Pixinsight - question regarding processing Pinwheel Supernova

[edarter]

Hi,

Back in January I gathered quite a bit of data on the Pinwheel galaxy and processed it, all good so far. When the SN kicked off though I revisited it and have gathered another 2-3hrs of data (unfortunately at a different camera rotation but thats another story).
I have two sets of integration, one with everything included (SN data as well) and one with just the SN data integrated. Both stacked in APP, no processing done.

My question is what would be the best way of processing this in order to maximise the SN data? I'm concerned that just processing the data where everything is integrated will have a reduced SN as it wasn't present in every frame and would therefore be calibrated out to some extent (it is still visible though - I've done a quick STF to check). This approach would of course give another 2-3hrs of the rest of the image though which can only be a good thing.

I usually separate the stars using Starnet 2 and did wonder about somehow adding the SN only star mask over the 'all included' starmask but a) not sure if thats a good idea and b) no idea how to do it!

Suggestions welcome

Ed

[Aramcheck]

With my 2022 & 2023 data I stacked the pre/post nova images separately & also as a combined set to maximise the signal to noise ratio. In Pixinsight I used LinearFit to make the sky background the same & applied the same stretch to each of the three masters.

I then made a mask of the supernova & used Pixelmath on copies the combined image to replace the pixels with a) the pre nova data & b) the post nova data. A slight curve adjustment was also needed hide the join...

Results looked ok, but when I later made a GIF (which only has 256 levels) the join is still a bit visible...

Cheers
Ivor

[tomato]

Top notch, the only thing I can see changing between the frames is the SN.

[edarter]

On 17/06/2023 at 19:35, Aramcheck said:

With my 2022 & 2023 data I stacked the pre/post nova images separately & also as a combined set to maximise the signal to noise ratio. In Pixinsight I used LinearFit to make the sky background the same & applied the same stretch to each of the three masters.

I then made a mask of the supernova & used Pixelmath on copies the combined image to replace the pixels with a) the pre nova data & b) the post nova data. A slight curve adjustment was also needed hide the join...

Results looked ok, but when I later made a GIF (which only has 256 levels) the join is still a bit visible...

Cheers
Ivor

Were all your sessions at the same camera rotation? Mine are at two different rotations so applying the same crop etc is something I'm not sure of doing.

[Aramcheck]

19 hours ago, edarter said:

Were all your sessions at the same camera rotation?

@edarter Yes - I used the image solved co-ordinates from the 2022  session and a framing mask in Astro Photography Tool to make sure they were the same orientation. What I didn't bargain for was that I'd had to rotate the OTA in the tube rings when I fitted an OAG... so the diffraction spikes were in different orientations between sessions. I made a mask of the pre-nova spikes from a copy of the Combined non-linear image & then substituted the pixel values from the post-nova image, and again used a curves adjustment to hide the joins.

If you look at the bright stars in the combined image (attached) the additional spikes have been removed reasonably well.

You can still try to incorporate data with vastly diffferent orientations by using the NormaliseScaleGradient script (in Script->Batch Processing menu). I've used that in the past to avoid the sharp changes in background sky values with data with different orientations. (I have the paid version of NormaliseScaleGradient).

Cheers
Ivor