StarTools

Hallelujah so it seems I was doing it correctly the thing that was causing the headache as Ivo pointed out was the second autodev , this time I only selected an area within the Galaxy using Roi , also remembering to choose non linear srgb source when loading the autodev wipe autodev Ha file when converted to rgb rgb legacy thanks Ivo and HappyKat , I will repost my efforts now

Update https://stargazerslounge.com/topic/3732 ... mpetition/

I could not layer it in as the got the error dimensions do not fit...or something like that, which i actually expected ...???

Make sure you bin and crop exactly the same for all files ,so i binned 50% and cropped border 30 pix on all 4 axis .
i did a video of processing the Ha and adding to the lrgb not quite the zak snyder cut but may be helpful to someone
https://youtu.be/mLll-V5d_m4

Ivo im currently playing with some god data ,L4hrs rgb,2hrsHa 11hrs and 03 8.5 hrs so would it be best to add ha as above ? but what about 03 would i layer that in same as Ha after ha added to lrgb ,doing exactly the same with o3 data loading as blue,legacy ,interpolation off? so load the o3 legacy data tiff then layer the LRGB<Ha data (lighten).

update O3 added same as Ha data using layer module

Burly wrote: ↑Sun Mar 21, 2021 12:19 pm Ivo im currently playing with some god data ,L4hrs rgb,2hrsHa 11hrs and 03 8.5 hrs so would it be best to add ha as above ? but what about 03 would i layer that in same as Ha after ha added to lrgb ,doing exactly the same with o3 data loading as blue,legacy ,interpolation off? so load the o3 legacy data tiff then layer the LRGB<Ha data (lighten).

update O3 added same as Ha data using layer module
LRGBHaO3_50.jpg

For most objects (for example star forming regions), I would not recommend that specific Ha layering technique (it is best used for Ha-poor objects such as galaxies).

For objects that are strong in Ha emissions, and where you have a lot of Ha signal, by far the cleanest way to create a composite is to use the Ha as luminance and RGB for coloring. You can extend that to (Ha+O-III) for luminance and RGB for coloring. You will get all the Ha + O-III detail, while getting correct visual spectrum coloring this way.

So I would use compose to add Ha and 03 together using Lrgb save as a tiff then load that as a luminance using lrgb Ivo ? Or could you tell me the correct loading sequence so I don’t get muddled .

The practices of adding Ha+R to become a new red channel, or using Ha (Ha+L) as luminance, are definitely more universal (note that the former will skew visual spectrum colors of course, while the latter does not and is usually preferred).

A query Ivo --- in v7, how does one create a (n exposure weighted) Ha+L luminance, when also loading the R,G,B colour channels in Compose? I feel like compose needs one more available slot, haha, since this Ha+L blend is quite a commonly done thing. Now, I can imaging creating a synthetic Lum using the layer module, but this probably involves processing the Ha and LRGB separately, and then merging them in Layer, using the correctly calculated % contributions (like the old workflow for synthetic L in earlier versions).

Amaranthus wrote: ↑Sat Apr 10, 2021 12:34 pm exposure weighted Ha+L luminance

The trouble is that it is impossible to create any meaningful exposure weighted L and Ha composite. One is a very wide part of the spectrum (that even includes Ha itself again), the other is just a single emission line.

The reason why you can do exposure weighting with individual red, green and blue channels is that, in aggregate, they make up the full spectrum (as recorded in L), e.g; R + G + B = L.

Adding together different narrow bands for luminance creation (for example for a SHO composite) is also "doable", because camera sensitivity is - roughly - equal for every slice of the spectrum. E.g. Ha + O-III + S-II = synthetic L (not to be confused with full spectrum luminance).

There is, however, something I've started work on adding to the Compose module (and rippling through the entire engine), which is narrowband accents. These are be processed separately in a third parallel path (like luminance and chrominance) and should not impact either luminance or chrominance during processing, until you are ready to apply them at very late in your workflow.

Thanks Ivo, that does make sense, now that I think on it after your explanation. To combine these would be more art, and not 'documentary'.

That said, I'm intrigued by the 'accent' approach you've tantalised us with --- will stay tuned!

I just discovered the IKO project in the SGL forum. At the risk of being very late in the game, may I submit my version of the M33 data?

Russ