Rosette Nebula

[Peakedge]

Last Thursday night was the best night for a long time to collect some much needed photons. I managed to get about 40 subs for this one, of which I stacked the best 30. I was really excited by this one but I definitely need to get a coma corrector.

As always, I would very much appreciate feedback.

30 x 360s subs @ ISO800
Darks, flats and bias
SW 200PDS
Canon EOS 100d
Processed in Pixinsight

David

[PeterCPC]

Nice image but I would get rid of the green. You can use Pigs in Space Pixinsight to do that.

Peter

[Peakedge]

Peter

 

Thanks for the comment.

In my excitement  I forgot to run that process (SCNR?)

 

David

[laudropb]

Looks good to me, but agree that there is a green tinge.

[Peakedge]

Redone with the SCNR process................................less green now!

David

 

[Guest]

I have no criticisms as its a beautiful image.

I've never seen this object processed like this before and I am really intrigued by the oranges and dark reds.

Usually I get a salmon pink in the centre and far less structural detail.

I see you ran SCNR but did you Colour Calibrate while it was linear? Did you use Local Histogram Transformation for the detail?

The more I look at the more I like it ...I've just not seen it like this before - shows how I've gotten "locked in" to processing it in a certain way.

David

 

[Peakedge]

9 hours ago, David_L said:

I have no criticisms as its a beautiful image.

I've never seen this object processed like this before and I am really intrigued by the oranges and dark reds.

Usually I get a salmon pink in the centre and far less structural detail.

I see you ran SCNR but did you Colour Calibrate while it was linear? Did you use Local Histogram Transformation for the detail?

The more I look at the more I like it ...I've just not seen it like this before - shows how I've gotten "locked in" to processing it in a certain way.

David

 

David

I did background neutralisation, colour calibration and DBE whilst the image was linear and applied LHE sometime after applying the non-linear stretch. I have also been a little locked in to processing my pictures as I've been trying out various published workflows.

I did however make a 'mistake' with this one - I added drizzle. I think it reduced a lot of the noise from the outset but it made the files that large that it made processing some of the steps very long, in fact some wouldn't run. My computer is barely the minimum for running Pixinsight. Anyway, in the long run I think it helped because it meant I had to open small preview windows to at least consider everything I was doing and allowed me to experiment with different settings. I was more critical of what I was doing rather than blindly following settings used by others in their tutorials because once I committed to the main picture, I knew it could take an hour or so to process! I think this is the main thing to take away - experiment, experiment, experiment - use 'undo and redo' to compare before and after.

I also tried a few new processes; a masked stretch and I think it is called 'darkstructureextract'. I'll check later on that but the difference with the later was subtle but effective. Interestingly, I couldn't denoise the picture whilst it was so large. I saved both the large drizzled file and also converted it to a smaller jpeg which I was then able to denoise at the very end. It was the last step prior to being reminded that I had forgotten to apply SCNR! Its this version published above.

I think I was lucky in both the subs I managed to gather and the processing path I went down but I've learnt a lot from it and possibly won't be as locked in to process as I was. Experiment, experiment, experiment!

 

David

[Sam]

I like the colour in your image, it's fantastic. Colour can be subjective so not need to worry too much about it. 

[Guest]

4 hours ago, Peakedge said:

David

I did background neutralisation, colour calibration and DBE whilst the image was linear and applied LHE sometime after applying the non-linear stretch. I have also been a little locked in to processing my pictures as I've been trying out various published workflows.

I did however make a 'mistake' with this one - I added drizzle. I think it reduced a lot of the noise from the outset but it made the files that large that it made processing some of the steps very long, in fact some wouldn't run. My computer is barely the minimum for running Pixinsight. Anyway, in the long run I think it helped because it meant I had to open small preview windows to at least consider everything I was doing and allowed me to experiment with different settings. I was more critical of what I was doing rather than blindly following settings used by others in their tutorials because once I committed to the main picture, I knew it could take an hour or so to process! I think this is the main thing to take away - experiment, experiment, experiment - use 'undo and redo' to compare before and after.

I also tried a few new processes; a masked stretch and I think it is called 'darkstructureextract'. I'll check later on that but the difference with the later was subtle but effective. Interestingly, I couldn't denoise the picture whilst it was so large. I saved both the large drizzled file and also converted it to a smaller jpeg which I was then able to denoise at the very end. It was the last step prior to being reminded that I had forgotten to apply SCNR! Its this version published above.

I think I was lucky in both the subs I managed to gather and the processing path I went down but I've learnt a lot from it and possibly won't be as locked in to process as I was. Experiment, experiment, experiment!

 

David

Excellent feedback - thanks!

You're right about the experimentation in PI, I've had a lot of luck trying out different approaches but you've mention a couple I haven't even tried - reminds me to keep investigating this great software.

 

David

 

[StargeezerTim]

That really packs a punch! great image.

[gorann]

Yes, a pretty and interesting image. Like David I end up with a pinkish centre and a bit more cloud-like appearance. Your version looks somehow more solid, like you could walk on this nebula. I assume we often try to make it look cloud-like since we imagine that nebulae are some kind of clouds, which they are not, at least not in an earthly sense. I wonder how much more or less unconscious feelings like that goes into our processing.

[Peakedge]

Many thanks for the comments guys. Much appreciated.

23 hours ago, gorann said:

Yes, a pretty and interesting image. Like David I end up with a pinkish centre and a bit more cloud-like appearance. Your version looks somehow more solid, like you could walk on this nebula. I assume we often try to make it look cloud-like since we imagine that nebulae are some kind of clouds, which they are not, at least not in an earthly sense. I wonder how much more or less unconscious feelings like that goes into our processing.

Not sure about unconscious feeling going into a picture but guess it must do to be able to process to our personal tastes. However, I did worry that others might find it a little overprocessed. As for the colour, other than a little saturation in PI, it may have already been there due to light polluted skies, though I do have a CLS CCD filter. One difficulty I had was doing a proper Colour Calibration and Background Neutralization since there was very little area without nebulosity. Either of these factors may have affected the final colour but I do like the outcome.

 

David

 

 

 

[Guest]

6 hours ago, Peakedge said:

Many thanks for the comments guys. Much appreciated.

Not sure about unconscious feeling going into a picture but guess it must do to be able to process to our personal tastes. However, I did worry that others might find it a little overprocessed. As for the colour, other than a little saturation in PI, it may have already been there due to light polluted skies, though I do have a CLS CCD filter. One difficulty I had was doing a proper Colour Calibration and Background Neutralization since there was very little area without nebulosity. Either of these factors may have affected the final colour but I do like the outcome.

 

David

 

 

 

Try the Linear Fit process before you start any other process in PI - its designed to balance out red bias prior to processing - my images (taken with QHY8) are nothing but red before I start processing.

Colour Calibration helps with this but I've found Linear Fit, used first, does a great job.

David

 

[Peakedge]

12 hours ago, David_L said:

Try the Linear Fit process before you start any other process in PI - its designed to balance out red bias prior to processing - my images (taken with QHY8) are nothing but red before I start processing.

Colour Calibration helps with this but I've found Linear Fit, used first, does a great job.

David

 

I've not tried the Linear Fit Process before and will definitely try it on my next picture, should the clouds clear. I wonder if it will make a difference to this one since I don't think I ran Colour Cal properly due to limited background. It might explain why there is a colour difference between this picture and what people are used to seeing. Its worth an experiment.

 

David

[Guest]

On 2/18/2016 at 07:49, Peakedge said:

I've not tried the Linear Fit Process before and will definitely try it on my next picture, should the clouds clear. I wonder if it will make a difference to this one since I don't think I ran Colour Cal properly due to limited background. It might explain why there is a colour difference between this picture and what people are used to seeing. Its worth an experiment.

 

David

Hi David - excellent - I would definately try it on this picture - works really well for me, as I mentioned earlier my images all have a heavy red bias prior to its use.

Here's my notes on how I run it - you probably know this but you can preview the linear image in screen transfer function - this give the image a stretch etc for preview.

Have a look at your .fit file (stacked) before / after in screen transfer function to see what a great job this does :-)

Notes (health warninig: they may be wrong but they work for me)

1.       Linear Fit (LF) for very bad colour inbalance (check image for red bias)

a.       Open linear image > Linear Fit

b.       Split the image to LRGB: Image > Extract > Split to channels

c.        Select reference image: > Red

d.       Apply LF to green and blue image by left clicking on the red image and dragging it to the green and then the blue

e.    Recombine: > Channel Combination > select each channel from the RGB drop down > Apply Global

[Peakedge]

1 hour ago, David_L said:

Hi David - excellent - I would definately try it on this picture - works really well for me, as I mentioned earlier my images all have a heavy red bias prior to its use.

Here's my notes on how I run it - you probably know this but you can preview the linear image in screen transfer function - this give the image a stretch etc for preview.

Have a look at your .fit file (stacked) before / after in screen transfer function to see what a great job this does :-)

Notes (health warninig: they may be wrong but they work for me)

1.       Linear Fit (LF) for very bad colour inbalance (check image for red bias)

 

a.       Open linear image > Linear Fit

 

b.       Split the image to LRGB: Image > Extract > Split to channels

 

c.        Select reference image: > Red

 

d.       Apply LF to green and blue image by left clicking on the red image and dragging it to the green and then the blue

e.    Recombine: > Channel Combination > select each channel from the RGB drop down > Apply Global

 

David

Glad you've included instructions. You can imagine my confusion when I tried it last night to discover that my single colour image needed a reference image etc and that it was used for part of the process for processing RGB images. I actually hadn't thought of splitting the channels and so this process is invaluable. Thank you.

 

I'll let you know how I get on.

 

David

[astroman001]

Beautiiful image.

Peter

[Peakedge]

On ‎19‎/‎02‎/‎2016 at 15:57, astroman001 said:

Beautiiful image.

Peter

Thank you Peter.

On ‎19‎/‎02‎/‎2016 at 14:07, Peakedge said:

David

Glad you've included instructions. You can imagine my confusion when I tried it last night to discover that my single colour image needed a reference image etc and that it was used for part of the process for processing RGB images. I actually hadn't thought of splitting the channels and so this process is invaluable. Thank you.

 

I'll let you know how I get on.

 

David

David

I re did the process using Linear Fit and it has made a difference. The nebula colour has come out closer to the pink that everyone expects. I don't know which I prefer TBH but I have learnt a new step when doing a 'normal' Colour Cal and Background Neutralisation is possible due to no background showing due to the nebula!!

Many thanks

 

David

[Peakedge]

And the redone picture......

[Peakedge]

Or if not pink, less red!!!!!