M83 - luminance

[codemonkey]

Finally captured more data (2hrs, 13mins) on M83 last night. I ended up ditching the focal reducer so I've not combined any of my previous data.

Guiding was really poor last night, not sure why. Maybe just seeing, maybe user error, maybe both. I definitely have some balance issues (dec) and that's contributing for sure, confusing PHD's calibration process. I think it still looks ok, stars haven't bloated too much...

Anyway, pretty happy with this. I need to figure out what the go is with blown out stars though, they seem to have a dark ring around the clipped core. Maybe this is anti-blooming at play? I should probably try and get some shorter exposures for the stars.

Now one day I'll actually get some colour on this. I've imaged it in mono several times now but I never end up getting around to the RGB. Hopefully this will be the last time before RGB :)

[rustigsmed]

lots of detail in there Lee, very nice, looking fwd to the RGB version!

Russ

[RickS]

Some nice detail, Lee. That's not the ABG doing that. Looks more like a processing artifact. Do you see it in the raw subs?

[codemonkey]

Cheers Russ :-)



Thanks Rick! I'm pretty confident that it's in the raw subs, thus my thinking. It's possible I inspected a sub that was calibrated or "cosmetically corrected" by mistake though. I'll double check when I get home.

[codemonkey]

Ok, you were spot on Rick, definitely a processing artefact. Wasn't there until registration. Changing interpolation algorithm to nearest neighbour removed the artefact.

Edit: Scratch that -- nearest neighbour lessened the artefact, it's still there but not as obvious until processing exacerbates it.

[RickS]

Can you put a raw FITS file somewhere it can be downloaded, Lee?

[codemonkey]

Thanks Rick, I can put one up tonight when I get home. I've managed to work around it by using nearest neighbour interpolation when registering the images, which make sit almost imperceptible, and then tidying up the (small) remaining artefacts that appear in processing using morphological transformation.

[codemonkey]

Well, turns out I didn't keep the stacked unprocessed image and now I can't replicate the results. Must have been a pebkac. Nevertheless, I've put up a stacked image (https://www.dropbox.com/s/2nm8khiitqka1i6/integration.fit.zip?dl=0) on dropbox if anyone is interested in having a play with it.

I also reprocessed it myself... better than the previous version I think!

[RickS]

Nice work, Lee. Hopefully, the problem is permanently solved :)