PDA

View Full Version here: : M83 Reprocess again


Startrek
27-04-2022, 02:57 PM
Another M83 version using scientific aware colour type with 50/50 Layering
Bit more dramatic with brighter core

Comments welcome
Thanks

AdamJL
27-04-2022, 06:57 PM
Yeah that's terrific. Not an easy one to shoot. Core kept well under control.
But my favourite is looking for the "fuzzies" in the background. So many little galaxies hiding!!

Great work, Martin

Can I ask what causes the multi-colour diffraction spikes on the big star to the right?

Startrek
27-04-2022, 08:00 PM
Thanks Adam,
When using Newtonian reflectors the larger brighter stars end up with multi coloured diffraction spikes due to the different wavelengths of light reflecting through the spider vanes and secondary mirror holder at specific angles
I don’t mind them a bit adds a bit of character
One huge star that accentuates this phenomenon is Alnitak
Martin

AdamJL
28-04-2022, 01:15 PM
thanks for the info, Martin

The_bluester
28-04-2022, 01:25 PM
That is looking very nice, and the stars look better too than the earlier ones. Have you used curves or similar to darken the background? I reckon you will reveal more faint fuzzies with the background a touch brighter.

To add a little to the comment above, if you are imaging with a newt and you don't get the rainbow diffraction spikes, it can indicate that your collimation is not quite right.

Startrek
28-04-2022, 03:31 PM
Paul,
Thanks
I use Startools which works differently to say PI or PS
No levels , curves or calculators etc..
It uses signal Evolution Tracking from start to finish , so each module you use knows what the other one has done previously

Worth viewing the YouTube below (if you have time )

https://m.youtube.com/watch?v=8Luu2-BMXzc

You simply can’t clip your data in Startools unless you do something absolutely stupid
Quite a few modules have algorithms that can reduce brightness but maintain detail
I think what I failed to do was mask out the faint fuzzies when I applied the Star mask in the Shrink module to reduce Star size and glare , the faint fuzzies shrunk as well
Yes the image had quite a number of tiny galaxies everywhere
Thanks again
Martin

RyanJones
28-04-2022, 04:55 PM
Question for you regarding Startools Martin. You say it’s impossible to black clip unless you do something rather stupid. I downloaded the image an had a look at your background and you’re correct it’s not clipped but it’s so close. I’m assuming from that that the software finds the darkest part of the image and sets that to 0 then brings the background down to that level ? I know we’ve had this discussion before about the blackness of your background and I guess that’s just horses for corses. To me they all seem too black but that’s just my opinion. So with evolution tracking and all the other startools toys the program KNOWS what’s going on but without levels and curves etc. how do YOU know what it’s doing ?

Cheers

Ryan

Startrek
28-04-2022, 06:23 PM
Ryan,
Yep not black clipped
The blackness is a bit to harsh but there’s a reason behind it (within the algorithm )
I strongly recommend to watch the video ( link above in my reply to Paul ) Explains how Signal Evolution tracking works in comparison to traditional post processing
I think you will be amazed by what Ivo has designed and engineered in this magical software , it’s pretty much fool proof
Cheers
Martin

RyanJones
28-04-2022, 06:52 PM
I will watch the video out of interest.

What I’m getting at is that it’s all good and well for the software to know what it’s doing but if for example “ the blackness is too harsh “ and you’re not controlling it or a part of the process then you’re at the mercy of what the software does for you.

Nikolas
28-04-2022, 07:03 PM
Looks great Martin very subtle

Startrek
28-04-2022, 07:41 PM
Your totally in control always and not in any “mercy of the software”
It just takes away the guess work as the algorithms are mathematically correct with respect to signal and noise architecture
Hopefully the video will explain the mechanics of the software
Cheers
Martin
PS By no means is Startools the be all and end all but it’s damn close ( I must admit though all Startools users including me have some difficulty with the colour module , the later versions are definitely improving in this area )

Sunfish
28-04-2022, 09:13 PM
Excellent colour. A great rendition.

Startrek
30-04-2022, 12:11 PM
Thanks Nik and Ray
Pretty happy with the level of detail at 1000mm fl
Just wished those damn clouds stayed away for a few more hours, conditions were magic , guiding was just about flat lining , best it’s been for 6 months on the south coast
Cheers
Martin

Dave882
30-04-2022, 01:01 PM
Really great image mate! You’d think it was shot with a different scope compared to the first version. Really getting a heap of detail all the way to the core. Well done!!

RyanJones
30-04-2022, 04:33 PM
I watched the video Martin. I get the ethos behind it. I just don’t feel like it works the same way my mind does. Magical ? Or highly assistive ? I guess I see it like a manual gearbox vs an automatic. Without knowing the inner workings of the valvebody/mechatronics it could seem like magic how it knows when to change gears...... I like driving a manual.

This aside, my opinion on these images are that the background looks too black to me. It’s not huge but I just don’t feel like space looks real at 0,0,0 or very close to it. I took your image and lifted the black point to 20. It’s only a subtle change, I just feel it looks more natural. You may feel differently ?

Cheers

Ryan

Startrek
30-04-2022, 09:04 PM
Thanks Ryan for taking the time to view the video
It would be probably better understood if you did a trial on a dataset but everyone has to be comfortable with what they use and this Startools is a different animal than your traditional processing workflows and outcomes
I’ll try next time to get further away from that black point !!
Cheers
Martin