Help - Don't Know How To Do Post Processing On my Milky Way !!!

[Elp]

13 minutes ago, Chuleton08 said:

No manual focus?

 

Not this particular lens. The majority of my primes are manual however.

[Oddsocks]

This was the quick and rough result I came up with for the Siril output, basically the same tools and processes that Wim used in PixInsight for the DSS file.

Always find it difficult to process Milky Way images of this scale with very few empty background areas to get a reference for gradient removal plus the horizontal banding that always lurks in the images from Canon cameras begins to show as the image is stretched in the dark areas after gradient removal is applied.

Could spend hours on this data set to get a really good result.....

 

[wimvb]

5 minutes ago, Oddsocks said:

Always find it difficult to process Milky Way images of this scale with very few empty background areas to get a reference for gradient removal plus the horizontal banding that always lurks in the images from Canon cameras

Agreed, these are not easy. PixInsight has a CanonBandingRemoval script just for that brand.

I think you lost reds in the two main nebulae. I got that when I used standard colour calibration. Photometric colour calibration worked much better for me as it uses stars as references.

[Elp]

19 minutes ago, Oddsocks said:

horizontal banding that always lurks in the images from Canon cameras

It's a two button press in Siril. You can also remove vertical banding which sometimes helps to smooth the result followed by a second horizontal banding removal.

[Oddsocks]

3 minutes ago, wimvb said:

I think you lost reds in the two main nebulae. I got that when I used standard colour calibration. Photometric colour calibration worked much better for me as it uses stars as references.

Hi Wim.

Good to hear from you....

I did use SpectroPhotometricColorCalibration on that attempt and interestingly got very different results using the standard default Sony sensor filters and the specific filters for Canon 500 series cameras.

The result I posted was with the Canon specific sensor filters selected in SPCC, which produced grey-brown in the major nebula areas but did keep the smaller Ha areas the correct colour, when I tried the default Sony sensor filters the result was overall too pink.

Probably one way to process this is with two versions using both sets of filters and then selective layering in Photoshop/Affinity Photo.Gimp etc.

I find working with monochrome cameras and separate filters so much easier than OSC.

All depends how much time you want to spend....

[Chuleton08]

26 minutes ago, Oddsocks said:

This was the quick and rough result I came up with for the Siril output, basically the same tools and processes that Wim used in PixInsight for the DSS file.

Always find it difficult to process Milky Way images of this scale with very few empty background areas to get a reference for gradient removal plus the horizontal banding that always lurks in the images from Canon cameras begins to show as the image is stretched in the dark areas after gradient removal is applied.

Could spend hours on this data set to get a really good result.....

 

Thats Nice. Why is there weird stuff happening in the bottom side? Is it not enough data when stacking? 

[Elp]

9 minutes ago, Oddsocks said:

I find working with monochrome cameras and separate filters so much easier than OSC

OSC is the worst in LP areas, maybe not so much emission with a narrowband filter, but LRGB..., why I also mostly do mono.

[Oddsocks]

Just now, Chuleton08 said:

Thats Nice. Why is there weird stuff happening in the bottom side? Is it not enough data when stacking? 

I don't have any reference images of this part of the border between Scutum and Serpens to compare with but looking at the raw stacked outputs from Siril and DSS I suspect you have some stray light contaminating that part of the field, possibly the moon, or light pollution etc., but stacking artefacts will also create a fuzzy glow in parts of the field where star matching is poor.

More data is always good because then you can selectively reject the poorest subs from the stack and only combine the best frames to give a clean image to begin post processing.

There are many tools in post processing to cleanup a poor stack but it is always better to have the best quality starting image possible before beginning to stretch and enhance the image.

[Chuleton08]

Just now, Oddsocks said:

I don't have any reference images of this part of the border between Scutum and Serpens to compare with but looking at the raw stacked outputs from Siril and DSS I suspect you have some stray light contaminating that part of the field, possibly the moon, or light pollution etc., but stacking artefacts will also create a fuzzy glow in parts of the field where star matching is poor.

More data is always good because then you can selectively reject the poorest subs from the stack and only combine the best frames to give a clean image to begin post processing.

There are many tools in post processing to cleanup a poor stack but it is always better to have the best quality starting image possible before beginning to stretch and enhance the image.

Hmm, Ok. Thanks!

[Oddsocks]

14 minutes ago, Chuleton08 said:

Hmm, Ok. Thanks!

If your Canon camera is one of the models that still has the separate viewfinder port cover attached to the camera strap then use it when taking long-time exposures to prevent stray light leaking into the mirror box, which can create a faint glow right around the periphery of the image.

Once the image is framed turn off the camera back display panel, which is another source of stray light (in the form heat) that will appear as a glow in the image.

When using a standard lens fit a deep hood to keep local light pollution from entering the front of the lens, anything will do, bit of rolled up plastic sheet covered in black fabric - or black waxed cardboard etc., and secured with an elastic band... is better than nothing.

Get away from local light pollution to somewhere really dark to grab the data.

[wimvb]

This is as far as I can push the Siril image. The problem with dslr cameras is that the images suffer from what Tony Halas calls "colour mottle", large scale colour noise. This gives a blotchy look if you start saturating colours.

Anyway, this is with all the bells and whistles in Pixinsight. This time I didn't remove the stars.

According to the fits header of your image, there were 17 x 28 s exposures in the stack. That is just about 8 minutes total integration time. For such a short time, this isn't bad. Is your camera astro-modded (ie modified to enhance sensitivity in the red especially for H-alpha)?

Edited Thursday at 22:26 by wimvb

[wimvb]

A few interesting areas in this image. I platesolved and annotated in Pixinsight. Only Messier objects and constellation lines

 

Edited Thursday at 21:31 by wimvb

[Chuleton08]

That's Nice! Yes, It was a very short integration time. Had a short night, and had to delete lots of images due to objects in the foreground. And my camera isnt astro modded (Don't want to risk anything). That's the part of the milky way I wanted