1.8 closed alpha feedback

General discussion about StarTools.
User avatar
admin
Site Admin
Posts: 3367
Joined: Thu Dec 02, 2010 10:51 pm
Location: Melbourne
Contact:

1.8 closed alpha feedback

Post by admin »

Hi all,

For those wishing to share feedback on the closed alpha - in particular the NBAccent and SVDecon modules, your comments, logs, results (before/after) and/or datasets are much appreciated!

Please feel free to share them here, via PM or via email.

I will endeavor to address any and all feedback here publicly, or privately (if contacted via PM or email).

Thank you!
Ivo Jager
StarTools creator and astronomy enthusiast
Wolfsrudel
Posts: 5
Joined: Thu Jul 16, 2020 5:39 am

Re: 1.8 closed alpha feedback

Post by Wolfsrudel »

Hi Ivo,

I am happy that the testing can begin.

First "bug" I encountered was that during sampling of the PSF Stars Startools crashed several times.

I have attached the logfile.

Greetings

Mario
Attachments
StarTools (2).zip
(515.1 KiB) Downloaded 152 times
Carles
Posts: 191
Joined: Thu Jul 30, 2020 5:06 pm

Re: 1.8 closed alpha feedback

Post by Carles »

Wolfsrudel wrote: Sat Jun 19, 2021 12:45 pm Hi Ivo,

I am happy that the testing can begin.

First "bug" I encountered was that during sampling of the PSF Stars Startools crashed several times.

I have attached the logfile.

Greetings

Mario
Hi Mario,

I'm also testing 1,8. I had a crash in Decon only once, and it was when I was trying to select too fast
the stars (although I thought that the reason was my laptop, which is pretty old ) Since everytime you select
one PSF Star it starts "working" with that, so selecting too many too fast might saturate it ? Just saying...because
after that crash I just selected slower and didn't crash so far!

Still, let's see what Ivo says about it!

Regards
Carles.
Carles
Posts: 191
Joined: Thu Jul 30, 2020 5:06 pm

Re: 1.8 closed alpha feedback

Post by Carles »

Hi Ivo!

One thing I've just realised, is that when using Compose to add the NarrowBand file,
after I do AutoDev to see where to crop, and crop it, I can change between Color and Luminance...but not NarrowBand
And when I went to Wipe, checked all three and my NarrowBand file wasn't cropped propperly (since I took it different day and aligned it with the others ) would it be possible to toggle NarrowBand too in Crop to be able to crop all 3 layers properly ?

Thanks!

Regards

Carles.
deanST
Posts: 17
Joined: Wed Mar 10, 2021 6:35 pm

Re: 1.8 closed alpha feedback

Post by deanST »

Hi Ivo,

I am having some issues with SVDecon module on M5 data, in particular the variable star 5 Serpentis.
Admittedly, the core is blown but the decon processing adds some unpleasant effects. It would be ideal
if not masking it would disable / exclude it from deconvolution. I tried no mask, auto star mask, grow / shrink blob
and experimented with the PSF parameters, but nothing alleviated the issue. I ran through the sampling process
and selected stars across the field having at least 1 green pixel within an identified outline. I liked what the decon
was doing for M5 itself.
Here is a screen cap after applying color post-SVDecon:
ST8_5_Serp.jpg
ST8_5_Serp.jpg (197.19 KiB) Viewed 4250 times
I was able to kluge a fix with the repair module:
ST8_5_Serp_repair.jpg
ST8_5_Serp_repair.jpg (51.38 KiB) Viewed 4248 times

link to calibrated stack: https://tinyurl.com/74cy2rcw
225 x 60" light dithered every frame, 40 dark, 40 flat, 40 bias, calibrated and stacked in Siril (no normalization)
ASI533MC Pro, WO GT71, 0.8x FlatA6iii, IDAS LPS-D3 filter

Code: Select all

-----------------------------------------------------------
StarTools 1.8.503alpha
Sat Jun 19 12:38:51 2021
-----------------------------------------------------------
File loaded [/media/dean/Samsung_T5/Astro/data/20210617/m_5_combine/r_pp_m_5_stacked_GESDT.fit].
Image size is 3008 x 3008
--- 
Type of Data: Linear and was Bayered, but not whitebalanced
--- Auto Develop
Parameter [Ignore Fine Detail <] set to [Off]
Parameter [Outside RoI Influence] set to [15 %]
Parameter [RoI X1] set to [0 pixels]
Parameter [RoI Y1] set to [0 pixels]
Parameter [RoI X2] set to [3008 pixels (-0)]
Parameter [RoI Y2] set to [3008 pixels (-0)]
Parameter [Detector Gamma] set to [1.00]
Parameter [Shadow Linearity] set to [50 %]
--- Bin
Parameter [Scale] set to [scale 70.71% / +1.00 bits / +0.41x SNR improvement]
Image size is 2126 x 2126
--- Crop
Parameter [X1] set to [307 pixels]
Parameter [Y1] set to [407 pixels]
Parameter [X2] set to [1839 pixels (-287)]
Parameter [Y2] set to [1784 pixels (-342)]
Image size is 1532 x 1377
Mask used (BASE64 PNG encoded)
--- Wipe
Parameter [Synthetic Dark/Bias] set to [Off]
Parameter [Gradient Edge Behavior] set to [Absorb 50%]
Parameter [Synthetic Flats] set to [Off]
Parameter [Sampling Precision] set to [256 x 256 pixels]
Parameter [Dark Anomaly Filter] set to [1 pixels]
Parameter [Gradient Falloff] set to [35 %]
Parameter [Synth. Bias Edge Area] set to [100 %]
Parameter [Gradient Aggressiveness] set to [80 %]
Parameter [Correlation Filtering] set to [0.4 pixels]
Mask used (BASE64 PNG encoded)
Redoing stretch of linear data
--- Auto Develop
Parameter [Ignore Fine Detail <] set to [1.6 pixels]
Parameter [Outside RoI Influence] set to [15 %]
Parameter [RoI X1] set to [623 pixels]
Parameter [RoI Y1] set to [557 pixels]
Parameter [RoI X2] set to [917 pixels (-615)]
Parameter [RoI Y2] set to [811 pixels (-566)]
Parameter [Detector Gamma] set to [1.00]
Parameter [Shadow Linearity] set to [33 %]
--- HDR
Parameter [Small Detail Precision] set to [Max]
Parameter [Channels] set to [Brightness Only]
Parameter [Algorithm] set to [Reveal DSO Core]
Parameter [Dark/Bright Response] set to [Full]
Parameter [Detail Size Range] set to [60 pixels]
Parameter [Strength] set to [1.2]
--- Contrast
Parameter [Expose Dark Areas] set to [Yes]
Parameter [Brightness Retention] set to [Off]
Parameter [Precision] set to [256 x 256 pixels]
Parameter [Shadow Detail Size] set to [10 pixels]
Parameter [Locality] set to [50 %]
Parameter [Shadow Dyn Range Alloc] set to [50 %]
--- Wavelet Sharpen
Parameter [Structure Size] set to [Medium]
Mask used (BASE64 PNG encoded)
--- SNR-aware Wavelet Sharpening
Parameter [Scale 1] set to [100 %]
Parameter [Scale 2] set to [100 %]
Parameter [Scale 3] set to [100 %]
Parameter [Scale 4] set to [100 %]
Parameter [Scale 5] set to [100 %]
Parameter [Mask Fuzz] set to [4 pixels]
Parameter [Amount] set to [300 %]
Parameter [High SNR Size Bias] set to [85 %]
Parameter [Low SNR Size Bias] set to [0 %]
Parameter [Dark/Light Enhance] set to [0% / 100%]
Mask used (BASE64 PNG encoded)
Mask used (BASE64 PNG encoded)
Mask used (BASE64 PNG encoded)
Mask used (BASE64 PNG encoded)
Mask used (BASE64 PNG encoded)
Mask used (BASE64 PNG encoded)
Mask used (BASE64 PNG encoded)
--- Spatially Variant PSF Deconvolution
Parameter [PSF Resampling] set to [Intra-Iteration]
Parameter [Synthetic PSF Model] set to [Circle of Confusion (Optics Only)]
Parameter [Sampled Area] set to [15x15]
Parameter [Synthetic PSF Radius] set to [1.5 pixels]
Parameter [Synthetic Iterations] set to [Off]
Parameter [Spatial Error] set to [1.00]
Parameter [Deringing Fuzz] set to [20.0 pixels]
Parameter [Deringing] set to [50 %]
Parameter [Dyn. Range Extension] set to [1.00]
Parameter [Linearity Cutoff] set to [60 %]
Parameter [Sampled Iterations] set to [10x]
Mask used (BASE64 PNG encoded)
Mask used (BASE64 PNG encoded)
--- Color
Parameter [Bias Slider Mode] set to [Sliders Reduce Color Bias]
Parameter [Style] set to [Scientific (Color Constancy)]
Parameter [LRGB Method Emulation] set to [RGB Ratio, CIELab Luminance Retention]
Parameter [Matrix] set to [Identity (OFF)]
Parameter [Dark Saturation] set to [2.4]
Parameter [Bright Saturation] set to [Full]
Parameter [Saturation Amount] set to [230 %]
Parameter [Blue Bias Reduce] set to [1.00]
Parameter [Green Bias Reduce] set to [1.91]
Parameter [Red Bias Reduce] set to [1.03]
Parameter [Mask Fuzz] set to [1.0 pixels]
Parameter [Cap Green] set to [0 %]
Parameter [Highlight Repair] set to [2 pixels]
--- Unified De-Noise
Parameter [Grain Size] set to [10.5 pixels]
Parameter [Walking Noise Size] set to [1.0 pixels]
Parameter [Walking Noise Angle] set to [0]
--- Super Structure
Parameter [Detail Preservation] set to [Linear Brightness Mask Darken]
Parameter [Compositing Algorithm] set to [Multiply, Gamma Correct]
Parameter [Brightness, Color] set to [Only Brightness]
Parameter [Brightness Retention] set to [Local Median]
Parameter [Mask Fuzz] set to [1.0 pixels]
Parameter [Airy Disk Radius] set to [50 %]
Parameter [Gamma] set to [1.00]
Parameter [Detail Preservation Radius] set to [20.0 pixels]
Parameter [Saturation] set to [50 %]
Parameter [Strength] set to [100 %]
Mask used (BASE64 PNG encoded)
--- Repair
Parameter [Radial Samples] set to [32]
Parameter [Sub Sampling] set to [4x]
Parameter [Algorithm] set to [Redistribute, Core Is Avg Location]
Parameter [Grow Mask] set to [0 pixels]
--- Unified De-Noise
Parameter [Grain Size] set to [9.4 pixels]
Parameter [Walking Noise Size] set to [1.0 pixels]
Parameter [Walking Noise Angle] set to [0]
--- Unified De-Noise
Parameter [Scale 1] set to [99 %]
Parameter [Scale 2] set to [99 %]
Parameter [Scale 3] set to [98 %]
Parameter [Scale 4] set to [97 %]
Parameter [Scale 5] set to [67 %]
Parameter [Equalized Grain] set to [15 %]
Parameter [Scale Correlation] set to [33 %]
Parameter [Color Detail Loss] set to [40 %]
Parameter [Brightness Detail Loss] set to [55 %]
Parameter [Grain Dispersion] set to [9.4 pixels]
Carles
Posts: 191
Joined: Thu Jul 30, 2020 5:06 pm

Re: 1.8 closed alpha feedback

Post by Carles »

Hi,
I've also found some artifact in the stars' Core using SVDecon
This one is taken with SCT. Tracking was OK but stars are elongated, not sure if
is anything to do with the focal reducer (read something somewhere) but I think
focus is OK.

Regards

Carles.
Attachments
SVdecon ARtifact. core..jpg
SVdecon ARtifact. core..jpg (314.89 KiB) Viewed 4198 times
Stefan B
Posts: 422
Joined: Sat Oct 31, 2020 8:59 pm

Re: 1.8 closed alpha feedback

Post by Stefan B »

Hi Ivo et al.,

tried 1.8 on two data sets yesterday. One of them was M33 with HII regions:
M33.jpg
M33.jpg (97.01 KiB) Viewed 4187 times
NBaccent made a huge difference to the image (has been late yesterday so I forgot to do a before/after screenshot). I think that this greatly simplifies adding NB to RGB data :thumbsup:

Second attempt was on an ongoing project so far only containing one night of l-eNhance data for NGC7000:
NGC7000, 2021-06-18, 46x5min.jpg
NGC7000, 2021-06-18, 46x5min.jpg (256.32 KiB) Viewed 4187 times
Has been loaded in Compose using the bi-color from OSC/DSLR option. Went for H(HO)O in color. SVDecon has been used here. Cannot say if it made a difference across the image, but it worked. I had one crash during selection of sample stars. StarTools just closed without further notice. PC did not crash. Just had to restart StarTools and do the workflow again. Sounds a bit like the issue already encountered by Carles.

Some remarks:
- WIPE: Before/After doesn't work for NBaccent
- WIPE: what is correlation filtering?
- WIPE: when entering duo NB data in compose and using the bicolor from OSC/DSLR option, one can only swith between NB and color instead of luminance and color. On purpose?
- SVDecon: In the description of the last image of the SVDecon documentation, left and right column are confused
- SVDecon: after decon the sampled stars are hard to inspect due to their marking
- SVDecon: Would it be possible for StarTools to do the sampling automatically since it already generates the apodization mask automatically?

Generally: Is it allowed to already share images created with 1.8? Not that I have one to share at the moment. But would be nice to know.

Best regards
Stefan
Carles
Posts: 191
Joined: Thu Jul 30, 2020 5:06 pm

Re: 1.8 closed alpha feedback

Post by Carles »

Stefan B wrote: Mon Jun 21, 2021 6:33 am Hi Ivo et al.,

tried 1.8 on two data sets yesterday. One of them was M33 with HII regions...

Hi Stefan!

Yes I've had that issue too..and once more. the second time is true that I was (un purpose) to select "faster" stars across the field.
When I selected many but slower, didn't crash.

Also, I found that sometimes after wipe, and AutoDev again, my image was BW ...even I loaded from "open" not Compose. (single image, not NB here )

Stefan, since it is to accentuate NarrowBand, you should load an RGB file in either L, RGB and the narrowband file in NBAccent slot only. And I am not sure you should select Bicolor OSC/DSLR, since the Narrowband/L-enhance in your case will be "layered" somehow (or that's the impression I have using it ) after in the NBAccent module. SO I am assuming it should be L-RGB or Synthtic L+L from RGB or something. Hope Ivo can clarify this.

I've also realized the confusion on the columins in SVDEcon last image in documentation while translating into Spanish (for future addition to manual, whenever it needs to) have notified Ivo on email yesterday.

About the "Correlated Fitering " is already in Wipe documentation of 1.8
https://aife.me/startools18/modules/wipe

"Fixed pattern noise and correlated artifact filtering
The 'Correlation Filtering' parameter specifies the size of correlation artifacts that should be removed. This feature can ameliorate correlation artifacts that are the result of dithering, debayering or fixed pattern sensor cross-talk issues. Correlated noise and artifacts can look like detail to both humans and algorithms. By pre-emptively filtering out these artifacts, modules will be able to better concentrate on the real detail in your dataset and image, rather than attempting to preserve these artifacts.

The usage of this filter is most effective on oversampled data where the artifacts are clearly smaller than the actual resolved detail."

That being said, awesome images, Stefan!! :D

Clear skies
Carles
Stefan B
Posts: 422
Joined: Sat Oct 31, 2020 8:59 pm

Re: 1.8 closed alpha feedback

Post by Stefan B »

Hi Carles,

thanks for pointing me to the explanation of the correlation filtering! Must have overlooked it as I have scrolled through the documentation yesterday (as already noted ... it got late ... my fault :oops:).

SVDecon crashed once more here, too, during sampling. But I didn't select the stars very fast IMHO. Maybe my notebook is too slow...
Stefan, since it is to accentuate NarrowBand, you should load an RGB file in either L, RGB and the narrowband file in NBAccent slot only.
Sorry, should have made that clearer in my first post. I did exactly this for the M33 image. Only for NGC7000 I used the bi-color option.
That being said, awesome images, Stefan!
Thanks, much appreciated :)

Regards
Stefan
User avatar
admin
Site Admin
Posts: 3367
Joined: Thu Dec 02, 2010 10:51 pm
Location: Melbourne
Contact:

Re: 1.8 closed alpha feedback

Post by admin »

Wolfsrudel wrote: Sat Jun 19, 2021 12:45 pm First "bug" I encountered was that during sampling of the PSF Stars Startools crashed several times.
Many thanks Mario, I think I have a fix for this now. A new alpha is coming later this week.
deanST wrote: Sat Jun 19, 2021 5:47 pm I am having some issues with SVDecon module on M5 data, in particular the variable star 5 Serpentis.
Many thanks for the report, log and dataset. The Dynamic Range Extension feature is not working properly (yet) and may be the cause of these sorts of artifacts. I will have a look at that dataset, thank you. The one star appears very strange though, almost as if it is the result of some sort of HDR feature?

Thanks again!
Carles wrote: Sat Jun 19, 2021 5:03 pm One thing I've just realised, is that when using Compose to add the NarrowBand file,
after I do AutoDev to see where to crop, and crop it, I can change between Color and Luminance...but not NarrowBand
And when I went to Wipe, checked all three and my NarrowBand file wasn't cropped propperly (since I took it different day and aligned it with the others ) would it be possible to toggle NarrowBand too in Crop to be able to crop all 3 layers properly ?
Indeed - I still need to implement this. Thank you!

I think I got the SVDecon crashing under control now. Expect an alpha later this week.
Stefan B wrote: Mon Jun 21, 2021 6:33 am NBaccent made a huge difference to the image (has been late yesterday so I forgot to do a before/after screenshot). I think that this greatly simplifies adding NB to RGB data :thumbsup:
Fantastic! :bow-yellow:
Stefan B wrote: I had one crash during selection of sample stars. StarTools just closed without further notice. PC did not crash. Just had to restart StarTools and do the workflow again. Sounds a bit like the issue already encountered by Carles.
Indeed, fix incoming.
Some remarks:
- WIPE
Indeed lots of little bugs to iron out :)
Thank you!
- SVDecon: In the description of the last image of the SVDecon documentation, left and right column are confused
I'm 42 years old grown man and still don't know my rights from my lefts. :oops:
- SVDecon: after decon the sampled stars are hard to inspect due to their marking
- SVDecon: Would it be possible for StarTools to do the sampling automatically since it already generates the apodization mask automatically?
Thank you. I will look at this!
Generally: Is it allowed to already share images created with 1.8? Not that I have one to share at the moment. But would be nice to know.
Absolutely!
Ivo Jager
StarTools creator and astronomy enthusiast
Post Reply