Not getting the results I'd like with LHaRGB

[BrendanC]

Hi all,

I've captured some Ha of Bode's Galaxy which I'd like to incorporate into my LRGB image.

I've tried every which way, using the NB Accent, layers, blending modes etc etc etc but cannot get anything that is appreciatively different from the LRGB version. The Ha just isn't coming through. I was kind of hoping this tutorial would help but it didn't: https://www.youtube.com/watch?v=mLll-V5 ... e=youtu.be

Anyone fancy taking a look at my files and seeing what they can get out of them?

Two choices:
* If you'd like to play around from first principles, the individual FITS files for L, R, G, B and Ha are here: https://1drv.ms/u/s!AqovBuVZMwj3k4RK7OE ... w?e=be1BGd
* If you'd prefer to skip all that and just play around with layers, the processed TIFF files for LRGB and Ha are here: https://1drv.ms/u/s!AqovBuVZMwj3k4RJ54b ... A?e=4qvhQq

Thanks
Brendan

[LuckyEddie]

Hi Brendan,
Here's my attempt, using NB Accent. Mostly for my own education, I'm not in any way experienced.

Bodes LRGBHa.jpg (369 KiB) Viewed 4595 times

And log...

StarTools.log.zip

(45.27 KiB) Downloaded 287 times

Regards
Ed

[BrendanC]

Thanks Ed, I'll take a look. It's better than my attempts.

[BrendanC]

OK, so I've looked at this and it's really interesting. You've done pretty much what I did EXCEPT that, when loading the channels into the Compose module, I see that you've allocated 3600s to the L, and then 10,800s to R, G and B.

Why is this? I've been diligently putting in the actual values for the amount of exposure I have, which is L 5880s, R 2520s, G 1980s and B 2340s. Is this wrong? Should I not be doing this? I thought that was the point of adding the total exposure times!

Also, I thought this was the wrong way around, and that L should have the same total exposure as R, G and B combined. In which case I'd have thought it would be 10800s to L and then 3600s to R, G and B.

I mean, I can't argue with the result! It's better than mine by quite a long way. And I'm really, really grateful for your help - thank you so much.

But I don't understand why I've been doing this by the book, and not getting something similar.

Any ideas?

[LuckyEddie]

I didn't have a clue about the overall exposure times as you didn't mention them, so I just went with the exposure times given in the filenames. I dialed them in as 60 & 180 minutes rather than seconds assuming that it's the ratios that matter rather than the absolute values. Did I mention I'm not experienced?
It sort of looked OK to me though.

[BrendanC]

Ah, apologies - the durations are in the FITS headers, which you can see in apps such as for example FITS Liberator.

I'll have another run through this sometime and see what happens. Thanks again!

[admin]

Hi,

I'm getting expected results from the NBAccent module;

Very quick log as follows;

--- Compose
Parameter [Luminance, Color] set to [L + Synthetic L From RGB, RGB]
Parameter [Color Ch. Interpolation] set to [On]
Parameter [NB Accents Type] set to [Ha/S-II from NB filter]
Parameter [Lum Total Exposure] set to [1h38m (98m) (5880s)]
Parameter [Blue Total Exposure] set to [0h39m (39m) (2340s)]
Parameter [Green Total Exposure] set to [0h33m (33m) (1980s)]
Parameter [Red Total Exposure] set to [0h42m (42m) (2520s)]
Type of Data: Linear, was not Bayered, or was Bayered + white balanced
--- Auto Develop
--- Bin
Parameter [Scale] set to [scale 50.00% / +2.00 bits / +1.00x SNR improvement]
Image size is 2505 x 1932
--- Crop
Parameter [X1] set to [722 px]
Parameter [Y1] set to [411 px]
Parameter [X2] set to [2173 pixels (-332)]
Parameter [Y2] set to [1828 pixels (-104)]
Image size is 1451 x 1417
--- Wipe
Parameter [Gradient Aggressiveness] set to [94 %] (too lazy to crop away the stacking artifacts again )
Redoing stretch of linear data
--- Crop
Aaaaand I just realised this is going to be very big for an animated GIF - better reduce some more.
Parameter [X1] set to [32 px]
Parameter [Y1] set to [78 px]
Parameter [X2] set to [1219 pixels (-232)]
Parameter [Y2] set to [1215 pixels (-202)]
Image size is 1187 x 1137
--- Auto Develop
Parameter [Ignore Fine Detail <] set to [4.4 px]
Parameter [Outside RoI Influence] set to [15 %]
Parameter [RoI X1] set to [148 px]
Parameter [RoI Y1] set to [192 px]
Parameter [RoI X2] set to [442 pixels (-745)]
Parameter [RoI Y2] set to [549 pixels (-588)]
Parameter [Detector Gamma] set to [1.00]
Parameter [Shadow Linearity] set to [50 %]
--- Color
Legacy preset (I'm using it to intentionally desturates lighter part of the image, to make the Ha we will be adding, stand out better).
Parameter [Saturation Amount] set to [336 %]
Parameter [Blue Bias Reduce] set to [1.00]
Parameter [Green Bias Reduce] set to [1.13]
Parameter [Red Bias Reduce] set to [1.27]
--- Visual Spectrum Narrowband Accent Compositing
The Ha data is noisy. And the trick is to separate as much noise from Ha detail as possible. Any speck of noise will make it into the final composite, so try to "remove" (clip) as much as possible without clipping/destroying faint detail.
Parameter [Ignore Fine Detail <] set to [15.4 px]
Parameter [Outside RoI Influence] set to [15 %]
Parameter [RoI X1] set to [148 px]
Parameter [RoI Y1] set to [192 px]
Parameter [RoI X2] set to [442 pixels (-745)]
Parameter [RoI Y2] set to [549 pixels (-588)]
Parameter [Threshold] set to [30 %]
--- Visual Spectrum Narrowband Accent Compositing
Galaxy preset.
Parameter [Response Simulation] set to [Ha/S-II (Pure Red)] (also makes the Ha stand out, as there is already Ha visible in the wide spectrum LRGB - here it is usually purple/pink colored due to Balmer series correlation, rather than pure red).
Parameter [Detail Size] set to [100 px]
Parameter [Brightness Correlation] set to [50 %]
Parameter [Gamma] set to [3.51]
Parameter [Strength] set to [109 %]

Hope this helps!

[BrendanC]

Outstanding, thank you Ivo. I'll check this later but it does look a lot better already - much more along the lines I was expecting, especially with all that lovely Ha around the centre of the Cigar.

[Mike in Rancho]

Hi Brendan,

Thanks for posting the sets. This was my first try at using NBA in a galaxy poke-through setting, so, total newbie. But it was fun.

Not very good probably. I didn't pull too much of Holmberg IX here. Nor does it flash like a Christmas tree.

I didn't like the way many of the stars also turned all hydrogen red on me, so I used the undo buffer right after NBAccent. Not sure if legal.

BrendanC HaLRGB M81-82 ST8 1B.jpg (414.35 KiB) Viewed 4527 times

[BrendanC]

Hey, Mike! Thank you again! And this is another version better than mine. Time to go through all this...