StarTools 1.9 Beta

[admin]

Hi all,

The first 1.9 beta is now available.
If you've been following along with the 1.9 preview thread, there should not be too many new surprises.
But if you haven't, then please have a look at the CHANGELOG file, which should list the many improvements.
Again, many thanks to @hixx @decay and @Carlesa25 for their massive contributions to the German and Spanish translations.

Clear skies!

[Carles]

Hi all,

The first 1.9 beta is now available.
If you've been following along with the 1.9 preview thread, there should not be too many new surprises.
But if you haven't, then please have a look at the CHANGELOG file, which should list the many improvements.
Again, many thanks to @hixx @decay and @Carlesa25 for their massive contributions to the German and Spanish translations.

Clear skies!

Hi Ivo
haha you always mention Carlesa25 <-- and that's why I never get notifications xD

[Mike in Rancho]

Great!

Too soon for a 1.10 wish-list?

I'm still not quite there on the 1.9 SVD deringing. The three controls allow correction of most induced star ringing and shadowing, but not 100% yet. I've tried it against my Newt, of course, along with a couple different refractors (one being Ron's) and RASA11 data. You really do have to customize the deringing sliders for the different optics.

On some larger and medium stars, the deringing effect field seems a little too small to fully cover the ringing, regardless of settings. Also and perhaps a little more troublesome as far as the elbow grease required for repair, some very small stars, if against nebulosity or even the diffraction ball of a very bright star, don't seem to be included in the deringing effect, again regardless of controls. Making me wonder if some sort of star sensing is used in the background as a deringing aid, and maybe can't dig everything out?

[dx_ron]

In case you are interested, here is a link to the refractor (AT130EDT) data that Mike referred to: https://www.dropbox.com/s/vqs6d46ltd1i1 ... s.fit?dl=0, where the stars that have saturated cores in the linear state get a distinct dark ring when stretched (this also occurs in 1.8, by the way).

If it is actually a problem with the optics or acquisition, I'd be interested in what that might be. Not happy, but interested...

[Stefan B]

Thanks for providing the new beta, Ivo!

As already mentioned in another thread I am struggling with the new deringing options. I just can't get the big star to look decent:

Deringing M8.jpg (142.94 KiB) Viewed 117622 times

The linear data can be found here: https://www.dropbox.com/s/grvo92airo17t ... .fits?dl=0

Processing for the above image was very basic. 50%bin, crop, wipe, Optidev (no ROI), contrast, HDR (tuned down), sharp (tuned down), SVD. I can't get a combination of deringing focus and fuzz that makes the artefact disappear. No problem in 1.9.542 - I just bumped up deringing detect until all rings were gone.

Can anyone help?

Regards
Stefan

[almcl]

Hi Stefan

Thanks for sharing your (very nice!) data.

I had a quick go at processing it in 1.9.557 beta and got the result below. Not sure if its the sort result you're after, but if it is, the SVD settings that seemed to help were Deringing fuzz 13% and Deringing focus 84%. I found lowering the fuzz setting a bit counter intuitive at first but it does seem to help.

M 8, deringing.jpg (184.72 KiB) Viewed 117617 times

[Stefan B]

Damn, that looks pretty good! Will try that...thanks!

[Mike in Rancho]

Processing for the above image was very basic. 50%bin, crop, wipe, Optidev (no ROI), contrast, HDR (tuned down), sharp (tuned down), SVD. I can't get a combination of deringing focus and fuzz that makes the artefact disappear. No problem in 1.9.542 - I just bumped up deringing detect until all rings were gone.

Can anyone help?

Regards
Stefan

Hey Stefan, that's a nice nebular field you have there. Corner stars are a little off, but it's pretty cool when you hit the first OptiDev or Wipe screen and so much emission structure pops up all over the screen.

At the moment I'm not much help though! This does seem to be a strongly ringing dataset, sadly. I thought I was getting pretty good at juggling the three controls too, but just haven't had much luck yet. At 50% bin, I used very low linearity (single digits) and very low low fuzz to get a close to acceptable result. I also added a bit of dynamic extension, though I'm not sure it helped. There was still some ringing and shadowing left over, but taking the image into Color all that mostly got lost, perceptually, absent zooming in close. I think at that stage Shrink deringing only would have cleaned things up.

I then tried again at 71% bin, figuring that would be better, but instead it was worse.

I wonder if it's something I'm doing early on with the Wipe, stretch, or other modules that is leaving my processing more prone to worse star ringing and shadowing? I will have to experiment more. A few versions ago DAF and IFD usage was having an effect, but I don't know if that whole scheme was jettisoned when the three sliders came in.

Anyway I otherwise did basically what you did. Bicolor composite, cropped the overlaps, 50% bin, Wipe, no ROI OptiDev but with IFD, contrast defaults, HDR 50 pixels 1.01 tame and highlights down to 15%, then SVD as stated. I think if I took it through Color and Shrink and then finished it off, the result would have been good with no readily visible artifacts at normal viewing scale.

[Stefan B]

Thanks for having a look, Mike. You're right, stars in the corners are ... ughhh ... this seems to be a hit and miss for me. My NGC 6914 data set was much better in that regard for no apparent reason.

I didn't have the opportunity to try the parameters suggested by @almcl on this data set but it worked fine on NGC 6914. A bit strange...

Regards
Stefan

[Mike in Rancho]

Thanks for having a look, Mike. You're right, stars in the corners are ... ughhh ... this seems to be a hit and miss for me. My NGC 6914 data set was much better in that regard for no apparent reason.

I didn't have the opportunity to try the parameters suggested by @almcl on this data set but it worked fine on NGC 6914. A bit strange...

Regards
Stefan

Let us know how you do on a re-try. I put the two images above into layers in Gimp, scaled up almcl's version to match, in order to blink the differences. Granted the scaling and then saving as an animation at a reasonable file size may have degraded some detail, but this is what I see..

stefan Deringing comparo M8 4.gif (679.27 KiB) Viewed 117515 times

almcl's process does turn out much cleaner through SVD, though the question becomes what was done in order to achieve that. Just looking at the field here, there seem to be potential variances in the OptiDev (maybe even Wipe), Contrast, and HDR, and of course maybe then the selection of star samples in SVD. But even his version does have some remnant ringing here and there, including a seeming size mismatch between the deringed area and the ring or shadow from deconvolution (just a wee bit too small, as I mentioned earlier), and also at least one of those bright stars has ended up with a dimmed center core, though perhaps it might need to wait for Color to see how that pans out.

Ponder ponder....