New OSC Narrowband Processing Technique

CLICK ON THESE AFFILIATE LINKS TO SUPPORT THE CHANNEL:
* Agena Astro: bit.ly/487tmnj
* High Point Scientific: bit.ly/49z9qdv
* Amazon: bit.ly/darksky...
OTHER WAYS YOU CAN SUPPORT THIS CHANNEL:
* Send a donation via PayPal: www.paypal.com...
* Click on the "Thanks" button in the toolbar located underneath this video.
* Join the Dark Sky Geek Community Discord server: / discord
* Like, subscribe, comment on this video, tell your friends about it, etc.
DESCRIPTION:
In this video, I present a new, innovative processing technique for narrowband targets when using an OSC (One Shot Color) camera.
Music credit: Scott Holmes Music

[@darkskygeek]

I posted a follow up video on my channel which discusses some important details that I unfortunately skipped over in this video. Check it out: kzbin.info/www/bejne/Z2rRfJh8hdpmbtU CLICK ON THESE AFFILIATE LINKS TO SUPPORT THE CHANNEL: * Agena Astro: bit.ly/487tmnj * High Point Scientific: bit.ly/49z9qdv * Amazon: bit.ly/darkskygeek-amazon-affiliate-link OTHER WAYS YOU CAN SUPPORT THIS CHANNEL: * Send a donation via PayPal: www.paypal.com/donate/?hosted_button_id=49UXY8F6VVYFA * Click on the "Thanks" button in the toolbar located underneath this video. * Join the Dark Sky Geek Community Discord server: discord.gg/Rg7JYF8nHw * Like, subscribe, comment on this video, tell your friends about it, etc.

[@stephen2615]

Thanks. This looks interesting. I will have to explore the possibilities. Most people who do this tend to focus more on stretching than anything else.

[@janelubenskyi1177]

Looking forward to your excellent videos in the future….

[@DSOImager]

Nice way to get the most out of your OSC with a dual NB filter. NIce shot of the crescent. CS!

[@Robservatory]

Great looking image you made there!

[@TheOriginalTommo]

Have just this week started capturing my first narrowband data using the same camera as you and the Antlia ALP-T filter. Glad I found this video; very helpful thank you. I love PixelMath 😆 Subscribed 👍

[@outdegree]

Thank you for the tip! I will give it a try and reprocess my Eagle and Omega I captured recently. And greetings from Alameda ;)

[@aerozg]

Nice. Thank you for showing it's possible. I will try this when i get my L-Quad Enhance filter. Cheers!

[@Nabby13]

The most important part, subtracting R from G to get Oiii, is incorrect. R ~ 0.8*Ha , G = 0.92*Oiii + 0.16*Ha ====> 0.92*Oiii = G - 0.16*Ha = G - 0.2*R. You are only subtracting 0.1*R. Your correction for the blue channel is also incorrect - 0.5*Oiii = B - 0.05*Ha = B - 0.0625*R.

[@darkskygeek]

Hi Oleg, take a look at the follow up video I posted on my channel, and you’ll see how I get the final PixelMath expressions. There is some method to my madness 😂 Thanks for watching!

[@janelubenskyi1177]

Thank you ….I was looking for a starting point in pixelmath for the Crescent Nebula ❤️😊

[@rafaelesther]

Thanks for helping a newbie like me

[@davidaylsworth8964]

Interesting video. I’m also using an ASI533MC Pro and L-eXtreme with a variety of optics. I’ll need to try this approach sometime.

[@zippis333]

Same here :)

[@peterlaubscher3989]

Many thanks, super helpful, great image. Will be giving this a go soon. Clouds, are you listening?

[@tullyfisher]

Just found your channel. Great and helpful video. Subscribed! :)

[@gregmckay666]

2/3rds of a great video. You were going through the process so well untill you get to 07:42 where you start the channel combination. Then you just sort of show the results of the steps without explaining how you did those steps. I guess you did this to save time but for some of us that do not understand the steps, we are lost at that point. I'd rather a 30min video that is thorough than a 10 minute video that skips important details. Still, I'm subscribing, hoping for more details down the line.

[@darkskygeek]

Yup, totally agree, and that is why I posted a follow up video shortly after to go over some of these points. My apologies for not including these important bits in the first video.

[@jeffratino5456]

Thanks for the video. When I try the run PixelMath to generate the OIII or Ha channel, I get an error message: "Couldn't execute PixelMath instance: There is no active image window" What am I missing??

[@darkskygeek]

Are you clicking on the blue square at the bottom left of the PixelMath window? Also, make sure you select “Create new image” (see the video, you might have missed this small detail) Hope that helps!

[@astrocatinfo]

Great video and nice approach. How do you compare these results to apply an Image Integration in PIX or similar? I see your point according the bandpasses, but what about SNR? If you apply PixelMath I guess you won’t get an improvement in SNR… and you have 2 channels (G & B) gathering [OIII] data… that could improve the SNR

[@textandtelescope8199]

That is great!

[@OnceShy_TwiceBitten]

How would you add Sii data to this? if you say used the Sii filter from optolong? (which I am doing)

[@darkskygeek]

It’s not possible using this technique, which is reserved for HOO images.

[@macubasu]

Is this applicable to SIRIL? What software are you using? There is no hint anywhere.

[@darkskygeek]

I do mention several times in that video that I am using PixInsight. I do not use Siril so I am afraid I cannot help with that.

[@steveby14]

Would you be able to give me the equations. I can’t read from the video. Thank you

[@darkskygeek]

Things will become much clearer if you watch the follow up video 😉

[@mammouth2727]

Génial.

[@galactus012345]

Hi, brilliant to isolate the wavelegnths signals. How can I get the same RGB starting equations for my qhy268c ? Same imx571 sensor but not same manufacturer

[@darkskygeek]

Hi! I published a follow up video on my channel to explain how I derived the PixelMath expressions, check it out, it should help. You can start with the same expressions and tweak them, your sensor is actually pretty similar to the one in the ASI533MC Pro. Thank you for watching!

[@deepd4393]

Excellent !! Explained very simply. Can we add separate SII data using the SII filters and mix this OIII and HA data? 9:03 you said, you captured RGB of the stars, do we need to use a separate filter for that? Could you please how to do that?

[@darkskygeek]

Unfortunately, you won’t be able to separate SII. In most cases, it’s ok. In my experience, SII is really valuable only on a fairly small number of targets. H-alpha and OIII will cover most narrowband targets well because very few narrowband targets emit strongly enough in SII to get a clean signal. If you really want to isolate SII, I am afraid the only solution is to use a mono camera with an SII filter. As for the RGB stars, since this entire video is about using an OSC camera, all you have to do is to use a luminance (UV/IR cut) filter.

[@deepd4393]

@@darkskygeek Got it. thank you so much for ur reply

[@Emil-se2er]

@@darkskygeek But why not collecting and extracting H-Alpha and OIII from dual band filter image and collecting SII signal in the ordinary way with SII filter and combining all afterwards. Probably after linear fit? Are there enough targets with strong SII signal is another story.

[@zippis333]

Very nice processing technique. May I ask at which step of the process did you stretch the image? Thank you.

[@darkskygeek]

Hi Jan! The purpose of this video was not to show the full processing workflow. But good news: I have a series of videos ready to go that will cover that. They are part of what I call “The Bubble Nebula Project” (I already posted the first video in that series) Hope you like it!

[@zippis333]

@@darkskygeek Thank you for the answer. I will check your other videos too. And I understand that in this one you did not want to go through the whole process. I was just curious about the moment in the process when you stretch. Hope I will learn it in the Bubble project video :)

[@fernandomauleon1]

Two great videos! However, I can't see how to arrive at the formula to add OIII_G and OIII_B and obtain OIII. I use an ASI 183 mc pro camera, and the tables give me values ​​that differ from yours by approximately 2%, not much, but I would like to work with the greatest possible precision so that errors do not accumulate. Thank you very much

[@fernandomauleon1]

Sorry, I think it's already answered in part deux...

[@darkskygeek]

Yes, see around 5 minutes in part 2 :)

[@jimcollins1922]

Great instructional video. I am going to have to try that out. Is it possible for you to post the pixel math equations? Also, I have the L-Ultimate with the ASI533C, are the equations you showed valid for it as shown. I understand that I may have to tweak things.

[@darkskygeek]

The instructions should work fine for the L-Ultimate, but since no two filters are perfectly equal, expect to do a small amount of tweaking.

[@WArockets]

Do you think this technique would allow splitting of Ha and NII signal when using a 5nm dual narrowband filter? It's really the only thing I'm missing since I've got both of the ALP-T filters so Ha, OIII, SII and Hb. There are good planetary nebulae that show good division between Ha and NII if imaged with separate 3nm filters

[@darkskygeek]

Highly unlikely! Also, NII emits across 2 spectral lines that are on either sides of Ha. Check out my recent intro to spectroscopy (part I) in which I show the high resolution spectrum of a planetary nebula.

[@darle94]

Bonjour à toi Peut-on utiliser les données que tu utilise dans pixel math pour une caméra zwo 183 mc pro . on te remerciant de tes vidéos super intéressante . Etant débutant en astrophotographie ,je les trouvent d'une aide précieuse . bien à toi et bon ciel . Patrick ( France )

[@darkskygeek]

Salut Patrick, Non, il faudra que tu détermines les paramètres de ta caméra par toi même, en utilisant les astuces qui sont dans cette vidéo et dans la deuxième partie. Il faut essayer et voir ce qui marche. Bon courage!

[@darle94]

@@darkskygeek merci à toi de ta réponse rapide. Je te souhaite bon ciel. Bien à toi Patrick

[@deniffler1946]

hi! Thank you for a great lesson. Please tell me how to combine then On and OIII using pixel math? What formula should I use? Thank you in advance for the answer

[@darkskygeek]

Sure! I did not cover that because it is explained in many other videos or web sites, but here is what I did. Open PixelMath, uncheck the option "Use a single RGB/K expression", and then, use this (assuming your H alpha image is named Ha, and your OIII image is named OIII): Red: Ha Green: (0.9 * OIII + 0.1 * Ha)

[@Emil-se2er]

Hi man. I have to admit your corrections to the image are impressive. Now please let me start asking the questions. There are a few things left uncleared in the video and in the CloudyNights discussion. 1. About the whole idea. What you are doing is getting OIII and H-Alpha signals out of the camera RGB interpretation of the dual band filter data. Theoretically your results should be identical to taking two images with mono camera and OIII and H-Alpha filters and combining them in HO image. Is this a correct statement? The whole purpose of this is to learn how to collect correct narrow band data two times faster by using single shot color camera and dual band filter. Is this a correct statement? 2. I am really lost with your equations and especially with the last one about OIII. Please explain better how exactly you derived this equation. The reason I am asking is because we use different cameras and if we need to tweak the parameters, we will need to know their relation. According to the math two equations should be sufficient for solving two variables - OIII and H-Alpha. Theoretically equitation 1 and equation 2 should be sufficient for solving H-Alpha and OIII. Same for any other two channels. It is not clear to me how exactly and why you are mixing up three equations for solving two variables. Please give us more of the evaluation of the last OIII equation so I could understand better. 3. Lastly, I didn't understand how exactly you combine OIII, and H-Alpha images. I was expecting this to be done with Channel combination, but in the video, you said with PixelMath. If it is with PixelMath what then what is the proportion? Did you just sum them 1:1? This would probably give us a lot of saturation in the final image. Please explain. Thank you, Emil

[@darkskygeek]

Hi Emil, Have you had a chance to view the follow up video I posted on my channel? It answers some of your questions. If anything is left unanswered, please post your question in the comment thread of the subsequent video. It’s easier to keep track of things that way. Thanks! And thank you for watching!

[@darkskygeek]

Here is a quick summary: 1. Not exactly. This video is for people who already have a OSC camera and a dual narrowband filter, and who would like to try a different method to process their narrowband data. Also, with that equipment, you don’t collect narrowband data twice as fast as mono. Actually, it’s a little slower than mono, although not as much as what people might think. There are several great CloudyNights posts on this topic. 2. The derivation of the PixelMath expressions is covered in much greater detail in my follow up video. Please check it out. 3. I also covered that in my follow up video. Also note that there are tons of ways to do this, I just present my way of doing things. Let me know if anything else is unclear.

[@Emil-se2er]

​@@darkskygeek Thank you for your prompt reply. I will definitely check the second video and if I still have questions will post them there. I am imaging with L-enhance, which is wider than L-extreme, but because I see the improvement with your method, I will try it with L-enhance too. I already found the curves for my ASI294 and will start taking the coefficients from there. You said that you took your photo in San Jose. If you are part of SJAA we probably already met somewhere in the dark. :)

[@AntonioPena1]

Hello thanks for the math, make sense, I noticed in your video you did not explain how you integrated the final image using the OIII and HA, You used RGB or LRGB?

[@darkskygeek]

Hi Antonio, there is a follow up video on this channel which goes over a few details I had left out in this original video. Please watch it. Hopefully, it clarifies a few things. Thanks!

[@cryhavoc38]

8:39 how do you use pixelmath to combine the images?

[@darkskygeek]

Check out part II. Also note that I have a much more in-depth processing series titled “The Bubble Nebula project” on my channel. Hope this helps!

[@flexable9256]

Nice - just a quick comment, in your PixelMath formulas you don't seem to take into account that Green pixels are twice as many as Blue and Red ones in a Bayer matrix. Why not?

[@darkskygeek]

Hi! Good question! The fact that there are twice as many green pixels as blue pixels only influence the signal to noise ratio, it does not affect the intensity of the signal, and therefore the value of each pixel after debayering. Hope this helps, and thank you for watching! Make sure to check out the part two of this video, which I published to my channel a while back. More processing videos (better produced and better explained) are coming soon.

[@ulfgranlund5900]

These equations seem to give the impression that we could reliably separate OIII and Ha signal from a single debayered OSC camera image. I don't see how that can be possible. There will always be a mix of photons of different wavelenghts giving signal in all the R,G and B pixels. Just assuming that their spectral distrubution is governed only by the spectral sensitivity of the sensor is by no means the whole thruth. It will of course also depend on the spectral distribution of the target. So what is done here is only an approximation, probably in most cases a reasonably good one, but it is not the whole truth. If I make a thought experiment and imagine that I am imaging a monochromatic target that emits photons only in OIII, or only in Ha, the equations given here will still produce some signal in the other band, wich in that case obviously is not correct.

[@darkskygeek]

If you have signal that is only in OIII or only in Ha, you will most definitely have signal in all 3 bands because of the transmission characteristics of the bayer matrix. These equations and this methodology is likely the most accurate out there. A simpler version is to assume that 100% of the red signal is Ha (see Bray Falls’ latest video) That works okay too, but it’s just not as accurate, and on the crescent nebula, it will cause undesirable artifacts. Clear skies!

[@haoyangmi3579]

Very neat idea, thank you! May I ask how did you integrate the Ha and Oiii channel back to color image? Using HOO format?

[@darkskygeek]

Very good question! I already posted a follow up video on my channel, and I answer that specific question (among other things) Check it out! Hope this helps :)

[@haoyangmi3579]

@@darkskygeek This is very helpful! Thanks!

[@jimwaters304]

Good process. I suggest you revise this video to include the information in the CN post. There are too many steps you skipped over.

[@darkskygeek]

I’m way ahead of you on that one 😁 A couple of months ago, I posted a follow up video on my channel, check it out! Thanks for watching :)

[@massimo541]

👏👏👏🤩

[@evemitchellanderson]

Why not just use mono and be done with it 😅

[@darkskygeek]

Hehe! Yup, that’s what I did, but a lot of people have an OSC camera, for a variety of reasons, and this technique may be useful to them.

[@j.s.3407]

@@darkskygeek How a comparison between the mono data and this processes synthetic data extracted from OSC camera?? That would be a GREAT video.