Blurxterminator… a game changer? - Forums

Topic starter

IOTD Staff

Dec 15, 2022 · Edited: Dec 16, 2022

In case you didn’t know this just came out:
https://www.rc-astro.com/resources/BlurXTerminator/index.php

im a big fan of Russells stuff… but this just takes it to a whole new level.

here is a long video of Adam Block going through it… he must say the word Blurxterminator over a 1000 times in the video so be warned 😂😂

https://www.youtube.com/watch?v=_rWEuW4MTrQ

If you have ever used deconvolution in pixinsight you would know that it is a painful process to get right and takes a long time… this does it in one shot 10x faster…

like with any process it is easy to over do though… However what I’m most impressed with is what it does to the stars. i’ve tested it a bit with some old data and any color aberration I may have had or small tilt issues, gone… see you later… and amazingly well done too… now this does not replace good optics of course, but for those of us who do not own an astrophysics refractor on top of a mountain in Chile with 0.5 seeing (one day that’s the dream…), this really helps…

have you had a test, what did you think…?

Engaging

Kevin Morefield

IOTD Staff

Dec 15, 2022

Oh yeah, it's awesome. For my images the default is wayyy too strong. I've been cutting the Sharpening down to 50 or lower. I'm a little oversampled and am using a long FL scope so it should have a big impact there. I think with my short scope where the FWHM is in the 2 pixel range it won't do as much for sharpening the object but the stars will like it a lot.

Steeve Body

Topic starter

IOTD Staff

Dec 15, 2022

Kevin Morefield:
Oh yeah, it's awesome. For my images the default is wayyy too strong. I've been cutting the Sharpening down to 50 or lower. I'm a little oversampled and am using a long FL scope so it should have a big impact there. I think with my short scope where the FWHM is in the 2 pixel range it won't do as much for sharpening the object but the stars will like it a lot.

Same, I have sharpening way way down. Anywhere between 30 to 60 for me

Rafał Szwejkowski

IOTD Staff

Dec 16, 2022

It has a tremendous impact on my Ritchey-Cretien images which require careful deconvolution due to FL and relatively large central obstruction.

Steeve Body

Topic starter

IOTD Staff

Dec 16, 2022

Rafał Szwejkowski:
It has a tremendous impact on my Ritchey-Cretien images which require careful deconvolution due to FL and relatively large central obstruction.

I can see how significantly more impactful this would be a longer focal length scope where the optics tolerances are so much tighter…

TurtleCat

Dec 16, 2022

I've already reprocessed a bunch of stuff with it. It's a game changer in my opinion. I know some people think that the mystic and secret art of image correction should be kept from the unwashed masses but I'd rather get the tedious stuff out of my way. The real craft, in my opinion, is on how and what you capture and how and what you present. It's not in what should be essentially routine stuff from calibration to basic corrections.

Engaging

Rafał Szwejkowski

IOTD Staff

Dec 16, 2022

Steeve Body:
Rafał Szwejkowski:
It has a tremendous impact on my Ritchey-Cretien images which require careful deconvolution due to FL and relatively large central obstruction.

I can see how significantly more impactful this would be a longer focal length scope where the optics tolerances are so much tighter…

That's absolutely true, for example this is what happens to @FL2800mm Maksutov Ha channel. It solves one of my biggest problems which was seeing-related blurring at such long focal lengths. (I feel as if I just carried my setup up a 4000m mountain). At short FL it has much less of an impact but what I found is that correcting stars in corners is a nice bonus regardless of FL.

I am a big fan of RC's work and I would at this point recommend anyone with a processing workstation to also invest in a nVidia video card (or M1/2 Mac) to speed up the eX line. Time and electricity cost savings are substantial.

Steeve Body

Topic starter

IOTD Staff

Dec 16, 2022

Rafał Szwejkowski:
Steeve Body:
Rafał Szwejkowski:
It has a tremendous impact on my Ritchey-Cretien images which require careful deconvolution due to FL and relatively large central obstruction.

I can see how significantly more impactful this would be a longer focal length scope where the optics tolerances are so much tighter…

That's absolutely true, for example this is what happens to @FL2800mm Maksutov Ha channel. It solves one of my biggest problems which was seeing-related blurring at such long focal lengths. (I feel as if I just carried my setup up a 4000m mountain). At short FL it has much less of an impact but what I found is that correcting stars in corners is a nice bonus regardless of FL.

I am a big fan of RC's work and I would at this point recommend anyone with a processing workstation to also invest in a nVidia video card (or M1/2 Mac) to speed up the eX line. Time and electricity cost savings are substantial.

Pretty subtancial!

Steeve Body

Topic starter

IOTD Staff

Dec 16, 2022

Here is how it fixed a chromatic aberation problem for me in the corner of one of my images

Jay Grevell

Dec 16, 2022

Steeve Body:
Rafał Szwejkowski:
Steeve Body:
Rafał Szwejkowski:
It has a tremendous impact on my Ritchey-Cretien images which require careful deconvolution due to FL and relatively large central obstruction.

I can see how significantly more impactful this would be a longer focal length scope where the optics tolerances are so much tighter…

That's absolutely true, for example this is what happens to @FL2800mm Maksutov Ha channel. It solves one of my biggest problems which was seeing-related blurring at such long focal lengths. (I feel as if I just carried my setup up a 4000m mountain). At short FL it has much less of an impact but what I found is that correcting stars in corners is a nice bonus regardless of FL.

I am a big fan of RC's work and I would at this point recommend anyone with a processing workstation to also invest in a nVidia video card (or M1/2 Mac) to speed up the eX line. Time and electricity cost savings are substantial.

Pretty subtancial!

I think you have overdone the sharpening on this, its quite clearly creating details that arent there? subtlety is the key with these applications.

Rafał Szwejkowski

IOTD Staff

Dec 16, 2022

Jay Grevell:
Steeve Body:
Rafał Szwejkowski:
Steeve Body:
Rafał Szwejkowski:
It has a tremendous impact on my Ritchey-Cretien images which require careful deconvolution due to FL and relatively large central obstruction.

I can see how significantly more impactful this would be a longer focal length scope where the optics tolerances are so much tighter…

That's absolutely true, for example this is what happens to @FL2800mm Maksutov Ha channel. It solves one of my biggest problems which was seeing-related blurring at such long focal lengths. (I feel as if I just carried my setup up a 4000m mountain). At short FL it has much less of an impact but what I found is that correcting stars in corners is a nice bonus regardless of FL.

I am a big fan of RC's work and I would at this point recommend anyone with a processing workstation to also invest in a nVidia video card (or M1/2 Mac) to speed up the eX line. Time and electricity cost savings are substantial.

Pretty subtancial!

I think you have overdone the sharpening on this, its quite clearly creating details that arent there? subtlety is the key with these applications.

That's your opinion, which I wasn't seeking or asking for in this thread. Default setting was applied for illustration purposes only -- using it as an opportunity to put down or assert expertise over other users was uncalled for on your part.

Well Written

Jay Grevell

Dec 16, 2022

Rafał Szwejkowski:
Jay Grevell:
Steeve Body:
Rafał Szwejkowski:
Steeve Body:
Rafał Szwejkowski:
It has a tremendous impact on my Ritchey-Cretien images which require careful deconvolution due to FL and relatively large central obstruction.

I can see how significantly more impactful this would be a longer focal length scope where the optics tolerances are so much tighter…

That's absolutely true, for example this is what happens to @FL2800mm Maksutov Ha channel. It solves one of my biggest problems which was seeing-related blurring at such long focal lengths. (I feel as if I just carried my setup up a 4000m mountain). At short FL it has much less of an impact but what I found is that correcting stars in corners is a nice bonus regardless of FL.

I am a big fan of RC's work and I would at this point recommend anyone with a processing workstation to also invest in a nVidia video card (or M1/2 Mac) to speed up the eX line. Time and electricity cost savings are substantial.

Pretty subtancial!

I think you have overdone the sharpening on this, its quite clearly creating details that arent there? subtlety is the key with these applications.

That's your opinion, which I wasn't seeking or asking for in this thread. Default setting was applied for illustration purposes only -- using it as an opportunity to put down or assert expertise over other users was uncalled for on your part.

Well I’m sorry I offered my opinion on the topic. Hopefully it is helpful to someone that appreciates it. Have a good day.

Well Written

Blaine Gibby

Dec 16, 2022

When can we expect this to be released to photoshop so us simpletons can use it?

Steeve Body

Topic starter

IOTD Staff

Dec 16, 2022

Blaine Gibby:
When can we expect this to be released to photoshop so us simpletons can use it?

I may be wrong but this may be a pixinsight exclusive...?

Rafał Szwejkowski

IOTD Staff

Dec 16, 2022

Jay Grevell:

Well I’m sorry I offered my opinion on the topic. Hopefully it is helpful to someone that appreciates it. Have a good day.

No problem, I am sure Russel Croman welcomes all feedback and the program "clearly creating details that aren't there" is the kind criticism he expects and it is for him to address.

Jay Grevell

Dec 16, 2022

Rafał Szwejkowski:
Jay Grevell:

Well I’m sorry I offered my opinion on the topic. Hopefully it is helpful to someone that appreciates it. Have a good day.

No problem, I am sure Russel Croman welcomes all feedback and the program "clearly creating details that aren't there" is the kind criticism he expects and it is for him to address.

Yeh I’m sure he expects all kinds of feedback. Mine was meant as constructive so sorry if it came across the wrong way. Like all sharpening algorithms it just needs trailering to each image id say. I’ve read a few with high end systems saying the default values are way to high for their images and it will vary user to user. It looks promising tho.

Gabe Shaughnessy

Dec 16, 2022

I've enjoyed the new plugin as well on both my high res iDK and moderate focal length refractors. The performance of this tool vs. the usual deconvolution procedure is night and day, both from a level of effort and final results perspective. You can more quickly iterate on different settings since it's so fast. It's also very nice to have independent control over the tightening of both stellar and non-stellar sources. I've found a little bit goes a long way.

Well Written Concise

Kevin Morefield

IOTD Staff

Dec 16, 2022

Blaine Gibby:
When can we expect this to be released to photoshop so us simpletons can use it?

Russell addressed this on the BDX page of his site. Because deconvolution must be done on linear data it can’t be used in Photoshop.

It would be a lot easier for me if I could use it after doing the LRGB combination but I can’t seem to work out a good way to combine Luminance with RGB in the linear state. But wow, is this product nice for OSC guys!

Helpful Concise

Matthew Proulx

Dec 16, 2022 · Edited: Dec 16, 2022

Kevin Morefield:
Blaine Gibby:
When can we expect this to be released to photoshop so us simpletons can use it?

Russell addressed this on the BDX page of his site. Because deconvolution must be done on linear data it can’t be used in Photoshop.

It would be a lot easier for me if I could use it after doing the LRGB combination but I can’t seem to work out a good way to combine Luminance with RGB in the linear state. But wow, is this product nice for OSC guys!

Interesting, I have never used Deconvolution in a linear state. I always do it at the end. Good to know.

Michael Ring

Dec 16, 2022

I really like what the tool does to galaxies, they go from Meeeh to pretty nice in a single step and the stars from my EdgeHD 9.25 go from ugly to kind of acceptable.
But although I reduced sharpening to 0.4-0.5 the results are sometimes a little to intense for my taste so in the end I do similar as with Topaz Denoise, I use masks in Photoshop to tame the parts that look overdone.
The only not so nice thing is that I have to stretch my data twice, one version with BlurXterminator and one without.

Michael

Helpful

Seung-Jun Kim

Dec 16, 2022 · Edited: Dec 16, 2022

I just attempted re-processing one of my previous image.

This is definitelly a game changer imo.

Btw, the image was taekn with GSO 10" RC and QHY268M

andrea tasselli

Frequent contributor

Dec 16, 2022

Let's cut the chase and make everything artificial and AI generated. At least we'd save on the glass expenses…

Wei-Hao Wang

Dec 16, 2022

On the Photoshop issue, there is some discussion about this on CN. There I argued that linear data are not absolutely necessary, and someone tried it on stretched images and confirmed what I said. Furthermore, you can absolutely work with linear data in Photoshop, using adjustment layers.

So this comes down to whether Russ wants to spend the effort to port it to Photoshop. According to himself (see the CN link, reply #60), the usage of Photoshop is minority, so he doesn't plan to do it, unfortunately.

All that being said, the tool looks quite amazing. I do use PixInsight. So his decision of not porting it to Photoshop does not affect me. For those who magically do not use PixInsight at all, I think this is the time you start using it. In my opinion, this IS a game changer.

ps, there are lots of before/after comparisons these days. If you see a comparison that's not presented at 100% pixel scale, forget about it. Everything you see there can be achieved with careful sharpening. The true power of this new tool can only be appreciated and set apart from skillful sharpening when viewed 100%.

Helpful Insightful

Stuart Taylor

Dec 16, 2022

this is great news! I too struggle with deconvolution, but I really like StarXT and NoiseXT from Russell, so I shall be trying this for sure!
Thanks for letting us know

Fritz

Dec 16, 2022

The indroduction of AI to astrophotography is a game changer for sure. To the better or worse is at the users hands as with all tools.

I have been playing around now with some of my bad data and as far as I can see now this tool can't save bad data. I have been using deconvolution in the past and will do in future. I have to play with this tool a lot more to see if for me the convenience and the gain in quality is worth the expense. For me processing is rewarding on its own and does not have to be convenient.

CS Fritz