9 replies to this topic

[oneredpanther]

Well, that month of cloudy skies and incessant rain finally lifted and I've been making strides at imaging again! 

 

About 11 months into the hobby and I'm converging on a style that I'm proud of. Potentially this is my best work and certainly in a similar ballpark to my recent M42 and Rosette. 

 

Such a shame that the nebula season is almost at a close, but quite honestly I'm exhausted by astronomy and could do with a little mental break for a while. I'm sure we can all identify with that burnout feeling where you eat, sleep, live, breathe this insane hobby for months at a time.

 

Eventually it starts to encroach upon your dreams, and then it's probably time to get it a rest. 

 

_________________

 

 

Here I've captured 25hours on another of this season's must-haves; the California Nebula!

 

About 8hours each on the the Antlia 3nm Pro filters, using a mix of 5-10 minute subs on the 1600MM Pro at -20'C.

 

Stacking and gradient removal in APP, then finished and balanced in Photoshop. 

 

• Here is the uncompressed full-resolution version.

 

Two personal developments on this project that have upped my game: 

 

1. Using HaSHO instead of SHO significantly improved my image quality. 

 

One thing I've developed recently as part of my 'personal style' is using almost exclusively Ha as luminance, then SHO as RGB.

 

Previously when using pure SHO, you would find that the image was terribly noisy; even with 10-20 hours of integration. By offloading the luminance into the strong hydrogen layer, I can significantly clean up the image. I find therefore that HaSHO is much better looking than SHO. 

 

2. A novel (?) method of ultra-clean star reduction. 

 

Another technique I'm working on is how to improve star reduction. I think a lot of tutorials have it wrong, in my opinion. Most people will tell you to select the stars then make them smaller with a 'minimum' (or similar) filter; but this leaves terrible artefacts. 

 

My approach has been to generate a pair of Starfull / Starless versions of the Ha image; then create a star-mask from the unstretched Oxygen layer; before using that to mask the 'starfull' Ha image back in. 

 

So instead of messily selecting stars in an image and making them smaller with bad artefacts - I just take an image with no stars at all; and then use the low-nebulosity oxygen layer as a mask to add a very clean and precise mask of them back to the starless image. 

 

This 'inside out' method is muuuuuuuuuuch cleaner than any other method I've seen described, and I couldn't find anyone else doing it this way. 

 

Maybe I can make a quick tutorial of my own in case anyone's interested? 

 

 

Thanks for looking!

 

_________________

 

• More of my images on www.instagram.com/baz_astra

_________________

 

 

TECHNICAL STUFF:

 

Spectral Palette: HaSHO (Ha luminance + SHO)

 

Exposures:

102 x 300s Hydrogen (8.5hrs)
52 x 600s Oxygen (8.7hrs)
100 X 300s Sulphur (8.3hrs)
 

= TOTAL 25.6hrs

 

Camera: ASI 1600MM Pro @ -20'c
Mount: Skywatcher HEQ5 Pro
Telescope: William Optics Redcat 51
Filters: Antlia 3nm SHO Pro narrowband
Guiding: ZWO ASIAIR
Software: APP + Photoshop
Sky type: Bortle 4
Location: Norfolk, England

Attached Thumbnails

• 

Edited by oneredpanther, 18 January 2022 - 09:55 AM.

[bobharmony]

That is an amazingly detailed California!  Thanks for the overview of your process, it produces a wonderful result.

 

Bob

[siriusandthepup]

Wow - beautiful! 

[ChiTownXring]

Well, that month of cloudy skies and incessant rain finally lifted and I've been making strides at imaging again! 

 

About 11 months into the hobby and I'm converging on a style that I'm proud of. Potentially this is my best work and certainly in a similar ballpark to my recent M42 and Rosette. 

 

Such a shame that the nebula season is almost at a close, but quite honestly I'm exhausted by astronomy and could do with a little mental break for a while. I'm sure we can all identify with that burnout feeling where you eat, sleep, live, breathe this insane hobby for months at a time.

 

Eventually it starts to encroach upon your dreams, and then it's probably time to get it a rest. 

 

_________________

 

 

Here I've captured 25hours on another of this season's must-haves; the California Nebula!

 

About 8hours each on the the Antlia 3nm Pro filters, using a mix of 5-10 minute subs on the 1600MM Pro at -20'C.

 

Stacking and gradient removal in APP, then finished and balanced in Photoshop. 

 

• Here is the uncompressed full-resolution version.

 

Two personal developments on this project that have upped my game: 

 

1. Using HaSHO instead of SHO significantly improved my image quality. 

 

One thing I've developed recently as part of my 'personal style' is using almost exclusively Ha as luminance, then SHO as RGB.

 

Previously when using pure SHO, you would find that the image was terribly noisy; even with 10-20 hours of integration. By offloading the luminance into the strong hydrogen layer, I can significantly clean up the image. I find therefore that HaSHO is much better looking than SHO. 

 

2. A novel (?) method of ultra-clean star reduction. 

 

Another technique I'm working on is how to improve star reduction. I think a lot of tutorials have it wrong, in my opinion. Most people will tell you to select the stars then make them smaller with a 'minimum' (or similar) filter; but this leaves terrible artefacts. 

 

My approach has been to generate a pair of Starfull / Starless versions of the Ha image; then create a star-mask from the unstretched Oxygen layer; before using that to mask the 'starfull' Ha image back in. 

 

So instead of messily selecting stars in an image and making them smaller with bad artefacts - I just take an image with no stars at all; and then use the low-nebulosity oxygen layer as a mask to add a very clean and precise mask of them back to the starless image. 

 

This 'inside out' method is muuuuuuuuuuch cleaner than any other method I've seen described, and I couldn't find anyone else doing it this way. 

 

Maybe I can make a quick tutorial of my own in case anyone's interested? 

 

 

Thanks for looking!

 

_________________

 

• More of my images on www.instagram.com/baz_astra

_________________

 

 

TECHNICAL STUFF:

 

Spectral Palette: HaSHO (Ha luminance + SHO)

 

Exposures:

102 x 300s Hydrogen (8.5hrs)
52 x 600s Oxygen (8.7hrs)
100 X 300s Sulphur (8.3hrs)
 

= TOTAL 25.6hrs

 

Camera: ASI 1600MM Pro @ -20'c
Mount: Skywatcher HEQ5 Pro
Telescope: William Optics Redcat 51
Filters: Antlia 3nm SHO Pro narrowband
Guiding: ZWO ASIAIR
Software: APP + Photoshop
Sky type: Bortle 4
Location: Norfolk, England

Please make a youtube video of your process for what you are doing with stars. If you don't have a youtube account then a step by step will work..

Thanks..

[Borodog]

I am a novice, but the method I've been contemplating for star "reduction" and retention of star color for OSC data is simply to add an un- or under-stretched version of the entire image to the starless stretched image of the nebulosity. Yes? No? Too dumb/simple/won't work because XYZ?

 

Fantastic image, by the way.

Edited by Borodog, 18 January 2022 - 11:46 AM.

[Look at the sky 101]

Very impressive processing,  love the tone .

[oneredpanther]

I am a novice, but the method I've been contemplating for star "reduction" and retention of star color for OSC data is simply to add an un- or under-stretched version of the entire image to the starless stretched image of the nebulosity. Yes? No? Too dumb/simple/won't work because XYZ?

 

Fantastic image, by the way.

Basically, yes - that's more or less the distillation of what I do here. 

 

For narrowband I just get the star data from my least-nebulous channel (oxygen) and then use that as the star-mask to restore Ha stars back to the Ha channel. 

 

Doing it this way has the advantage of collecting none of the junk that is selected when using the traditional 'youtube astro-influencer' method (you know who they are) - because instead of starting with everything and making it smaller; you're starting with nothing and making it bigger. 

Edited by oneredpanther, 18 January 2022 - 12:01 PM.

[ConnorM]

It seems the stars have some weird shapes. I'd consider using less data to maintain the stars but the nebula looks awesome.

[bobzeq25]

PixInsight has a method of star reduction/star shape improvement that incorporates your philosophy.  And (standard PixInsight thing) it does so in an _extremely_ (understatement) adjustable way.  MorphologicalTransformation (includes a _lot_ of parameters) with a mask (ditto).

Edited by bobzeq25, 18 January 2022 - 03:31 PM.

[oneredpanther]

It seems the stars have some weird shapes. I'd consider using less data to maintain the stars but the nebula looks awesome.

I think that might be the compression on CloudyNights... I don't think the original shows weird shapes. 

 

 

PixInsight has a method of star reduction/star shape improvement that incorporates your philosophy.  And (standard PixInsight thing) it does so in an _extremely_ (understatement) adjustable way.  MorphologicalTransformation (includes a _lot_ of parameters) with a mask (ditto).

My friend has so far been unsuccessful in converting me to PI since I have 23 years of Photoshop experience; and any marginal improvement in perceived quality is vastly offset by the infinitely slower and more complicated way of doing things.

 

I'm here to produce nice-looking art for my wall. If it's slightly less scientifically accurate according to some signal processing algorithms... honestly I could care less about that.