StarTools

Hi,

I am using my new L-Ultimate filter to try to do an HOO image of the Crab Nebula. However I extract and load the Ha and OIII data in Compose and regardless of settings in the Color module, the image comes out heavily blue-green. That seems to me to be an overly unrealistic emphasis of OIII over Ha. I use APP to stack my images and for some of my runs in ST (I've tried both ST8 and ST9) I used its extract algorithms for the Ha and OIII to load in ST. Using APP's basic post processing, the result seems to have a better balanced display of Ha vs. OIII. See two attached representative results. I can't remember all of the different workflows I have tried. If there is a basic "textbook" best workflow for me to try for the correct balance of Ha vs. OIII, I would love to have a recommendation.

Thanks,
Jeff

Hi Jeff,

what a fantastic dataset ! So much detail. I tried M1 last night (visual spectrum) as well, but it only was a big mess. So jealous. What scope do you use?

AFAIK you don't need to extract the channels beforehand. ST will do that for you. Just load the same file into the RGB slots. Set parameter [Luminance, Color] to [L + Synthetic L From R(2xG)B, R(GB)(GB) (Bi-Color from OSC/DSLR)].
Later in Color Module set parameter [Matrix] for example to [HOO Duoband 100R,50G+50B,50G+50B]. Then try to use Bias sliders to set the desired colour balance between channels.

At least this is what I did with my duoband datasets. Good luck!

Best regards, Dietmar.

Hi Jeff,

Assuming you are importing unadulterated data as Ha and O-III, and have performed a Wipe afterwards, then the Color module should absolutely allow you to perfectly and accurately balance Ha and O-III at will. As of ST 9, the Color module will even name the two bands as such rather than saying "red" and "green"/"blue" (provided you set the "spectrum and filters" to "Duo/Triband/Quadband" and Luminance, Color set to on of the bi-color options).

If this is not the case, then something is very wrong. In that case the next step would be to upload your dataset somewhere so we can have a look.

The simplest textbook workflow is literally that; Compose > Wipe > Stretch > Color.

Hi Dietmar,

For this image I used my 10" LX200 Classic on a CEM70g mount, guided with an OAG and ASI174MM mini. I imaged with my ASI294MC Pro. These are among my first efforts with a new Optolong L-Ultimate. Things are promising given I am in a Bortle 8+ setting with Christmas lights now in the yards of most neighbors. I am still learning that filter and the subs used here were either 4 or 6 minutes each. I played with a lot of combinations of the subs I had but the total integration time was on the order of 4 to 6 hours. BTW, in my brief experience 3 minute subs are not really long enough with this filter, especially with an f10 scope. I am kind of surprised with the amount of OIII data that showed up which is why I posted my original question. Maybe the Crab Nebula truly has that much OIII as compared to its Ha? Maybe the simple answer is the L-Ultimate is as good as advertised and StarTools is accurately displaying the comparative strength of the Ha and OIII?

One really nice thing about the L-Ultimate is that with its very narrow Ha and OIII capture bands, light pollution is greatly reduced even in my very light polluted sky. (The poor Wipe module has very little to do as compared to the images I usually ask it to clean up).

Thanks for the suggested loading of Ha and OIII in Compose. Is there only one usable non-false color HOO setting in the Color module Matrix? Another question I have is what is the best positioning of the sliders for creating the synthetic luminance in Compose. I have read that either G or B should be set at zero or both reduced by 50%. Would R alone be the best source for the Luminance with this type of narrow band filter? I've tried different settings and didn't see a big difference in the result. Would it be worth the extra nights and time it would take to capture a regular set of luminance subs?

Thanks again,
Jeff

Hi Ivo,

I am starting in ST with linear RGB or linear Ha and OIII FITS extracted in APP using your suggested settings for stacking there and calibrate with darks, flats and darkflats. As mentioned in my reply to Dietmar, when I get to Wipe there seems to be minimal gradient problems, especially when I mask out the nebula. (I of course crop the dithered edges). In my second stretch I use an ROI covering the nebula and reduce the outside influence to 0 which along with a little Ignore Fine Detail, seems to give a decent result at that point. In various runs I apply Contrast, HDR, and maybe SuperStructure before Color. Sharp sometimes. Decon doesn't seem to help.

I have tried ST9 only very preliminarily. It looks very promising but I need to spend more time trying to understand all of its new features. For example, I cannot figure out its Decon. (BTW, in my ST cheat-sheet I have already done a search and replace AutoDev with OptiDev).

I would be happy for you to have a go with some of my data. Please give me your recommendation on how to make 150 MB, more or less, files accessible to you?

Thanks,
Jeff

Hi Jeff,

Not everyone probably agrees with me (big surprise lol), but I tend to consider any duoband through OSC to be false color narrowband by nature. Perhaps even moreso with an L-eXtreme or L-Ultimate. But yeah HOO matrix is probably the closest you are going to get to where those emission lines would fall across full spectrum.

I suppose you could just open the file as straight-up RGB, but legitimate emission throttling could be problematic (Ha dominant for most targets), as could maintaining the as-acquired ratio of the OIII in the G and B channels, which would be a product of your sensor QE along with the % passed by the filter.

I do believe that the Crab may be pretty strong in OIII. Also, your APP version strikes me as overbalanced towards Ha. Which is totally okay if that's the balance and appearance you want to show. But, the proportion of red and red-leaning stars is quite high, especially when most normal stars are probably going to be quite strong in the OIII band and weak in the Ha band. Just the way they shine, I think. Your ST version, with just a smattering of red stars, strikes me as closer to the as-acquired ratio.

Google Drive and Dropbox are probably the two most go-to methods of linking a big file for public download.

Since you used a 294MC, do you not just have one OSC stack? If so you'd load it up just as Dietmar said. If you don't have a single stack but instead are pre-extracted into Ha and OIII, then I suppose you could just treat them as if they were mono filter files. There's a bicolor option for that too (the one that doesn't say from OSC), and as you say you would either just load your OIII into either G or B, or if both, zero out the exposure on one of them.

I'm not sure if such APP extraction allows the little boost of 2xG that performing that in ST would?

Looks like you've captured some great structural detail here.

Hi all,

Here are three of the FITS files I used: the first one is a full RGB file; the other two are the Ha and OIII files extracted in APP:
https://www.dropbox.com/s/v39rfv9lnk8e2 ... .fits?dl=0
https://www.dropbox.com/s/dcvsr1prvt2pj ... .fits?dl=0
https://www.dropbox.com/s/wg8q5yf9ke290 ... .fits?dl=0

Jeff

Hi Jeff,

I had a quick go with your nice dataset ('RGB'). That all seems fine to me, nothing unusual. No problem to balance the channels to taste.
I'm not sure - is there a problem to get it to work for you or is it more the question what the 'right' colour balance should be?

Regarding Ha/OIII balance I guess it should be considered, that M1 shows strong continuum emission as well, due to it's synchrotron radiation emitting pulsar in centre (AFAIK).

So a whole bunch of the captured 'OIII' could actually be just continuum emission.

Best regards, Dietmar.

Hi Jeff,

I too could not find anything wrong with the datasets.

You'd want to import them like so; After a simple Wipe>AutoDev>Color you get a perfectly balanced HOO image;
Turning off artistic mode so we can better judge the emissions in all areas (rather than having colors desaturate by brightness); Which, for example, looks exactly like this rendition.

Note that you can linearly balance Ha or O-III in any way you see fit; there is no right or wrong here. You can choose precisely how much Ha vs O-III you wish to let poke through. StarTools' defaults (and the creator of the linked Astrobin image) merely chose the balance that lets just as much Ha as O-III through, as this tends to be the most informative for your audience.

As Dietmar points out, some increase in saturation is needed due to the strong correlation of emissions across all bands in this object.

There does seem to be something "off" about the APP image though. I can see slightly different hues of red, as well as lots of Ha-dominant stars (which is virtually never the case, or in your words, would be an unrealistic emphasis of Ha over O-III). It's as if the color channels were stretched independently (which would be a big no-no)...

Hi Deitmar and Ivo,

Thank you for your helpful comments and settings suggestions and the much better results than I could do with my data. Also, I had no idea that the Crab's super nova remnant pulsar could be the source of a lot of the narrow band blue-green captured by the L-Ultimate. That may explain the confusion I had in my original post regarding too much OIII. (Or, maybe it just has a lot of OIII as Mike suggests).

Thanks again. I'm still on the steep end of the learning curve.
Jeff