CaLIGHTs Version 3.1.0.80 has been released

[wols...@gmail.com]

Latest Revision 3.1.0.80

-Dark Optimization would cause CaLIGHTs to run out of memory if the image was too small.  This bug has be eliminated.

-Finally got the math working correctly for the Calibrate Preview screen.  Lots of research into color balance, sRGB color space, gamma and asinh stretching.  The Calibrate Preview uses a fixed color balance of Red=2.0, Green=1.0 and Blue=1.5. Hopefully this screen will be easier to use now.

-I have noticed that a few DSLRs don’t implement the equivalent of BlackADU and some DSLRs(Canon EOS 1000D) reports the wrong BlackADU.  CaLIGHTs did use BlackADU in an attempt to optimize the available resolution for the calibrated FITs files that it creates.  The dedicated astrocam manufacturers don’t use the concept of BlackADU so I decided to abandon the concept of BlackADU.  CaLIGHTs will still display a BlackADU value if one is defined but it will not be used to calibrate your LIGHT frames. The main issue will be some clipping of the brighter pixels that have been brightened because of FLAT frame compensation. It all depends on how significant the vignetting of the LIGHT frame is and whether there are bright stars in these areas.

-Modified all of the buttons so they have the same look and feel.

-Changed how CaLIGHTs creates FITS files.  The FITS file format was invented in the late 1970’s. It has many requirements that were specific to the computer equipment in use in those days. Magnetic tape and computer languages like FORTRAN were popular then.  This creates some quirks that don’t make sense today but are still required…some of which caught me by surprise. Many of today’s astronomy programs no longer care about these quirks but Deep Sky Stacker, which uses the CFITSIO library does care and DSS V4.2.5 was not liking CaLIGHTs FITS files.  To bring CaLIGHTs in line with this standard the following changes were made…

a)The file size is now padded with zeros so that the size is a multiple of 2880. I believe this was a magnetic tape requirement.

b)Keywords like SIMPLE, BITPIX, NAXIS, NAXIS1, NAXIS2, EXTEND, BZERO, and BSCALE were reformatted.  All this means is that these keyword were formatted by CaLIGHTs as…

SIMPLE  = T                    / file does conform to FITS standard

BITPIX  = 16                   / number of bits per data pixel

NAXIS   = 2                    / number of data axes

NAXIS1  = 4164                 / Width of Image

NAXIS2  = 2794                 / Height of image

EXTEND  = T                    / FITS dataset may contain extensions

BZERO   = 32768                / offset data range to that of unsigned short

BSCALE  = 1                    / default scaling factor

 

And now they are formatted as per the FITS standard V4 as..

 

SIMPLE  =                    T / file does conform to FITS standard

BITPIX  =                   16 / number of bits per data pixel

NAXIS   =                    2 / number of data axes

NAXIS1  =                 4164 / Width of Image

NAXIS2  =                 2794 / Height of image

EXTEND  =                    T / FITS dataset may contain extensions

BZERO   =                32768 / offset data range to that of unsigned short

BSCALE  =                    1 / default scaling factor

 

c)CaLIGHTs would either add or modify keywords such as ISOSPEED, EXPTIME, CCD-TEMP, INSTRUME, DATE-OBS.  Now CaLIGHTs doesn’t modify or add them if the original LIGHT frame was a FITS file.  If the original LIGHT frame was a DSLR RAW file or a TIFF, CaLIGHTs does add them because CaLIGHTs needs to create a FITS file so some sort of header needs to be created.  The ISOSPEED keyword is recognized by Deep Sky Stacker which causes DSS to display the ISO.

-A new strategy has been adopted by CaLIGHTs for the special case where the original LIGHT frame is a COLOR image and the user has chosen to perform 2 x 2 or 4 x 4 MONO binning.  For this specific case, if the MOSAIC, COLORTYP or BAYERPAT keywords are used in the LIGHT frame then these keywords will be removed when CaLIGHTs creates the MONO, or luminance, calibrated FITS file.

[Scott Kuchma]

Hello Peter ,

So , if I'm reading this correctly , your previous version *did* use BlackADU to calibrate Light Frames from DSLR's and the new version will not ?

Since it now sounds like there will be the potential for issues for DSLR Users , I think I will stick with the previous version as I have not had any of the issues you mentioned .

If I'm wrong please let me know .

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

[wols...@gmail.com]

Scott,

I would sooner see you start using the latest version and let me know if you experience a problem.  If you stay with the previous version then you will never be able to update and I will never know there is a problem. To the best of my knowledge, the only way this change could affect your calibrated LIGHT frames would be to risk saturating your stars at the very edges of your images because of the FLAT frame compensation.  Please understand that all that FLAT frame compensation does is try to brighten the stars at the edges of your images back to the same brightness they would have been if they were in the centre of your image.  In other words...if the star would have been saturated at the centre of your image it will most likely be saturated at the edges of your image.

Peter

[Scott Kuchma]

Hello Peter ,

Why would I never be able to update if I stay with the previous version ? Are the Releases accumulative or stand-alone ? I don't understand .

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

[wols...@gmail.com]

I guess what I am saying is that if you don't update and help me verify that everything works for your specific camera then any future update will have the same stumbling block for you.  If I invent a feature that is perfect for your camera and release a new version it will also include this latest code that no longer uses BlackADU.  Every new release includes all of the changes from the previous version.  Your risk is extremely small...after all...if you find a problem and let me know, I fully expect I will be able to fix it in very short order.

Peter

[Scott Kuchma]

OK , I understand what you are saying now . Sadly I won't have any new Images to process until Spring . Too cold for me now .

I'm happy to help but it might be a while before I have anything to contribute .

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

[wols...@gmail.com]

Scott,

One of the great things about astrophotography is when you have a new program or a new version of a program you can go back to previous imaging sessions and try different approaches to see if you can improve your technique.  If you wish you could recalibrate your LIGHT frames from a previous imaging session using the new version of CaLIGHTs to both help me test and for you to try different approach...maybe adding filtering...or experiment with binning.  You might find a combination that gives you a better final product.  I plan to use the winter months to recalibrate and reprocess some of my old DSLR images to see if the latest version of Startools can improve my results.  I also plan to learn how to use AI Denoise. I expect to have lots of free time this winter due to this pandemic.

Peter

[Scott Kuchma]

Hi Peter ,

You're right of course , I can always re-Process some earlier Images and see if there is a change in the outcome . I've got a couple I could try and it's going to be a long Winter .

I keep waiting for SiriL to finish their Beta Testing and release a stable v1.0 Candidate but there has been no activity there for a few months now .

With the latest version of DeNoise Ai , the Comparison Mode is very useful and both John and I have noticed that the better of the three Options seems to be the Low Light Mode rather than DeNoise . Interesting .

Each Image will be different based on the amount of Data present . I tried the Trial version of Sharpen Ai and it struggled to provide any improvements in the Astro Images . I'll remove it later this Month .

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

[wols...@gmail.com]

Scott,

You got me thinking about this BlackADU issue.  Honestly, I'm not happy abandoning BlackADU.  It's just that I discovered recently that the Canon 1000D reports a BlackADU of 1023 and when I look at the pixel values they are way down at ~300. When this happens CaLIGHTs would try to create a calibrated LIGHT frame but it would contain virtually all zeros which is just horrible bad news. I've even found situations where the MaxADU value returned by LibRAW is smaller than the pixel values in the data file. 

I still think that BlackADU does offer the ability to avoid data clipping at 65,535 and that is worth keeping.  I have an idea rattling around in my head. I might be able to give users the ability to keep the concept of BlackADU but they would be responsible for telling CaLIGHTs what it should be.  It's obvious to me that I can't trust the camera manufacturers to give me reliable values.  Luckily, this BlackADU issue does not affect your MASTERs.

Peter

[Scott Kuchma]

Hi Peter ,

Do you need examples from my Canon T6i DSLR to help with this Topic ? I know John has been using a T6 up until recently so I'm sure he could supply some Data from that Model as well .

Where would I find that Information ?

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

[wols...@gmail.com]

Scott,

Sure...if you can put a couple of LIGHT frames up on dropbox or simular I can keep them for testing purposes.

Peter

[wols...@gmail.com]

Sorry Guys...I found a bug in CaLIGHTs Version 3.1.0.80 that only applies to DSLR users.  Hopefully I will release an update later today.

Peter