DDP. It is a simplistic solution to a problem (nonlinear stretch of linear raw data) for which we have much more efficient and sophisticated solutions.

Regards,

[enzosantin]

Hi Juan, hi all,
I have reprocessed my image of ngc 6946 + ngc 6939 following Juan's advices.
Here the links:

old version:

http://www.enzosantin-astro.com/pagine_immagini/HiRes/NGC6946_6939_old.htm

new version:

http://www.enzosantin-astro.com/pagine_immagini/HiRes/NGC6946_6939_new.htm

I have processed the new version with histogram and/or curves stretching only= NO DDP stretching.

Because the stars was affected by large halos, I have corrected the halos creating "halos masks" with the wavelet transform, by isolating the appropriate layer scale in which the halos were evident and then applyng curves or histogram transform to the image masked with these "halo masks".

The same for bloated stars, but creating "stars mask" and applyng curves with the mask active. Thanks for looking.

Enzo Santin

[Carlos Milovic]

Hi Enzo

You are doing a very good job Congratulations. Keep those fine images coming.
 



May I ask you something? Wich tools/processes do you use in your processing workflow that are not implemented in PI? We know that image calibration (bias+darks+flats) is one of our flaws right now...

[enzosantin]

Hi Carlos,
thanks for your compliment!

I calibrate the images with MaxIm DL 4.0 or in Mira Pro.

Then I make alignment in MiraPro and this is a critical step for me: I know that also in PixInsight PE there is a very effective tool for aligning images, but I have too much images (more then 50!) for aligning only two image at once. It would be useful to align one entire "stack" of images also in PixInsight.

Then I "normalize", make data rejection and stack the images in CCDStack.

When I have a luminance and a RGB image, I work in PixInsight with great satisfation, but I would like some kind of "free hand"/"lasso" tool with feathering for "correcting/masking/excluding" some zone of the mask (I make this in Photoshop), because obviously not all zone of the image need the same kind of stretching/processing: for example the galaxy and the open cluster in my last image.

Sometimes, but rarely, I make some deconvolution in my images with AstroArt.

Sorry for my bad english, and thank you again!

Enzo