First attempt at colour CCD - M51

[steviemac500]

So fairly happy I got the focus sorted out then managed to get 8 x 8 min lights on L filter. Trees were starting to encroach so i had to lose some of the colour which im pretty sure is why the final image is lacking and is heavily blued, only 2 x 5 mins in each RGB. I'm pretty sure the image is not bang on aligned as i can see pixels (red) just to the left of many points of light. But, I think there's progress so i'll need to try and add more colour next time. All advice welcomed......

 

Edit.....just a thought, should i be processing each channel seperately before merging? Could someone tell me the workflow i should be following for each step?

[Knighty2112]

Very nice. Great image. Well done!  

[Shelster1973]

Very nice indeed

Am new to CCD myslef and from what I have read up, you process each channel as a seperate image (including calibration frames for each channel) and then the final images for RGBL are aligned and then combined to produce the final colour image.  Am standing by to be corrected on this, but that is how I understand it

[steviemac500]

Thanks Shelster. I'm really asking if each colour should have levels, curves etc applied after stacking and before merging. 

[Shelster1973]

I think each channel should be stacked and kept in a linear state, with the levels and curves coming after the combination, but may be wrong on that one.

Have you had a look at the Light Vortex tutorials.  These are based around using Pixinsight, but the theory behind the combination and whether things should or should not be done will stand

[MattJenko]

Different thoughts on this one. I reckon you could experiment and see. I use a slightly different processing flow due to the software I use, but I in effect keep things linear, ie no stretching, but the relative levels of the colours are balanced against each other and the backgrounds matched using PixInsight tools. I also prefer to remove gradients on each individually, although that can be done on a combined image as well. In PS land this would equate to a very slight levels adjustment to match them up before stacking I believe and whatever action or workflow you use to remove light gradients.

[ollypenrice]

5 hours ago, Shelster1973 said:

Very nice indeed

Am new to CCD myslef and from what I have read up, you process each channel as a seperate image (including calibration frames for each channel) and then the final images for RGBL are aligned and then combined to produce the final colour image.  Am standing by to be corrected on this, but that is how I understand it

No, I wouldn't do that. I combine R and G and B into a single linear (unstretched) RGB and then stretch that. If you don't do it that way you will have to be very careful and scientific in stretching each colour channel in just the same way. This is a needless complication. I would do a simple log stretch of the RGB, not trying to wring the last drop of faint stuff out of it but going for low noise and high colour saturation.

I would then process the L with the priority on faint stuff and sharpness.

Next, add the L to the RGB in Ps blend mode luminosity.

I could write ten pages on the best way I know of doing this, but that would be a start.

Olly

[2STAR]

I think its a lovely image as it is to be honest

[Shelster1973]

11 hours ago, ollypenrice said:

No, I wouldn't do that. I combine R and G and B into a single linear (unstretched) RGB and then stretch that. If you don't do it that way you will have to be very careful and scientific in stretching each colour channel in just the same way. This is a needless complication. I would do a simple log stretch of the RGB, not trying to wring the last drop of faint stuff out of it but going for low noise and high colour saturation.

I would then process the L with the priority on faint stuff and sharpness.

Next, add the L to the RGB in Ps blend mode luminosity.

I could write ten pages on the best way I know of doing this, but that would be a start.

Olly

Olly

Thanks for clarification.  May not have made myself clear (have been doing a lot of that this week).  By my 'process each image' it was meant to be up to the point of doing levels and curves, so each channel is still in its linear state.

Am looking forward to getting some data myslef to work on, but seem to have perma-cloud here in the 'sunny southwest'

[Feral Rabbit]

Nice picture and looks like there is another DSO right on the top edge in the middle. My knowledge is poor so I don't know which one it is.

I'd certainly be happy with that myself....But then I am a rank novice when it comes to AP and astronomy.

Thanks for sharing