Back to forums

How to add H alpha data to OSC data in Pixinsigh?

Optolong L-eXtreme ZWO ASI2600MC Pro Calibration Processing H-alpha One-Shot Color Deep Sky Galaxy M31 Messier Objects
8 replies•241 views
oymd avatar
oymd avatar

Hi everyone

I was wondering if someone can give me advice on how to add Ha data to OSC data in PI?

I use NINA for my data acquisition.

The camera is 2600MC Pro, which has UV/IR window.

I’ve taken enough data of M31 without using any LP filters, but would like to add some Ha data I have taken with the same camera and the Optolong L-extreme filter.

I have 10hrs total integration time on M31, and a further 5hrs taken with the lextreme.

Obviously both sets of data have separate flats and darks, as the unfiltered OSC data was taken using 120s subs, and the Ha were 300s subs.

At what point in the preprocessing stage do I combine the data?

Do I do it in WBPP, or preprocess the data separately and end up with two separate masters, and then combine in PI?

Would appreciate some help with this.

Many thanks

Ossi

Joel85 avatar
Joel85 avatar

Hi,

May be you will find a solution here :

https://astro-photographie.fr/traitement_pixinsight.html?fbclid=IwZXh0bgNhZW0CMTEAAR25m0gA0-dEE57fR6gMhz-MNIeDRtyhhv30qxaRXtsTzXeX7M6h-Y3M-wc_aem_pzSRLl090dGziJpQdT3eLA

(case : data preprocessed separatly / 1 integration = 1 filter or 1 shoot without any filter)

CS

Mikołaj Wadowski avatar
Mikołaj Wadowski avatar

You should end up with two masters, one for the broadband data (UV/IR), and another one with the narrowband data (L-eXtreme). You combine them during processing.

The first two steps should be gradient extraction from both stacks and extracting Oiii from the narrowband stack. Due to how bayer matrix works, it lets quite a significant amount of Ha through the G and B filters, so your initial “Oiii” isn’t pure. I explained how to do it in my emissionless processing tutorial (step 1, you can ignore the rest). Whether you extract the gradient before or after Oiii doesn’t really matter, but I like to do it before.

Once you’ve done that, you can follow this tutorial by Charles Hagen. They key step here is continuum subtraction, which is necessary when adding narrowband data to broadband images. The tutorial mentions two methods, but method 1. has since been superseded by Charles’ script.

Helpful Insightful
andrea tasselli avatar
andrea tasselli avatar
Mikołaj Wadowski:
They key step here is continuum subtraction, which is necessary when adding narrowband data to broadband images


Nope, not really necessary at all.
Mikołaj Wadowski avatar
Mikołaj Wadowski avatar

andrea tasselli · Jan 2, 2026, 06:28 PM

Mikołaj Wadowski:
They key step here is continuum subtraction, which is necessary when adding narrowband data to broadband images



Nope, not really necessary at all.

I guess if you like your continuum being incorrectly colored red/blue then it’s not needed, sure. When doing galaxy images without continuum subtraction you’re bound to end up with overly red cores, or in case of nebulae - discolored dust.

andrea tasselli avatar
andrea tasselli avatar
Mikołaj Wadowski:
I guess if you like your continuum being incorrectly colored red/blue then it’s not needed, sure. When doing galaxy images without continuum subtraction you’re bound to end up with overly red cores, or in case of nebulae - discolored dust.


No, not really...
John Nedelcu avatar
John Nedelcu avatar

I struggled to do this for ages with different techniques like continuum subtraction, but my last image (IC405) worked like a charm by following the guide from Dark Sky Geek.

At least, it’s miles better than anything I have been able to produce so far. There is a bit of fiddling with the numbers, but after getting it dialled in, it worked.

If you’re interested in trying, these two videos explain the workflow:

Helpful Engaging Supportive
John Nedelcu avatar
John Nedelcu avatar

Mikołaj Wadowski · Jan 2, 2026, 06:41 PM

andrea tasselli · Jan 2, 2026, 06:28 PM

Mikołaj Wadowski:
They key step here is continuum subtraction, which is necessary when adding narrowband data to broadband images



Nope, not really necessary at all.

I guess if you like your continuum being incorrectly colored red/blue then it’s not needed, sure. When doing galaxy images without continuum subtraction you’re bound to end up with overly red cores, or in case of nebulae - discolored dust.

I found the opposite for me. Granted, I’m not very good at CS, and all the guides I’ve seen don’t work with my data at all.

Mikołaj Wadowski avatar
Mikołaj Wadowski avatar

John Nedelcu · Jan 2, 2026, 06:54 PM

I struggled to do this for ages with different techniques like continuum subtraction, but my last image (IC405) worked like a charm by following the guide from Dark Sky Geek.

At least, it’s miles better than anything I have been able to produce so far. There is a bit of fiddling with the numbers, but after getting it dialled in, it worked.

If you’re interested in trying, these two videos explain the workflow:

This is a good tutorial. This is not continuum subtraction, but rather a separate step in processing OSC data. It’s effectively the same thing I talked about in the first part of my comment.

Mikołaj Wadowski · Jan 2, 2026, 06:19 PM

The first two steps should be gradient extraction from both stacks and extracting Oiii from the narrowband stack. Due to how bayer matrix works, it lets quite a significant amount of Ha through the G and B filters, so your initial “Oiii” isn’t pure. I explained how to do it in my emissionless processing tutorial (step 1, you can ignore the rest). Whether you extract the gradient before or after Oiii doesn’t really matter, but I like to do it before.

The only difference is I don’t use any sensor characteristics as a baseline for the factor, since in my experience I would have to play around with them quite a bit anyway

John Nedelcu · Jan 2, 2026, 06:56 PM

I found the opposite for me. Granted, I’m not very good at CS, and all the guides I’ve seen don’t work with my data at all.

With OSC cameras it can sometimes be tricky. Especially if you don’t CFA drizzle - debayering does not preserve photometry, so it may not work properly, especially if you’re using automated scripts like PSC for continuum subtraction.

Either way, given a set of broadband and narrowband masters, it’s always possible to perform continuum subtraction. If you don’t you end up with results like these:

To be clear, in my opinion these are all good images, but display a pretty glaring technical error.

Helpful Insightful Respectful Engaging