Quick BlurXTerminator test on old data

[rustigsmed]

Hi all,

Thought I would check out some BlurXTerminator on some old images. Very quick, a couple of clicks and its done. Works well on M16 but in the Chicken it overdoes it on some of the boks. Noticed on shorter FL its much less noticeable (not shown here) than on longer FL scopes.

Note this is edited the old finished processed images and not doing it with the original raws.

Anyway interesting to see how it works - i'm sure everyone will eventually use it, but like other tools it can be easily over toasted.

Star reduction is good too, although I will need to better understand the halo setting.

Cheers

Russ

[RyanJones]

Very interesting Russell. It’s good to see a direct comparison with and without. I agree it has worked better ( more subtle ) on your Pillars than your Chicken. Looks like it certainly has promise inline with what I’ve heard elsewhere. Even more reason for me to jump from PS to Pixinsight.

[alpal]

I'm not sure about it -
it seems to make dim stars disappear especially on that chicken pic.


cheers
Allan

[gregbradley]

Good result.
These sample images though have had a stretch done?
That makes it nonlinear data and therefore not ideal for blurrxterminator which requires linear data like Startools does.

So if that is the case then you would expect better results if applied to the data before any curves are applied.
Greg

[Startrek]

Greg,
When I get time I might show some comparisons using the Startools Spatially Variant PSF Deconvolution which came out early last year in version 1.8
Ivo is currently working on version 1.9 which is supposed to be even better , more intuitive and less sampling involved
At present I getting great results with SV Decon 1.8 equal or better than this BlurXterminator thing and it usually takes me 5 or 6 mins
Spatially Variant PSF Deconvolution is performed as normal whilst tracking your data, true high fidelity signal processing ( it’s not an add on later )

https://www.startools.org/modules/sv-decon

Cheers
Martin

[Dilsh]

Thanks for that Russell,

Seems like this a must for PIs users. Think I will take the plunge as well.

Cheers,

Dilshan

[Dilsh]

Ok, So just tried it on a stretched highly processed image (I know this is incorrect).

Still the results speaks for itself.

In the words of the Colonel.

"take my money"

[rustigsmed]

hi Ryan, yes I am fairly new to pixinsight myself - i'm trying to force myself to spend the time in pixinsight but end up using a few programs at the moment!



cheers Allan, yes the chicken didn't work so well seems to be hit and miss.



Hi Greg,

Yes correct, these images have already been stretched - I haven't had a chance to try it - starting processing from the beginning just yet. it may end up with a more appropriate chicken.

definitely promising with minimal effort applied (m16).



Sounds like an interesting tool Martin - looking forward to the comparisons. It sounds like it might get more acceptance if there is no AI involved.





Glad the comparison was useful! Looks sharper definitely! :thumbsup: