Tarantula Nebula separated into OIII and Ha emissions

[sjastro]

Generally I try to avoid imaging common objects unless I am trying out new techniques.
One technique under development is to separate out OIII and Ha dominant features in an object and enhance each wavelength by remapping the colour.
In this case OIII is mapped to cyan and Ha to red.

The Tarantula is a good target, because it is bright in both OIII and Ha.
The challenge is identifying the dominant wavelength and is achieved by using a (OIII-Ha) and a (Ha-OIII) mask.

A conventionally processed image can then be remapped through each mask revealing and enhancing details that are not commonly seen.
The other advantage is that star colour balance is maintained.

2 hrs OIII + 2 hr Ha as a luminance image. 1 hr RGB each.
BRC-250 and STXL-11002 CCD.

High resolution image.
http://members.iinet.net.au/~sjastro/Tar_OIII_Ha.html

Clear skies

Steven

[gregbradley]

Great image Steven.

Greg.

[alpal]

Hi Steven,
that's an amazing picture - shows the quality of your optics.
That's a huge FOV for a 10" f/5 and sharp to the corners.
I'd like to see the separate mono frames of O111 & Ha.
Thanks for sharing

cheers
Allan

[marc4darkskies]

Unconventional, but compelling Steven. Nicely done.

Cheers, Marcus

[sjastro]

Thanks Marcus.



Here they are Alan.
http://members.iinet.net.au/~sjastro/Tar_OIII.jpg
http://members.iinet.net.au/~sjastro/Tar_Ha.jpg

To see the differences in greater one needs to subtract the Ha image from the OIII image and vice versa.
These were used as maps to extract the detail in the OIIIHaRGB image.

http://members.iinet.net.au/~sjastro/OIII_map.jpg
http://members.iinet.net.au/~sjastro/Ha_map.jpg



Thanks Greg.

Clear skies

Steven

[alpal]

Steven,


Thanks - you can really see now where the signal was.
Interesting that you subtract to make masks.

cheers
Allan

[Merlin66]

Steven,
That's amazing!!
Your efforts are well appreciated!

[RickS]

I really like it too, Steven. Good to see some new processing ideas.

Cheers,
Rick.

[sjastro]

It's very simple in PI.
Use pixel math to subtract the images, make sure the mask is registered with the main image and use a manual colour calibration command to work on the areas of interest.



Thanks Ken.



Thanks Rick. As you are a PI aficionado, I use the Assisted Colour Calibration command to work on the masked areas.
I'm sure the developers of this function in no way intended the function to be used as I am using it.
It allows me to accurately assign colours to the respective wavelengths.
OIII is blue green or cyan in colour, and Ha is deep red.

The challenge is to now add SII to the mix and assign a colour.
The most "realistic" colour for SII is a deeper shade of red which can be developed using the Assisted Colour Calibration command.

Clear skies

Steven

[multiweb]

That looks really cool. I like the palette. :thumbsup: Details are simply superb.

[sjastro]

Thanks Marc.

Steven

[sjastro]

Hello Marc,

It's not a different palette, but a (HaOIII)RGB image processed to differentiate between OIII and Ha regions and assign the correct colour to these emissions.

I have found SII doesn't add much to process, not surprisingly as SII is a weak emission and close to the Ha wavelength making it hard to differentiate.

Steven

[atalas]

Good work Steve.

[sjastro]

Thanks Louie.

Clear skies

Steven

[Stevec35]

Interesting processing technique Steven and it's certainly effective because that's a very nice Tarantula.

Cheers

Steve

[Ross G]

Hi Steve,

Your technique works well.

That is a great looking Tarantula photo.

Ross.

[Shiraz]

interesting approach and a very fine image Steven.

[sjastro]

Thanks Steve,Ross and Ray.

This is still a work in progress.
It needs further refinement.

One aspect I need to consider is to normalise the OIII and Ha images as part of the mask creation.

I'll post a revised version of the Tarantula.

Steven

[Peter.M]

I am interested in how you did this, did you account for differing responses at each wavelength. Filter transmission and ccd spectral response will both change the relative amount of intenintensity of each channel

[sjastro]

This is the next step of the process.
The way of dealing with these issues collectively (including sky glow) is to normalize the sky background of the stacked OIII and Ha images before constructing the masks.

The masks are applied to the OIIIHaRGB image and enhancement is done through the masks exclusively with the R channel.

When I get the time I'll process and post the final result.

Regards

Steven

[sjastro]

Analysis of the sky background found that the OIII stacked image was 1.2X brighter than the corresponding Ha stacked image.

This meant there was a bias towards OIII in the enhancement process and less Ha detail coming through in the masks.

Normalizing the OIII background relative to the Ha background has resulted in a better balanced image with a cyan appearance corresponding to the OIII emission and more red Ha detail near the core.

The white core is mainly due to near ultra violet.

Reprocessed image.
http://members.iinet.net.au/~sjastro/Tar_OIII_Ha_new.jpg

Steven