2022-05-11: APP 2.0.0-beta 1 has been released !!!
Download links per platform:
New OSC Quad-Band filters on the market, how best to process them?
Hello, I have been seeing the latest multi-band filters for OSC come out on the market that are quad band (Sii, Ha, Oiii, and Hb) and thinking of getting one as an upgrade to my current tri-band filter (Ha, Oiii, Hb) I was wondering is there an algorithm that night work best so all the bands are pulled into the processed image or a way to extract the Sii, like we can currently do for the Ha and Oiii for a SHO palette? Would there be any plans on the horizon to create an algorithm to support these new filters? Thanks and clear skies!
I cannot speak for Mabula or APP, but let me give you my view on this as a user of STC-DUO and APP.
APP can separate HA and OIII because: HA is passed by the STC-DUO AND the red filter in the bayer matrix. There is no HA signal under the red and green filters of the Bayer matrix of the sensor. OIII is passed by the STC-DUO filter AND the green and blue filters of the sensor's Bayer filter. HA is not recorded by the green and blue filters. This is the way APP can distinguish between HA and OIII and separate them out.
Now for SII and Hb: Hb is green/blue and will be recorded together with OIII (through the green and blue Bayer filters) . SII is in the red part of the spectrum and will be recorded together with HA (through the red Bayer filters). There is no way with the bayer filters as is to distingish between Hb and OIII and/or SII and HA.
These quadband filters capture more line emisssion than the duo or tri filters and so gather more signal (although SII is very faint normally).
"There are two types of filter:
Tri-Band - combines Ha into a red zone, and H-beta + Oiii into a green/blue zone. (Recommended for high to severe light pollution).
Quad Band - combines Sii AND Ha into a red zone, and H-beta + Oiii into a green/blue zone. (Recommended for moderate light pollution). "
Note that there are two approaches here: The one as in the altair quad band filters where the two passbands are broad (I guess 35nm or even broader), so that they encapsulate 4 emission lines (two emission lines in each passband). The fact that these are broad passband explains why the manufacture says: "Recommended for moderate light pollution"
En then there is the narrow 4x bandpass from this manufacturer: https://optcorp.com/products/radian-telescopes-2-inch-triad-ultra-filter
Very impressive and very effective in any kind of light pollution, but have a look at the price!
I hope this helps
Thanks for the response and I think I'm following. So if I understand correctly, even if I use one of the new quadband filters, I could still use Ha-OIII Color, or choose to extract either Ha or OIII and in actuality it would extract Ha and SII in the case of the Ha and Oiii and Hb in the case of extracting Oiii? Based on the spectrum these bands fall under, Red or Blue/Green?
I'm leaning more towards the Altair Quad because I live in a Bortle 5 area and it's a much lower price point compared to the OPT. It seems to be a bit broader band pass which I think will be better for using with my RASA 8 and ASI294MC and ASI183MC OSCs.
"So if I understand correctly, even if I use one of the new quadband filters, I could still use Ha-OIII Color, or choose to extract either Ha or OIII and in actuality it would extract Ha and SII in the case of the Ha and Oiii and Hb in the case of extracting Oiii? Based on the spectrum these bands fall under, Red or Blue/Green?" That is correct.
Bortle 5 is not too bad. I live in Bortle 5 (or even 6, depending which way I look) and the STC DUO helps tremendously. But STC-DUO has two times 10nm bandwith, which is more restrictive and better against light pollution.
Using very narrow bandpass filters on the RASA8 is not a happy thing. From the OPT website: "Since the filter has extremely narrow bandpasses, telescopes with focal ratios faster than f/4 will move transmission off band. Telescopes with focal ratios between f/2.9 and f/3.9 are compatible with Triad Ultra, with lower transmissions. Telescopes with focal ratios faster than f/2.8 having diminished throughputs at the nominal wavelengths. Between f/2.2 and f/2.8 will have about 0.8nm band-shift." (On the Triad Ultra Quad-Band).
It all depends, but it might be that going from your current tri-band to a quad-band does not give a great improvement. But that you only know after experimenting.
Thanks for that in-depth information @annehouw! These filters I have no experience with, but if they fall in the same spectrum-bands it should be extracted with OIII-HA yes, not 100% sure though how APP separates those exactly though.. but I think you're correct as that makes a lot of sense. 🙂
Please do post your results if you decide to use the new Quad filter. I use the STC-DUO and while it is a great and affordable filter in my view, it does has internal reflections on bright stars. At least in my f/3 system.
Interesting topic, as I am about to buy one of these filters myself.
I'm wondering if the various bands could be separated on pixel values? All these, Hb, OIII, Ha and SII have a spesiffic vawelenght wich is represented by a spesiffic pixel value (RGB). If a pixel were selected based on this value, including pixels in a band to each side of the this value, these could be used to create a separate image.
For Ha, and SII, BG pixel values could probably be ignored, and for Hb and OIII the R pixel value could be ignored. Then it could be worked out which values of Red should be assigned to Ha and which to SIII. And similar for Hb and OIII bands.
We would then have four separate images to play with like if they had been captured with spesiffic filters.
I'm not sure if this could be done or if I have even understood how RGB values relates to frequency band/vawelenghts. Don't chop my head of if I have got it all wrong. 🤔
Isn't this basically what the options in tab-0 do? Under the "algorithm" list.
Yes, I suppose it is. But it only splits into two "bands", Ha and OIII as I understand it. I was thinking if these two could be split further into Ha + SII and Hb + OIII. Then you would have four different images to play with. Could further be used/combined into any palette.
Just a thought, may not be worth the efforth.
I have not used the existing options in tab 0, so I'm not sure how that works or what it does.
There are also SII options, but indeed separate. The options there basically do what you want, they separate Ha and OIII from something like duo-band filter data and create separate integrations automatically.
That is fine then. I will have a go at this as soon as I have got some data with such a filter. But first I need the filter.👍😁
Interesting topic! I bought the STC Astro Multispectra filter (Ha, Hb, OIII and SII) and recently gained data of several objects with this filter.
So, I'm curious how I can process these data in APP with the algorithms in tab-0. I've already used the great "extract Ha" and "extract OIII" algorithms.
But the idea of Heno, in his message of October 2, 2019 15:10, is the same idea that I had in mind and would like in APP...if it's possible.
I don't think it is possible, there is no information in the data to point to the different wavelengths, only R, G and B signal. Just like AnneHouw above explains. I think it's very hard, if not impossible to really know for sure what signal belongs to what band.
If I had the dual band filter and a separate SII filter, then I am assuming that I could get all 3 bands with just two separate images. Is this correct? Or would it be better to just use 3 different filters (using a OSC is assumed).
Yes that would work indeed. SII is also in the red part so it would then be best to check the overlap of the filters for that, otherwise it might work a bit less well.