How do I avoid this?

Hard to gauge the quality of the stars off of a processed image, there appear to be some convolution artifacts and strange clipping effects going on here. Upload a raw image with only a stretch to get a better look. And for smaller stars and better clarity, dial in your focus, your autoguiding, and get as much data as you can.

I can't seem to save as a .raw file but here's a jpeg of that data with only a STF applied to it...



...and here's the same but from a single sub

Thanks Andy! What were you using for that? I've used Starnet++ before now but I think that's more for getting rid of stars entirely, and I've never really got my head around masking. Was it one of those techniques?

FWIW:  this is what your Rosette nebula photo looks like with smaller stars.  You have some really nice captures that with a tiny bit of re-processing would be absolutely amazing:

Hi Piers,  

I think the advice above from Andy is spot on.  I would only add a couple of things I have learned, starting out in this hobby only two years ago.

1) Nothing beats good signal-to-noise.  You don't say what Bortle Sky you are observing from, but if you live in Cumbria, I would hazard a guess it is likely to be Bortle 3 and above.   I also live under Bortle 3 and I now wouldn't give Markarian's chain anything less than 3hours, even with my Hypergraph 8 at f/3.2.   [Earlier I used to dash around the sky at 1-2hour exposures of as many targets as I could fit in - but since then, I have gone back to most to get better data.]

2) Post-processing is critical.  Along with longer integration times, the biggest thing that has improved my data has been better post-processing with PixInsight.  In that regard, the EZ suite of scripts in PI are *really* helpful, particularly EZ Decon, EZ DeNoise and EZ StarReduction.  My image post-processing has undergone a huge leap forward using these techniques.  There is a fabulous couple of tutorials posted in just the last few days on Astrobin by @Patrick Cosgrove on deconvolution in PI and the EZ Decon script.    I can't recommend them highly enough, or indeed the use of the EZ suite of to do deconvolution, noise reduction and StarReduction.  Finally, I did note that, to my eye, your images had quite a green cast to them.  Have you used SCNR in PI to tone this down?  I tend not to remove 100% of the green with SCNR, but my personal preference is around the 70-90% mark, depending on the image.

Hope this helps.

Brian

<deleted>

Walter Leonhard Schramböck:
At first I always thought that my camera was bad, then that my telescope was bad, meanwhile I realized that bad seeing and/or bad guiding is mostly responsible for bloated stars. My tracking device is very used and has never been maintained, and my autoguiding equipment is not top-notch either. That's why I often struggle with giant stars in post-processing. The more focal length, the worse it gets with bad guiding.

That could well be the case here too - the jet stream is a frequent visitor!

That's a great feature! I do need to spend time getting to grips with Pixinsight and with the eternal twilight just around the corner, it'll give me something to do in the summer!

Sorry - it's an ASI533MC Pro. Having had a quick look on Cloudy Nights perhaps it would help when using the FRA400! I've always put the much "nicer" stars from my R200SS down to just one of those things.

If I got something like an Astronomics L2 filter, would it matter where in my imaging train it was placed? Also, if I was using my NBZ filter, would I still want a UV/IR filter in place?

Finally…I assume my R200SS doesn't need one? It does have a Vixen Coma Corrector 3 in it but I think it's a very gentle intervention!

Thanks for that. Seeing as I share the camera between the two scopes the less I have to change things the better, so if it will work with both, perfect. 

I can’t quite understand why so many people have their DSLRs modified to remove the IR filter thing but now I have a dedicated camera, I would benefit from adding one!