CaLIGHTs v3.1.0.90 Released

[wols...@gmail.com]

I tried to take my time testing this version and I hope it works well.  Please let me know if you find issues.

Peter

Latest Revision 3.1.0.90

-Flat compensation logic has been fixed. This logic was applying too much compensation because of a logic error I had made when preparing V3.1.0.80.

-CaLIGHTs now remembers the LIGHTs Directory and the directory where your masterBIAS for LIGHT frames are.  When you start CaLIGHTs and select your LIGHTs you will find that CaLIGHTs remembers where you were last time.  When you specify your masterFLAT or masterDARKFLAT, CaLIGHTs will take you to where your LIGHT frames are kept.  Typically these masters are kept in a subdirectory of this folder.

-I have revived the usage of Black ADU.  This is primarily a DSLR feature that is not widely supported in the dedicated astrocam market. Black ADU is defined as the pixel value that corresponds to complete darkness.  It is the equivalent of the mean pixel value in your masterBIAS for LIGHT frames.  Enabling it’s usage is done via the Preferences button.  There is a checkbox for enabling Black ADU usage.  I had found there were a few DSLRs that either did not report a value for Black ADU or reported the wrong value for Black ADU.  Black ADU is only used if you have not supplied a masterBIAS for your LIGHT frames and you have enabled Black ADU’s usage via the checkbox.  If you do not enable Black ADU’s usage, you will find that CaLIGHTs will not allow you to calibrate until you specify a masterBIAS for LIGHT frames.  If you enable Black ADU’s usage and CaLIGHTs detects a non-zero value for Black ADU you will find that CaLIGHTs will allow you to calibrate without specifying a master BIAS frame.

One very important outcome of this work with Black ADU is the creation of a value called “Pedestal”. The Pedestal value will either be equal to Black ADU or the mean value of your masterBIAS for LIGHT frames. The value for Pedestal is displayed in the Calibration Progress Report.

This change has resulted in many benefits. It has allowed CaLIGHTs to continue to be able to view your LIGHT frames, using the Calibrate Preview feature, without specifying any masters.  I use this feature a lot as it makes it quicker for me to perform software testing. It also allows me to better help fellow DSLR astrophotographers who only supply a LIGHT frame. A big benefit is that it gives CaLIGHTs more “headroom” when creating calibrated LIGHT frames…fitting all the data into a 0 to 65,535 range is better optimized.

-CaLIGHTs now allows more values to choose from for the Multiplier.  With a 14b DSLR the choices for Multiplier were 1,2 and 4.  Now a choice of 3 is also available.  For a 12b DSLR the choices were 1, 2, 4 and 8.  Now the choices are 1, 2, 3, 4, 5, 6, 7 and 8.  This is useful when you discover that there are several Clamped Pixels @65535. If you reduce the Multiplier by one this may eliminate the Clamped Pixels @65535.  Typically these clamped pixels @65535 occur at the edges of your LIGHT frame where the FLAT compensation has brighten the image.  Reducing the Multiplier is a trade-off.  I believe using the highest available Multiplier minimizes quantization errors.  Most DSO details are extremely faint so having as many brightness levels as possible available is a significant concern.  When you examine your LIGHT frame you will typically discover that the clamped pixels @65535 are all stars. With my post processing I am always applying star shrinking or dimming because the stars are way too bright. Maybe…in a future version I will figure out how to use the FLAT Compensation logic to pro-actively dim only very bright stars that are affected by vignetting…so that they cannot exceed 65535.  Watch this space…

[Scott Kuchma]

Hello Peter ,

Well , I've finished my run through of NGC7023 , the Iris Nebula , and I'm really pleased with the final Image . What I noticed in the final Image were a *lot* more Stars . Then I realized that was because my original attempt was severely cropped .
This was done , as it was in several of my Processing attempts this year , to try and remove the vertical colour banding I'd see on the left and right side of the Images . I didn't get that today as you saw in the previous Email .
So , just for completeness , I'm using CaLIGHTs v3.10.90 , DSS v4.2.3 (still) , SiriL v09.12 (waiting on RC v1.00) , DeNoise Ai v2.3.5 (v2.3.6 came out an hour ago) and then FastStone v7.5 .
A couple of Screen captures of the Process will be below .

I can only imagine what I'll be able to Image if/when I can upgrade my Mount (think Wedge). I think I'm a little further up the Learning Curve and a big chunk of that is thanks to you and CaLIGHTs .

I'll be going back and retrying a few other Imaging and Processing attempts with this release of CaLIGHTs . It's working very well !

First up is the Calibration Preview in CaLIGHTs...lots of noise but smoothly and evenly spread across the Image .

Next is the DSS Linear Stacked Image . Much more Stars this time...

Then the final TIF from SiriL...

Then a quick run through DeNoise AI....

...and finally final colour tweaks and repairs using FastStone v 7.5...

I never noticed in my first attempt of this Nebula but can you see the Crescent Wrench to the left of center in the Nebula ? Very cool !

Not too shabby for an unmodded DSLR , Alt/Az Mount , no filters , ISO6400 and 40s Exp . I am "satisfied" with this effort . It can only get better .

Looking forward to great things from CaLIGHTs in the future .

Cheers ,

Scott................

[Scott Kuchma]

Hello Peter ,

Interesting , and I think good , result from another ReProcessed Image from this last Summer . M20 , the Trifid Nebula .

The first Image is from this Summer and probably used CaLIGHTs v3.10.70 along with all the Software I currently use .

This second Image is from the original Light Frames , Calibrated with CaLIGHTs v3.10.90 , with the same Settings and Masters as the ones I used for the Iris Nebula above .

There are some very interesting differences in detail in a lot of places .

This is the 24July2020 completed Image .

Below is today's completed Image . Can you spot the differences ?

There's a Dark Cloud detail completely missing in the earlier Image as well as the two very bright Stars near the center of the Nebula .

There is of course an obvious colour difference and I'm not sure if that is just down to my current Processing techniques . I tend now to not aggressively approach this .

I am much happier with the current Image . More subtle , more information , more accuracy to the actual Nebula (according to Web Photos) .

Interesting eh ? Can the recent changes to CaLIGHTs account for a change like this ? I'm sure it has nothing to do with my current prowess at the keyboard . LOL .

Let me know what you think . I'll probably start looking at a lot more of my previous Images and run them through this new CaLIGHTs version over the next week .

There's some kind of magic in there .

Scott............

[Scott Kuchma]

Just a Post to correct an error in my Post above as there are no Functions in Google Groups to allow me to simply Edit my own Post . Strange .

Above I wrote..." with the same Settings and Masters as the ones I used for the Iris Nebula above ."

I should have typed..." with the same Settings and Masters as the ones I used for the Trifid Nebula ."

Scott......................

[wols...@gmail.com]

Scott,

Great job re-processing your Trifid image.  The colors are warmer and more saturated.  There is also less noise causing graininess.  There is much more detail in the dark dust lanes.

The latest version of CaLIGHTs does a much better job identifying and fixing bad pixels which will help with post processing.  I also hope that the calibrate preview feature makes it easier, and more fun, for users to fine tune their calibration options.  I like to think that a small improvement during calibration of each LIGHT frame results in a cumulative improvement through-out stacking and post processing.

Peter

[Scott Kuchma]

Hi Peter ,

Is there anything in the new CaLIGHTs that explains the incredible inclusion of an entire Dust Lane and new Stars into the new final Image ? It's as if I had someone else's LIGHTs .

I'm still going through some of my old Captures to see what the new differences are . Here , below , is my first attempt at NGC1333 , the Embryo Nebula . Not a lot of Light Frames for this to shine or pop .

Then a complete run through with CaLIGHTs and all the previous old Software...

Really can't make a silk purse out of a sow's ear but there is now more colour , in the correct shades , and a smoother finish overall . 

Very , very interesting . This was only 22 Light Frames Stacked at ISO6400 and 40s Exp .

I will keep trying more of my old Images as I am intrigued by what this new version of CaLIGHTs is doing .

Cheers ,

Scott..................

[Scott Kuchma]

OK , I think this will be my last Post on this Topic as I don't have anymore finished Images which show a huge difference in the final Image . This is still done through v3.10.90 .

This was the last best Image I had done of M51 and was about the third attempt at improving on the original...

Then a complete rerun through from the beginning including some surprise success with Sharpen Ai which hasn't been able to do much with any of my Astro Photos before....

.

A few issues with a  number of Stars looking a little bloated and odd but overall I'm pleased with this latest attempt .

Cheers ,

Scott.................

[wols...@gmail.com]

Scott,

That is a stunning improvement.  Are you sure your not just posting images taken by Hubble??? :)

Peter

[Scott Kuchma]

HaHa , thanks Peter . Actually the NASA Hubble Image looks like someone put blotches of red paint here and there around the Galaxy . It doesn't look "real" somehow . LOL .

Also , I think I may have found what's been causing some of my Stars to have a black hole in the centre on some of my Images . It's DSS ! More specifically it's the Cosmetic Settings Tab .

According to a Site today..."having "Cosmetic Corrections" checked in DSS can cause the same issue when it thinks the center of a star is a hot pixel and removes it."

I've now turned that selection OFF in DSS so we'll see if I get any more holes in my Stars . Can't remember why I turned it on in the first place ? It wasn't every Star and not every Image I processed .

Is there anything about the latest DSS , with the new LibRaw , that we should know about or consider when working with CaLIGHTs ? I'm still using v4.2.3 and am reluctant to Upgrade .

Thanks again for all your hard work creating and maintaining and upgrading CaLIGHTs .

Scott...................

[Scott Kuchma]

Just an interesting update to my ReProcessing sequence . I was sure I could get a better Image of the Trifid Nebula using Sharpen Ai but it always came out worse .

So I went back to my finished SiriL Image and ran it through Sharpen Ai first *then* DeNoise Ai and eventually through FastStone again .

Interesting result . Here is the Image I posted above which was , at the time , a huge improvement over the original...

Then this is the result of simply switch the Processing sequence...

I think I'll update my Sequence Workflow and keep this last Image .

Scott...............

[wols...@gmail.com]

SCott,

Those dark tendrils look crisper.  I don't use any of the tools you use so I am excited by what you are doing and achieving. I know that my post processing is where I fall down.  I always try to recover star colour.  My stars easily blow-out.  I would like to see you strive to push back the brightness of your stars and see if you can pull colour back into them.  I always find that diming and shrinking the stars always compliments the nebula and makes an astrophoto more fascinating.  Out of curiosity...how many LIGHT frames did you use for your Trifid? What exposure/ISO did you use?  I suspect you used your 8" EdgeHD scope at F10 which is very high magnification so I also suspect that there were very few photons collect in each photo. I'm very impressed with how smooth the sky blue central region is.

Peter

[Scott Kuchma]

Hi Peter ,

This may be hard to believe but I was using 35s Exposures at ISO6400 . Out of 66 Lights I only could keep 38 after screening them in DSS .

DSS did find 85 Stars to align with though which is better than most Images .

With CaLIGHTs I always use SkyGlow Comp , 2x2 Binning and the Row Noise Filter . That's all . Plus FLATS , DARKS , BIAS and DARKFLATS .

Then into SiriL for about ten minutes , now Sharpen Ai , then DeNoise Ai , and finally into FastStone to tweak the colours , heal any garbage left over and maybe one last crop .

I wish I could do something with the Stars but I don't know how . I know GIMP and StarTools have ways of doing that but I just don't understand it . They are way above my skill grade .

Fingers crossed that StarFixer comes out of Alpha Testing (https://www.starfixer.org/) and is made available . I see they're on v2 now so he may be close .

I've tried it , it works , but you send in the Image and it gets returned with their Beta Watermark .

This is why my ears perked up when you mentioned something like this might be in a future design/release of CaLIGHTs . Very interesting .

Trying to redo my Comet Neowise Images but DSS is being cranky tonight so I'll leave and try again tomorrow .

Cheers ,

Scott.............

[wols...@gmail.com]

Oh Man!! so much detail with such a short exposure and high ISO.  The sharpness of your image is largely due to the short exposures.  At that high a magnification image rotation and polar alignment errors will turn your stars into ovals if you tried a longer exposure.  And only 38 images...that's just slightly longer than 22 minutes of total exposure.  I typically take hours of exposures.   I'm sure you have seen awesome astrophotos that claim total exposure times of 10 hours or more.  There's a lot to learn here.

This year I was taking 5 to 10 minute exposures at unity gain with my QHY294C.  My photos have high dynamic range but I also suspect that the long exposures are resulting in fuzzy images because of atmospheric seeing.  I believe that shorter exposures can have much tighter stars simply due to being subject to less seeing fluctuations.  Your focus is also very good.  I use an automated routine to focus my scope and I have confirmed it is accurate.  Somewhere in all of this is the secret to taking awesome astrophotos. Maybe this winter I should invest in an off-axis guider.

Scott...are you using the latest version of CaLIGHTs?  I have tested it with your LIGHT frames but I need feedback.

Peter

[Scott Kuchma]

Hi Peter ,

The "sharpness" may be down to some influence from the shorter exposures but I think a lot more happened using the new Sharpen Ai . It's Preview selection had suggested Focus as the route to take during its Processing .

I don't think my manual knob turning is giving me the best Focus at all . I did 3D print a large knob to go over the main small rubber one on the OTA . This gives me a little more small adjustment control .

Next year I will have the Celestron Focus Motor on the Scope . The iPad SkyPortal App already has a function for this as does the normal Hand Controller . I have only been eyeballing the Focus using the zoom function on the DSLR .

So , manual focus , no Guide Scope or camera , no laptop control . Just the basics . I'd like to add a Wedge next year so I can do a little longer exposures without the Field Rotation issues . The StarSense camera makes this easy too .

Off-axis Guider hmmm . John M. is your man to talk to on this . James Lamb had a great Video on using the Celestron one and using Stellarium to set up the correct rotation to get Guide Stars on the Guide Camera sensor . John says it is working quite well .

I am not using the latest CaLIGHTs...yet . There will be a Posting later from John M on this as he is still having some interesting issues . We talked quite a bit last night about what he is seeing and we shared some Screen captures on a few differences . So , for now , I'll finish my ReProcessing tasks with the 3.10.90 and wait to see what you think about John's interesting findings .

Scott....................

[wols...@gmail.com]

Scott,

I attached a photo of my "solution" for focusing my EdgeHD scope.  This mechanism is electrically identical to the Robofocus and to the Moonlite focuser.  It's also me tinkering.  When John contacts me I will discuss his experiences with an off-axis guider.

This year I used a program that I wrote that tries to address differential flexure.  The program analyses each LIGHT frame and identifies all the stars.  It compares the star locations between images and tells PHD2 to ramp the guiding setpoint to stop the stars from drifting.  It works well but it's not as definitive as using an off-axis guider. I am very interested in whether John has seen a marked improvement in focus when using an off-axis guider.

Peter

[wols...@gmail.com]

Scott,

The other aspect of focusing a SCT is to ensure that your last rotation of the knob is in the direction that moves the primary mirror upwards...against gravity.  It's basically a backlash issue...BUT... I calculate FWHM for every LIGHT frame and I have not see any significant degradation in FWHM regardless of which direction I last rotate the knob so I am not convinced that backlash or thermal effects are significant.  I suspect that FWHM is more influenced by seeing because even though I consistently focus, I notice shifts in FWHM from night to night.  Also I notice that FWHM begins to suffer in tandem with autoguiding, which I believe is a seeing issue.

Peter

[Scott Kuchma]

Hello Peter ,

I took a chance and redid one of my most favourite Images from 2020 , M82 , the Cigar Galaxy . I was so happy when the final Image appeared on my Monitor that I haven't dared to retry it , until today .

I still think there is a tiny bit of improvement to be found because I'm not using all the tools in all the Software Packages but this latest effort has me smiling yet again .  Enjoy !

This is the original effort...

...and now today's latest efforts...

This will cost me a few bucks as I've been having my "best" Images printed for show and tell and I'll have to do a lot of them again . LOL .

I think this improvement was down to simply adjusting my Sequence so that I do the Sharpen AI first then the DeNoise AI . It is quite the obvious change .

Now onto a few more...

Scott................