BlurExterminator 2.0

[vlaiv]

5 minutes ago, tomato said:

I suppose the problem is we can’t use the own eyes test on most Astro based images. 

We really don't need to. We have enough understanding and have good working theory of color and light that we can be certain we can process even astronomical images realistically.

That does not mean that we should do that and in most cases we need to "augment" them to show detail which would otherwise not be visible if we matched processing to what our eyes would see. We can however be selective in the way we apply that augmentation - for example we can simply elect to represent image as it would be if light was brighter than it is - either by having brighter sources, or sources being closer to us.

All of that is not mandatory for making of nice image, but it is there should one like to utilize such an approach.

[ollypenrice]

AI 4 is a disaster on our RASA data and I have gone back to 1. We have elongated stars on one edge and it turns them all into doubles. Yes we should fix it but we want to continue a number of projects before disturbing the rig.

Since Sxt gives control over star size I don't feel the need for it there but do like the nebula sharpening. BTW, another sharpening tool I like is, of all things,Topaz denoise!

Olly

[Stuart1971]

6 minutes ago, ollypenrice said:

AI 4 is a disaster on our RASA data and I have gone back to 1. We have elongated stars on one edge and it turns them all into doubles. Yes we should fix it but we want to continue a number of projects before disturbing the rig.

Since Sxt gives control over star size I don't feel the need for it there but do like the nebula sharpening. BTW, another sharpening tool I like is, of all things,Topaz denoise!

Olly

It really is only your settings you are using in BlurX that does that Olly, I had the same issue, try the correct only option, that will solve it…👍🏻

[ollypenrice]

1 hour ago, Stuart1971 said:

It really is only your settings you are using in BlurX that does that Olly, I had the same issue, try the correct only option, that will solve it…👍🏻

Thanks, I'll try that.

Olly

[Stuart1971]

2 hours ago, ollypenrice said:

Thanks, I'll try that.

Olly

If you do need some sharpening too, then use the normal setting and not correct only, but drop the star sharpen setting way down from the new default of 0.5 to what it was before, 0.25, and that also will stop the double star issue, or you could drop it even lower if needed….not sure why they have changed the default from 0.25 to 0.5

[ollypenrice]

1 hour ago, Stuart1971 said:

If you do need some sharpening too, then use the normal setting and not correct only, but drop the star sharpen setting way down from the new default of 0.5 to what it was before, 0.25, and that also will stop the double star issue, or you could drop it even lower if needed….not sure why they have changed the default from 0.25 to 0.5

Great, thanks. I've just fixed a physical fault on my processing PC (a miracle!!) so I can give this a go now.

lly

[Horwig]

This might be OT, but I got into BXt quite late, (not really done much imaging over the last couple of years, too much cloud.) I'm impressed with what it can do, I measured some images before and after, using PI's FWHMExetricity script. On my long FL scope I image at 0.5 arc sec/pixel, PI shows FWHM of 3.6 on one image before BXT, and 0.8 after correction, that is an amazing difference.

I've managed a couple of images over the last few weeks, M33 and The California Nebula, and whilst I'm impressed with the sharpening on the galaxy, the sharpening on the nebula was not that great. Anybody else seeing this, or am I just rusty and out of practise?

Huw

[tomato]

Here is a very brief (48 min) NGC 1499 taken with the RASA8/QHY268c/NBZ rig, imaging at 1.94 arcsec per pixel, with BXT and without BXT. It makes a difference but works much better, I think, on images with a lot more integration.

Without BXT

 

With BXT

[Stuart1971]

Both nice images, but the second one is deffo dimmer and less contrast…

[Horwig]

4 hours ago, Stuart1971 said:

Both nice images, but the second one is deffo dimmer and less contrast…

Hmmm, I think Stuart has a point, after copying both images into Photoshop, equalising fluxes with curves, I can see hardly any difference, will investigate more in the morning

[tomato]

No it isn’t much, Adam Block commented on one of my posts on here for the need for lots of integration to get the most out of BXT. 48 mins definitely doesn’t cut it.

[Stuart1971]

28 minutes ago, tomato said:

No it isn’t much, Adam Block commented on one of my posts on here for the need for lots of integration to get the most out of BXT. 48 mins definitely doesn’t cut it.

Yes I agree, I have tried it on a single sub, as an experiment, and it’s does an OK job, but then when you run on the stack, it’s so much better…

Edited January 14 by Stuart1971

[ollypenrice]

It largely depends on the target, surely?  Some don't have much fine detail, some have lots. I'd call the California relatively low on small scale detail at shorter focal lengths.

Since I remove and replace stars anyway, and give them an entirely different stretch from the subject, I find myself using the standard stars over the BlurX ones but I like BlurX on fine target detail.

Olly

[Horwig]

4 hours ago, ollypenrice said:

It largely depends on the target, surely?  Some don't have much fine detail, some have lots. I'd call the California relatively low on small scale detail at shorter focal lengths.

 

Olly

Good to have your input Olly, I've been comparing my California image to Kinch's superb 'clouds of California', he has so much clean detail, which is missing from mine. Interestingly, FWHM on mine is over 10"/pixel, which BXt is struggling to tame, but applying it twice seems to offer hope, the second application ruins the stars, but the nebulosity stands up well. I'm still experimenting.

Huw

[Horwig]

15 minutes ago, Horwig said:

FWHM on mine is over 10"/pixel

Sorry, brain not in gear, 'FWHM was over 10" is what I should have written 

[Horwig]

OK, so it appears you can apply BXt more than once, but with caveats. This is a 100% crop of the Ha from my California Nebula image. It was a very windy night, guiding graph was a sawtooth, hence the bad sharpness:

Worth viewing at 100%. Left, no BXt, FWHM 5.48 pixels, Center, one application, FWHM 2.32 pixels, Right, two applications FWHM 1.39 pixels

The difference between one and two applications is subtle, but definitely there, the downside is the dark halos on the double application, but maybe playing with halos might change matters, either that or remove stars and replace with the ones from the single application.

[happy-kat]

I find the second application too strong, looks over sharpened

[Stuart1971]

2 hours ago, Horwig said:

OK, so it appears you can apply BXt more than once, but with caveats. This is a 100% crop of the Ha from my California Nebula image. It was a very windy night, guiding graph was a sawtooth, hence the bad sharpness:

Worth viewing at 100%. Left, no BXt, FWHM 5.48 pixels, Center, one application, FWHM 2.32 pixels, Right, two applications FWHM 1.39 pixels

The difference between one and two applications is subtle, but definitely there, the downside is the dark halos on the double application, but maybe playing with halos might change matters, either that or remove stars and replace with the ones from the single application.

Rather then run through twice, have you tried using the settings within BlurX, the star sharpening setting now defaults to 0.5, which I find too much, but might be better to use that rather than run through twice on a lower setting, also you should run once on correct only to sort the stars shapes out, then run through again with all settings at default after other processing such as SPPC and background removal, but before yo go non linear..

[Horwig]

5 hours ago, Stuart1971 said:

Rather then run through twice, have you tried using the settings within BlurX, 

This was a very bad data set, my usual seeing is of the order of 3.5", this was about 11", so remarkable what BXt achieved really.

It was run with default star reduction, and non-stellar on max for both runs, but yes, I should run the second pass with star reduction reduced to minimum so as not to ruin the stars

Huw

[Horwig]

7 hours ago, happy-kat said:

I find the second application too strong, looks over sharpened

Yes, might be , I applied LHE to all three, probably a bit cack handed.