Rosette Nebula - Ha Oiii - Epsilon 180ed and Atik 383L+

[tooth_dr]

Here is an image from two short sessions - total of 3 hours and 10 minutes - 1 hour Ha and 2 hours 10 mins Oiii.  I think with a bit more data, I will get a better colour balance, but I see all the lovely images of a blue centre, and I couldnt manage to achieve that so a little disappointed.  Tips and pointers welcomed.

 

Thanks for looking

Adam

 

 

[MarkAR]

Lovely image, no idea how they get all the blue. Maybe they cheat and add Ha to blue channel.

[johngm]

Hi Adam, I think thats a cracking image 👍. What were the lengths of the individual subs ? I only ask, as i'm looking at a 383L myself. At the moment, its a toss up between that and the ASI1600MM. The thing that keeps swaying me away from the ASI is the microlensing issue that particular CMOS sensor has.

Regards John, and keep up the good work.

Forgot to add, yes, i do realise that the 383L+ uses a ccd sensor, but ive been interested by the images ive seen on astrobin

Edited March 16, 2020 by johngm
Missed text

[vlaiv]

If you want to try to get that "natural" look of this nebula, here is workflow that I would suggest:

1. Make synthetic luminance - use approaches like - weighted average, or max(ha, OIII). If Ha image contains signal in all areas where OIII signal is - use Ha as luminance as it is almost always higher in SNR.

2. Use color picker to pick your blue. Closest color to OIII is something like this: but you might want to choose a deeper blue. Once you select the color you want - check RGB ratios for it

3. Stretch Ha image but don't over do it - if you want to have distinct blue - leave center dark when you stretch - this will be your red channel

4. Stretch OIII image and apply it to RGB with channel mixer with specific ratios for blue color of your choice.

5. Combine red stretched Ha and stretched OIII colorized to blue to get color information

6. Stretch synthetic luminance and do denoising / sharpening, all the processing that you like in your image

and finally

7. transfer color from 5 to 6 for finished image

[geeklee]

Great image Adam. Loads of subtle colour and detail even with the selected palette. 

[Adam1234]

Great image

[B4silio]

Hey Adam!

(If you're using Photoshop...) A thing that helps quite a bit is to take the O3 image, create a starless layer from it, and re-add it as a "screen" layer.

Easiest way to get get a starless layer:

1. Copy your layer (important!) and rename it to Starless
2. Select > Color Range --> put into Highlights mode and drop the intensity to 190 or thereabouts (depends on the stretch of your O3), Fuzzyness at 20 is ok
3.  Select > Modify > Expand by 5-6 pixels
4. Shift+F5 for Content-Aware Fill
5. Filter > Noise > Dust and Scratches with ~6 pixels and 10-12 threshold to get rid of the dimmer stars
6. Spot Healing Brush to get rid of the halo on the larger stars
7. Done!

Now you can stretch and adjust your layer to get the blue core of the nebula to pop out a lot more.

I've taken the liberty of testing this out (just used the blue channel as the "O3" even if it's obviously a shortcut)

Hope this helps!

 

p.s. To protect the stars (so that they don't all become blue) you can:

1. Put your Starless just above your Original
2. Set blending mode of the Starless layer to Subtract
3. Ta da! You now have all your stars back

Having the stars as an added layer on top means you don't burn them out or re-color them when you add your blue core.

Edited March 16, 2020 by B4silio
Added test screengrab

[Hallingskies]

18 hours ago, tooth_dr said:

Here is an image from two short sessions - total of 3 hours and 10 minutes - 1 hour Ha and 2 hours 10 mins Oiii.  I think with a bit more data, I will get a better colour balance, but I see all the lovely images of a blue centre, and I couldnt manage to achieve that so a little disappointed.  Tips and pointers welcomed.

 

Thanks for looking

Adam

I much prefer your subtle handling of this to the lurid blues that seem so prevalent these days.  
 

[astro mick]

Very nice indeed Adam.

A lot of advice here,bit over my head I,m afraid.

Mick.

[tooth_dr]

On 16/03/2020 at 16:15, johngm said:

Hi Adam, I think thats a cracking image 👍. What were the lengths of the individual subs ? I only ask, as i'm looking at a 383L myself. At the moment, its a toss up between that and the ASI1600MM. The thing that keeps swaying me away from the ASI is the microlensing issue that particular CMOS sensor has.

Regards John, and keep up the good work.

Forgot to add, yes, i do realise that the 383L+ uses a ccd sensor, but ive been interested by the images ive seen on astrobin

Hi John

I used 600s subs for NB.  I have three cameras with the 8300 sensor.  I think it's perfectly good, and is very good value second hand.  The limiting factor here is me not the CCD.  The advantage of the CMOS is shorter exposures, but that in itself has some disadvantages around storage and processing.