Thanks for your comparisons. I installed the latest version of PI yesterday, and I have also compared the results of GraX and PI. Much like your results, it very much depends on the image. I got better results with GraX on some, and better results with PI on others. I think that the take-away is that it is great that we are now having multiple options to work with our images - and it's a good thing.
I fiddle with the percentage slider on GraXpert when using it inside PixInsight and found that one setting is not useful for the different targets. One is often too much regardless so I start at 0.4 and experiment. As far as nebulousity, if it's faint, I don't care about it. Call me jaded but not grumpy.. Keep up the good work.
Thanks for the video. I think seeing the subtracted gradients is crucial. Using Graxpert on images of the LMC and Rho Ophiucus it totally stripped out the nebulosity - the subtracted gradient looked exactly like the target, not good. Resorted to using the LP tool in APP, that did a very good job, slower with boxes but a good job. Hopefully the new gradient correction tool will be better than Graxpert.
Nice video Shawn. Once again, proving my thoughts. GraXpert is great for galaxies and planetary nebula but in my opinion and your video it shows that GraXpert destroys too much of the nebulosity in images where there is a lot of coverage in the frame. You didn't spend much time on the background model from GraXpert but it clearly shows all of the nebulosity being removed. My recommendation for folks is still to use DBE manually on this type of image.
Thanks for this... I saw the new tool release and told some friends that I'd be doing a side by side comparison like this. I think you said it at the end... use both (neither take long to run) and see which one produces the superior result for the particular image you're processing. So, will these tools replace the ABE/DBE? I really disliked the manual placing points of and all the adjusting with DBE but it did give good results.
I played around with it today and found that increasing the High Threshold value took care of a very similar gradient I had on an M42 data set. I ended up at a value of 0.30. With all the adjustable parameters in this new tool I think it will be much more controllable than GraXpert in it's current version. I searched for those videos on the Pixinsight site today and couldn't find them. Do you have a link?
At least for me, applying Gradient Correction to the galaxies I have photograph ( in a Vortel 4 area, but more polluted that it seems ); and applying all the technics they describe at their own tutorial videos, I have to say that GraXpert keeps giving me a better result. GradientCorrection tents to overcorrect all the time around the galaxy, adding black areas ( and applying different combinations of parameters and values ). In the other hand, GraXpert provides a worst overall result but doesn't add black areas around the galaxy, as GradientCorrection does in my images. I will try with nebulas, but I'm not sure it is a better solution for now. May be when MARS is out next year, things change ( but GraXpert will evolve too, so may be it's too late for PixInsight GC to win this "battle" ).
Thanks for the video, Shawn! The stripping of nebulosity is my #1 concern when using GraXpert
Thanks Shawn. Been a long time, have gravitated away from astro to birding photography over the past 18 months.
Excellent comparison Swan... thank you very much.
Updates this morning and I’m loving it.
Thank you. Good quick review. No more DBE
Totally oblivious to the new gradient tool in PI, so thanks Shawn. Nice comparison.
Looks like the new corrections tool did a better job at preserving the M51 colors than what Graxpert did.
Thanks for your time Great job as usual
Shawn, another terrific video. Many thanks.
Thanks for your comparisons. I installed the latest version of PI yesterday, and I have also compared the results of GraX and PI. Much like your results, it very much depends on the image. I got better results with GraX on some, and better results with PI on others. I think that the take-away is that it is great that we are now having multiple options to work with our images - and it's a good thing.
I fiddle with the percentage slider on GraXpert when using it inside PixInsight and found that one setting is not useful for the different targets. One is often too much regardless so I start at 0.4 and experiment. As far as nebulousity, if it's faint, I don't care about it. Call me jaded but not grumpy.. Keep up the good work.
I my testing the new tool affects nebulosity less so it gets my vote
Thanks for the video. I think seeing the subtracted gradients is crucial. Using Graxpert on images of the LMC and Rho Ophiucus it totally stripped out the nebulosity - the subtracted gradient looked exactly like the target, not good. Resorted to using the LP tool in APP, that did a very good job, slower with boxes but a good job. Hopefully the new gradient correction tool will be better than Graxpert.
Seems like the pixinsight background extraction maintained the color on M51 a bit better than Graxpert? Looks like the reds and blues are still there.
Nice video Shawn. Once again, proving my thoughts. GraXpert is great for galaxies and planetary nebula but in my opinion and your video it shows that GraXpert destroys too much of the nebulosity in images where there is a lot of coverage in the frame. You didn't spend much time on the background model from GraXpert but it clearly shows all of the nebulosity being removed.
My recommendation for folks is still to use DBE manually on this type of image.
Agree. I have abandoned GraXpert totally. DBE is a bit of a PIA but does a better job.
Love these videos 😀
Encore un excellent script…!et rapide.fait bien le job…!.
When Shawn is grumpy, he pulls out the bag of Ketchup flavored chips and feeds them to Dan
Thanks for this... I saw the new tool release and told some friends that I'd be doing a side by side comparison like this. I think you said it at the end... use both (neither take long to run) and see which one produces the superior result for the particular image you're processing. So, will these tools replace the ABE/DBE? I really disliked the manual placing points of and all the adjusting with DBE but it did give good results.
Whenever I use GraXpert it takes away some nebulosity. Siril works better for me.
i update pixinsight but i don't have the gradientcorrection tool in my process list, someone have some idea to fix my issue ?
I played around with it today and found that increasing the High Threshold value took care of a very similar gradient I had on an M42 data set. I ended up at a value of 0.30. With all the adjustable parameters in this new tool I think it will be much more controllable than GraXpert in it's current version. I searched for those videos on the Pixinsight site today and couldn't find them. Do you have a link?
At least for me, applying Gradient Correction to the galaxies I have photograph ( in a Vortel 4 area, but more polluted that it seems ); and applying all the technics they describe at their own tutorial videos, I have to say that GraXpert keeps giving me a better result. GradientCorrection tents to overcorrect all the time around the galaxy, adding black areas ( and applying different combinations of parameters and values ). In the other hand, GraXpert provides a worst overall result but doesn't add black areas around the galaxy, as GradientCorrection does in my images. I will try with nebulas, but I'm not sure it is a better solution for now. May be when MARS is out next year, things change ( but GraXpert will evolve too, so may be it's too late for PixInsight GC to win this "battle" ).
J’aime bien graxpert…aussi.
Nice
Hi
Does this works with osc images?
Yes it does.
Curious, I installed 1.8.9-2 and there is no Gradient Correction in the process list. I see @axelp1972 also does not have it with the latest update.
Update: I am wrong: I was out of date. I thought the "1.8.9-2" part was the version number. I failed to notice the date suffix.