Make your own M20 narrowband pic

[Bassnut]

Now try this M20 Triffid nebula Narrowband image.

The 3 subs are labeled with "Hubble" pallete RGB mapping, but you can mix em up anyway you want ;) . I havent tried other combinations, ittl be interesting to see what other combinations you can make look good !.

I also used the HA (green) channel as an extra Lumanence layer to add sharpening (it has more data and can stand more processing before getting noisy).

Processing is exactly the same as in RGB, only the stars will have a pink halo (typical in narrowband pics), to avoid that, you need to reduce the star size in the red channel, and a bit in the blue channel if your fussy, but its not essential.

You will get strong colours, narrowband is good for that.

The 3 tiff files are the full res from my cam.

http://www.box.net/shared/dgjth43qfy
http://www.box.net/shared/84cm8i6ni2
http://www.box.net/shared/cffm8n976h[/URL]

Just click on the images name and select save.

The thumbnail is my effort, and here

[URL]http://fredsastro.googlepages.com/M20TriffidNebulaNBfinalBlrg.jpg/M20TriffidNebulaNBfinalBlrg-full.jpg (http://www.box.net/files#0:f:7082971)

Its not hard, just treat it like an RGB and youll be fine.

Cheers
Fred

[mill]

The site is asking me to login :(

[Bassnut]

Mill, sorry, im working on it:P

[mill]

:mad2: :shrug: No problem :)
:D :thumbsup:

[Bassnut]

Mill, try this (ive edited the 1st post). I havent used this service before, it logs me automatically (I must be special ;-), so I cant tell if you need to.

Fred

[mill]

That works :)

[mill]

Hmm all my viewers and processing software see all the files as grayscale.
Even adobe cs2.

[Bassnut]

And one star has a bit of blooming, non antiblooming sensors do that when they saturate, heal it out if your keen, along with the slight artifacts caused by the deconvolute filter. Astrophograhy is a challenge, rock on :thumbsup:

Fred

[Bassnut]

Mill, thats correct, they are mono subs. Thats the idea, you make them RGB anyway you want.

In CS, convert the red image to RGB and copy and paste the green image into the green channel (as is, no RGB conversion) and the blue image into the blue channel. The red channel already has the red image from the 1st RGB conversion.

Now you can process each channel seperately or the RGB layer altogether, its already a "flat" single layer.

I couldve posted a colour image, but the idea is to do it yourself. If you like, I could post a photoshop file ready loaded but it will be big.

Would you like a PS file?.

Fred

[mill]

Aha :)
Ok will make something out of them.
Dont expect anything good.

[Bassnut]

If your confused with mono subs, heres a loaded PSD file ready to use in Photoshop CS. Half the res and 8 bit, but pre curved so easy to use, probably usable in earlier Photoshop versions too. 7.5 meg.

http://www.box.net/shared/s3nr83bdk6

Im keen to see what creative spirits are out there :thumbsup:

Go dudes

Fred

[Doug]

G'day Fred, I made a bit of a hash of your data. I tried something that backfired to get rid of the red halos..........Oh well at least now I know that won't work. Might try something else.
cheers

[Bassnut]

Nice one Doug, the stars look OK, a slight halo is acceptable, a NB signature. Push the colour, its all very subjective.

Fred

[Astroman]

Firstly I would like to thank you Fred for making this data available. As I am still getting used to the various techniques in image manipulation and various setting used with filters this gives me the opportunity to have a fiddle with data that is not normally available unless we go out and get the filters etc ourselves. Although not a great fan of the false coloured images it still gives us a good view of the detail found in these wonders. Anyway here is a quicky from the above data.

[Bassnut]

Andrew, I know what you mean, NB false colour is not for the purist, its an "artistic" effect unrelated to reality. Your processing is dark, cold and menacing. Striking! (mmm, perhaps NB processing reflects ones soul, being so arbitrary, do you have you a dark, menacing soul Andrew....just kidding, reading personality through processing could be fun with this hehe), you do have the artistic bent though, you must admit its a free form type of astrophotography.

Excellent :-)
Fred

[Garyh]

I would have a play but it would take too long to download on our snail dialup here...

[jase]

Thanks for providing the data Fred. Below is my one of the versions I’ve come up with. I started out as HOS, but the SII channel was rather dirty – perhaps needs more data collected. OIII is come up good after smoothing. The Ha is well captured, very little noise present and clean – good work. Fixed the blooming, this isn’t a big problem, but wasn’t very careful in doing so which affected some nebulosity. So based on this, I decided to do an RGB rendition using your data. Thought about using a synthetic green channel, but M20 does not emit much in the green wavelengths (more cyan). So the colour palette is Ha:OIII:OIII+Ha @ 25% opacity – mapped as RGB respectively.

This was then flattened and smoothed to create a standard RGB, that was later combined with the Ha for more depth – I couldn’t resist as the Ha channel is clean. I flatten this again to create a super RGB (that included the Ha as luminance), then added the Ha channel one more time to complete the LLRGB combination. The star colours went a little strange on me (began going teal). For NB this doesn’t matter so much, but as I was creating a pseudo RGB out of NB data, I wanted to keep it realistic. So I use the colour selection tool to rectify this (slightly increased yellow and magenta). I may try working with the SII frame and produce the HOS I originally intended to do.

[Bassnut]

Wow, nice work Jase, very colourfull. M20 suits the lolly pop look, its a popular way to present it. I like the LLRGB treatment to boost colour. Ive read Rob Gendlers original triest on this process, but havent tried it myself.

You also got the bue nebulae out on the side, I buried it by useing Ha as a Lum channel, Ha is not the right way to make a Lum channel in NB, but I got more sharpening with it. I have since sharpened this image even more with sucess, but still with much loss of the blue neb.

I often select out the stars for seperate processing, and exclude them when adjusting colour/saturation/sharpening on the rest, this avoids star artifacts and colour shift on them.

Yes, the SII data is lean, not much in M20, although I could have gotten more I guess. Its hard going back to to get more and do it all again later, you tend to move on ;-).

It will be interesting how you go with your own images once you get set up and capture some.

Thanks for the effort and describing how you did it, somewhat out of the ordinary, I have learnt more from the way you did it.

Cheers
Fred

[Dr Nick]

Wow! Well done to everyone I think, ty are all great! ;)

[jase]

Hmmm. M20 doesn’t look right when I use HOS palette (aka Canada-France-Hawaii Telescope – CFHT Palette). Just couldn’t get the right tones. The Hubble (SHO) palette appears to be the best fit.
So, I’ve process this based on OIII:SII:HA. Considering M20 (as we know it) has strong tones of blue, I thought the excellent signal from the HA image would do nicely. Processed this slightly different. As the Ha image dominates in clarity, I used this layer as an inverted layer mask to selectively Gaussian blur the OIII and SII channels. This reduced the inherent background noise, but there is still some present. Don’t think I went hard enough (still learning). I used the Ha image for luminance again, though typically NB images don’t need them. I lowered the white point output in levels to 200 (original 255) before applying an unsharp mask. Lowering the white point output ensure no data gets clipped as the unsharp mask increased the brightness. Star colour balance could have been better. Was trying to work out your arcsec/pixel as the images are probably oversampled and need some further sharpening/deconvolution in general. I haven’t experimented further with that. I miss NB imaging.

Yes, can wait to get my set up going Fred. Was at it last night (7pm to 4am) tuning the mount after I re-greased the worms. Reprogrammed PEC, now down to +/- 1.4 arcsecs, but still have a few polar alignment issues at the moment – having problems getting under 2 arc-min from the pole. Will get there eventually.

[mill]

Here is my humble try at this picture.

[Bassnut]

Jase,again, a very nice rendition, striking colour.

Sorry, perhaps I should have mentioned the Ha is binned 1*1 (image scale 0.92 arcsecs/pix) 140min and the other 2 are binned 2*2 (1.84) 50min, as I had in mind Ha as Lum from the start. Thats why the stars are pixelated in this pic I guess ;-). I also double the res before processing (Im not sure if you did), tends to smooth things out. The stars look better than your 1st effort.

I have found in the past SHO is often the most attractive, perhaps because we are used to seeing Hubble shots as the standard.

The inverted layer mask is interesting, have only used cumbersome fancy layered selections when filtering.

Sub 2 arcsec (RMS?) is a worthy pursuit, excellent work if you reach that ;-).

Thanks
Fred

[Bassnut]

Hi Mill

Hey, humble?, top work, close the conclusion I came to, cant be that humble hehe.

Stars have just a slight NB halo hint to them, sweet. If you havent done much Astro CS gear, this is a fine effort, youve pretty much nailed it.

Fred

[Ric]

Hi Fred, thanks for the chance to play with some great images like this.
I had a quick go in Maxim but had to crop it a bit to get it down to the right JPEG size.
I'm thinking about a set of these filters at the 13nm level so it was a great opportunity to play with this type of data.

Cheers

[Bassnut]

Nice one Ric. It looks weird upside down, ive looked at it too many times ;-).

Cheers
Fred

[2020BC]

This is great.