WR-134

[MarkAR]

Not wrong at all, I never realised until I did some research that NB integration is best done non-linear.

Improved the stars a bit and slightly stronger Oiii, it's a winner.

 

[Rodd]

25 minutes ago, MarkAR said:

Not wrong at all, I never realised until I did some research that NB integration is best done non-linear.

Improved the stars a bit and slightly stronger Oiii, it's a winner.

 

Thanks, Mark.  Came as a surprise to me.  I’ll have to revisit some of my other NB images

[Kinch]

On 20/06/2020 at 17:18, Rodd said:

this was new to me

....surprised you said that.....I recall telling you (quite some time ago) this was how I worked on my images. I have never seen it written (that I recall) - but fo me it sure is an easier approach and allows more control.

[Rodd]

28 minutes ago, Kinch said:

....surprised you said that.....I recall telling you (quite some time ago) this was how I worked on my images. I have never seen it written (that I recall) - but fo me it sure is an easier approach and allows more control.

I guess my memory is not what it once was.  

[MartinB]

I'm only looking on my phone at the mo.  The ha seems a better colour and their is some nice OIII in the background so, overall I would say better.  Will have a look on my calibrated monitor tomorrow.  

[Rodd]

6 minutes ago, MartinB said:

I'm only looking on my phone at the mo.  The ha seems a better colour and their is some nice OIII in the background so, overall I would say better.  Will have a look on my calibrated monitor tomorrow.  

Thanks Martin. I stared out with much more Oiii...but maybe too much.  Things were pretty blue.   So I knocked it back to what I posted.  It’s a tricky balance.

[MartinB]

Looking on my laptop now, and yes, the latest version has pulled out the faint OIII nebulosity and the Ha colour looks great.

I was interested to hear that the PI gurus were suggesting that you should stretch each channel before doing the combine.  I stretch in PS but the principles are the same.  It makes sense to do an optimal stretch for each channel separately and then combine and I've certainly tried this plenty of times.  Certainly combining at the linear stage and then boosting OIII and SII to match the Ha permaturely can trash the Ha.  However, gradually bringing the histograms to a balance giving the Ha a reasonable stretch before they are fully matched does give you a great deal of control over the final colour balance at the different brightness levels.  Using the autostretch in PI for each channel definitely doesn't work very well! 

However you went about it, it has worked like an absolute dream.  This is a cracking image.

[ultranova]

Great Capture and processing Rod,

Never seen  the blue before, so thanks for that, 

it would be a hard target to gain data in the uk for the blue side of it ,

Thanks for sharing, great  job.

Paul

[Rodd]

On 21/06/2020 at 13:19, Kinch said:

....surprised you said that.....I recall telling you (quite some time ago) this was how I worked on my images. I have never seen it written (that I recall) - but fo me it sure is an easier approach and allows more control.

What I have noticed after trying this method now on a few images is I can no longer use Ha as a luminance--to be fair, I suppose the right way to say it is using Ha as a luminance is no longer necessary.  But I was surprised to find this to be the case.  Without exception, if I channel combine NB stacks the normal way (the RGB way), using Ha as a luminance (channel combination in Clab mode with Ha chosen for L) invariably improves the image....I don't think there have been any exceptions to this (except bicolor images, where I usually cant insert the Ha OR OIII as lum--I have had some success integrating the Ha-OIII and copy of Ha (gotta love PI) into a synthetic L and inserting that--but not the Ha or OIII alone).

So--a couple of my reprocesses resulted in an improved image and a few did not.  I think, therefor it is target specific (or maybe data specific).  Not sure its any easier as you have to stretch 3 stacks instead of one--but there is less fiddling afterwards--so I think a wash as far as ease.  I find NB processing fairly easy anyway compared to LRGB

Anyway--just some thoughts.  

[ultranova]

Your not wrong,

Looking at all the ones you have processed this I would say is the best take yet

But as they say Beauty is in the eye of the beholder,

in other words what me and you think is better other people might not,

every one see thing differently.

you have done a top job on gathering the data and processing it,

I would be very proud

Well done Rodd

[Kinch]

Re your thoughts about adding in that Ha as Lum - I actually find it the other way around......in that it almost always will work for me with only an occassional problem image not accepting it. Strange......but you may be right in that it is data/target specific.