StarTools 1.9 preview

General discussion about StarTools.
User avatar
admin
Site Admin
Posts: 3338
Joined: Thu Dec 02, 2010 10:51 pm
Location: Melbourne
Contact:

Re: StarTools 1.9 preview

Post by admin »

happy-kat wrote: Tue Mar 28, 2023 9:41 am Thank you for the updated version Ivo, though the download page is pulling 1.9.546 still
Thanks @happy-kat! :)
It's probably your browser hanging on to old content. Try SHIFT + CTRL + R in your browser to force a refresh.
Ivo Jager
StarTools creator and astronomy enthusiast
User avatar
admin
Site Admin
Posts: 3338
Joined: Thu Dec 02, 2010 10:51 pm
Location: Melbourne
Contact:

Re: StarTools 1.9 preview

Post by admin »

1.9.549 alpha is now available

Biggest change (apart from tons of multi-language work) is that the auto mask generator is now real-time.
Ivo Jager
StarTools creator and astronomy enthusiast
Mike in Rancho
Posts: 873
Joined: Sun Jun 20, 2021 10:05 pm
Location: Alta Loma, CA

Re: StarTools 1.9 preview

Post by Mike in Rancho »

admin wrote: Sat May 06, 2023 3:38 am 1.9.549 alpha is now available

Biggest change (apart from tons of multi-language work) is that the auto mask generator is now real-time.
:thumbsup:

To mangle some Yoda: Keep, or keep not. There is no do.
User avatar
admin
Site Admin
Posts: 3338
Joined: Thu Dec 02, 2010 10:51 pm
Location: Melbourne
Contact:

Re: StarTools 1.9 preview

Post by admin »

Mike in Rancho wrote: Sun May 07, 2023 8:30 pm
admin wrote: Sat May 06, 2023 3:38 am 1.9.549 alpha is now available

Biggest change (apart from tons of multi-language work) is that the auto mask generator is now real-time.
:thumbsup:

To mangle some Yoda: Keep, or keep not. There is no do.
:lol:
I made the button green in His honor.
Ivo Jager
StarTools creator and astronomy enthusiast
Mike in Rancho
Posts: 873
Joined: Sun Jun 20, 2021 10:05 pm
Location: Alta Loma, CA

Re: StarTools 1.9 preview

Post by Mike in Rancho »

Ok here's a weird one. If you enter Wipe with a custom mask already set, everything goes kaboom. :shock: Seems fine if you create the mask once in Wipe, though.
Carles
Posts: 155
Joined: Thu Jul 30, 2020 5:06 pm

Re: StarTools 1.9 preview

Post by Carles »

Many processes seem faster, but still, I'd like some more flexibility on Decon module. For my datasets, 1.9 decon grabs only noise, no stars.
So for now I'll stick to 1,8.
User avatar
admin
Site Admin
Posts: 3338
Joined: Thu Dec 02, 2010 10:51 pm
Location: Melbourne
Contact:

Re: StarTools 1.9 preview

Post by admin »

Mike in Rancho wrote: Tue May 09, 2023 5:13 am Ok here's a weird one. If you enter Wipe with a custom mask already set, everything goes kaboom. :shock: Seems fine if you create the mask once in Wipe, though.
Does that still happen in this version?
https://download.startools.org/StarTool ... _alpha.zip
Carles wrote: Tue May 09, 2023 8:58 pm Many processes seem faster, but still, I'd like some more flexibility on Decon module. For my datasets, 1.9 decon grabs only noise, no stars.
So for now I'll stick to 1,8.
Fix coming soon. Are you still using the QHY8? I have some test datasets from you that exhibit that specific correlated noise pattern that indeed looks like little stars in places. I'm now leveraging the Dark Anomaly filter settings from Wipe and AutoDev (given they deal with the same sort of issue) to pre-filter the input to the star detector.

I've also been rewriting the de-ringing code.
Ivo Jager
StarTools creator and astronomy enthusiast
Mike in Rancho
Posts: 873
Joined: Sun Jun 20, 2021 10:05 pm
Location: Alta Loma, CA

Re: StarTools 1.9 preview

Post by Mike in Rancho »

admin wrote: Thu May 11, 2023 6:01 am Does that still happen in this version?
https://download.startools.org/StarTool ... _alpha.zip
Nope! No crash going into Wipe with the mask already made. :thumbsup:

admin wrote: Thu May 11, 2023 6:01 am Fix coming soon. Are you still using the QHY8? I have some test datasets from you that exhibit that specific correlated noise pattern that indeed looks like little stars in places. I'm now leveraging the Dark Anomaly filter settings from Wipe and AutoDev (given they deal with the same sort of issue) to pre-filter the input to the star detector.

I've also been rewriting the de-ringing code.
Looking forward to it. :D Nearly full-auto SVD can't be easy, what with the vast spread of possible datasets, not to mention differing user preferences for processing including desired resolution.

:bow-yellow:
Carles
Posts: 155
Joined: Thu Jul 30, 2020 5:06 pm

Re: StarTools 1.9 preview

Post by Carles »

admin wrote: Thu May 11, 2023 6:01 am
Mike in Rancho wrote: Tue May 09, 2023 5:13 am Ok here's a weird one. If you enter Wipe with a custom mask already set, everything goes kaboom. :shock: Seems fine if you create the mask once in Wipe, though.
Does that still happen in this version?
https://download.startools.org/StarTool ... _alpha.zip
Carles wrote: Tue May 09, 2023 8:58 pm Many processes seem faster, but still, I'd like some more flexibility on Decon module. For my datasets, 1.9 decon grabs only noise, no stars.
So for now I'll stick to 1,8.
Fix coming soon. Are you still using the QHY8? I have some test datasets from you that exhibit that specific correlated noise pattern that indeed looks like little stars in places. I'm now leveraging the Dark Anomaly filter settings from Wipe and AutoDev (given they deal with the same sort of issue) to pre-filter the input to the star detector.

I've also been rewriting the de-ringing code.
Great to hear, Ivo, thanks :)

I do use still qhy8L and recently did some tries with a planetary Player One Uranus-C for DSO, if you want I can send you recent qhy8L datasets and some for the PO too (no cooling, but surpisingly good..)
User avatar
admin
Site Admin
Posts: 3338
Joined: Thu Dec 02, 2010 10:51 pm
Location: Melbourne
Contact:

Re: StarTools 1.9 preview

Post by admin »

Ok, 1.9.550 is up.

@Carlesa25 see if this works better for your datasets.

A fair bit of work has gone into SV decon in general.

De-ringing should be much improved. It's now not so much an "amount "any more, but rather a continuum of different "looks".
I am trying to get away with just one parameter, rather than having multiple sliders allocated to de-ringing in the UI.

Internally, the slider blends ringing-affected areas with pixels from the original (pre-deconvolved image), a Gaussian blurred version of the new image and a transition from new image into the two. The goal is to - in no circumstance - introduce new high frequency "detail". Detail should either omitted or recovered form the original image.
Ivo Jager
StarTools creator and astronomy enthusiast
Post Reply