Ha-OIII integration - which algorithm in 0)RAW/FITS ?

I always use Adaptive Airy Disk for RGB data.

Thank you

@mabula-admin … are these different algorithms?

Hi all.

I found the answer - see this post, with Mabula's explanation.

Answer (I quote):
==> For Ha-OIII duoband filter use "Ha-OIII color/mono/extract Ha/extract OIII"
==> For OIII the Adaptive Airy Disc and these "Ha-OIII" algorithms are equally good.
==> But for Ha, the "Ha-OIII" algorithms are better than Adaptive Airy Disc.

@mabula-admin Bonus question:
Is the same true for an SII-OIII duoband filter?

Thank you and best regards.

Michael 

PS: Proposal for next Release: 
a. Additional Option "Ha-OIII extract both channels as mono"
b. Duplicate and add same selections for "SII-OIII ..."
c. For the mono channels: add "Ha", "OIII", "SII" into the FITS header.

Hi Michael @michaelacg and @digitaliz-se,

Answer (I quote):
==> For Ha-OIII duoband filter use "Ha-OIII color/mono/extract Ha/extract OIII"
==> For OIII the Adaptive Airy Disc and these "Ha-OIII" algorithms are equally good.
==> But for Ha, the "Ha-OIII" algorithms are better than Adaptive Airy Disc.

Yes, the same holds for SII-O3 dualband. The Adaptive Airy disc algorithm uses the green pixel data to interpolate red and blue channels, so it works equally well for O3 data as the Ha-O3 color algo. But for Ha data, there is no info in the green channel and thus the Adpative airy disc performs less well than the Ha-O3 color. It can be a bit confusing, but it means that Ha-O3 color is a better choice for Ha-O3 data if you want a RGB result.

Yes, the proposals are already on our list to implement, thanks!

Mabula

Hi All,

Hopefully my question is closely related enough to be added in this thread, let me know if new one required!

I'm working through a multi session project using double dual NB filters (Ha-Oiii + Sii-Oiii - Altair 4nm) on OSC camera (ASI2600mc).

Questions:

1) What is the optimum way to load the subs and integrate to obtain monochrome NB stack for each channel? I feel I might be making this more long-winded than it needs to be, so appreciate any advice!

2) As I plan to add more data when I get clears nights, is it ok to re-stack using the saved registered & normalized frames, or should I calibrate / full process each time I add more data? Any other shortcuts to add more subs but keep the quality?

Context:

I've been loading all my lights (with their corresponding calibration frames) from both filters at the same time using multi session (corresponding to each nights worth of data) and multi filter (custom named filters HaOiii and SiiOiii) then extracting each NB channel over the full set using the 0-Load: Extract Ha (and the repeat the exercise after integrating for Extract Oiii, extract Sii).

If I then integrate using the 6-Integrate: MultiChannel option Integrate per channel & all and MultiSession option: integrate all, I end up with 3 sets of 3 images that I'm not sure which one is correct for the extract purposes.  

Using my 'best guess' of each, I was able to produce a decent looking result, however I'm afraid it may be completely inaccurate!

https://app.astrobin.com/u/Zakky2k?i=lflwk3#gallery

Quick update - the following thread contained great info and is almost the same process I’ve been following: https://www.astropixelprocessor.com/community/postid/30580/

Just wondering if the ‘extract Oiii’ is effective when data from two dual-band filters is included in the same process, and if an option could be added to output all three mono NB channels ?

Hi Zak @zakky2k,

Thank you very much for your questions. 

1) Best way is to use the extract algorithms in 0) Raw/FITS per wavelength, so make them for Ha, O3 and Sii, for SII use H-alpha extract. The next APP version will have SII-extract as well 😉

2) The most ideal way is to save normalized, calibrated frames, but not yet registered. Once you want to combine all into a stack, disable normalization. Load all the calibrated, normalized lights and register them and then stack 😉

That result looks very fine to me, great result ! I will make sure that in the next APP version, the correct Filter assignment will be included in the FITS header.

Just wondering if the ‘extract Oiii’ is effective when data from two dual-band filters is included in the same process, and if an option could be added to output all three mono NB channels ?

That is just fine to use extract OIII directly on data loaded from separate dual-band filters provided they both actually have O3 data in them of course.

Why would you want 3 mono channels from dual band filter data? Is that for a special filter that has 3 narrowband signals? One in red, another in green and even another signal in blue is what I mean.

Mabula

@mabula-admin Thanks for your feedback! Good to hear I’m on the right track, looking forward to the next update 😅

Regarding the output into 3 monochrome channels, I was thinking this would be a convenient shortcut in the use case where data from both Ha-Oiii and Sii-Oiii filters were loaded at the same time, avoiding to have to reload and reprocess each data set separately, if that makes sense? 

So oiii would be processed from both sets, but Ha only from those with the appropriate filter and then the same for Sii. 
I assume this is not feasible as the ‘Extract Xxx’ performs different normalisation depending which narrowband is chosen? Wondering if some ‘intelligent’ solution is possible to apply the correct extract algorithm for different sets within the same process run 🤷‍♂️

Hi Zak @zakky2k, 

Ah, yes, I understand what you mean. For this to be possible, we need to make the 0) Raw/Fits settings be set for each image batch that you load into APP. Currently, the settings are global for all images. So to make that happen we need a bit more work, but this is for sure on our ToDo/Issue list, because it will make processing even more easy of course.

Mabula