Release date January 4th, 2022
-I have changed the minimum permission logic for calibration. I also changed the BLACK ADU Usage definitions. This is the result of a lot of detail work that focused on FLAT compensation. DSLR users will not notice any change in how their images are calibrated. This is because all DSLRs use the data convention called BLACK ADU. AstroCAMs do not use this data convention…yet…and this can cause headaches when trying to get acceptable FLAT compensation. I hope to post something to my website [astrohobby.ca] that explains this issue. Here are the main changes:
The BLACK ADU Usage selection on the Preferences dialog has been changed to the Pedestal Priority. There are four possible choices available here which influence the minimum calibration permission logic and the Pedestal value used for calibration. The choices are as follows:
Master DARK This choice should be considered the default. The minimum calibrate permissive requires either a Master DARK or that the LIGHT frame has a BLACK ADU value. For DSLRs there is always a BLACK ADU value. For AstroCAMs, if a Master DARK is specified, the Pedestal will be determined by the Master DARK. If there is no Master DARK specified then the “BLACKADU” FITS keyword value from the LIGHT frame will be used for the Pedestal…if the FITS header does not contain a “BLACKADU” keyword, calibration will not be allowed.
LIGHT Frame This choice causes the minimum calibration permissive to insist that the LIGHT frames have a BLACK ADU value. This is no problem for DSLRs but I believe that I am the only AstroCAM owner who has an AstroCAM that writes a “BLACKADU” value to the FITS header. I suspect this will happen with other AstroCAMs at some point…I have seen some FITS keywords, that look like they may be equivalent, that have been made available for some AstroCAMs. The BLACK ADU value for each LIGHT frame will be used to determine the Pedestal. I believe this will yield the best FLAT compensation…for my AstroCAM. The only quirk with this choice is that when the calibrated LIGHT frame is written to disk, the BLACK ADU for the first LIGHT frame is used for all the calibrated LIGHT frames. This ensures that the Sky Glow Comp feature works correctly and that all of the calibrated LIGHT frames have the same total darkness setpoint.
Master BIAS This choice causes the minimum calibration permissive to be either that a Master BIAS is specified or that the LIGHT frame has a BLACK ADU value. There are some astrophotographers that don’t bother taking DARKs and just use BIAS frames. This is why this choice is available.
Custom This choice causes the minimum calibration permissive to always be TRUE. The Pedestal value is set by the custom value that is specified. This custom value can range from 3 to 65,535.
-Added an Output Pedestal selector which gives users more control on how the data in there calibrated FITS files is optimized. This selector is available on the Preferences window. There are two settings:
Same as Input This is the default choice which causes CaLIGHTs to do what it always has done when creating calibrated output files.
Custom This choice allows the user to select a custom output pedestal value that can range from
-20000 to 10000. This is an advanced feature which is explained in detail here Preferences
-Modified the DARK Optimization logic so that it does not use pixel values close to the edges of the image. I don’t allow it to use pixels closer than 10% to the edges of the image. I hope this improves DARK optimization when cameras have Amp Glow. Typically, Amp Glow is only visible along the edges of an image.
-Modified the Noise Filter…a lot! I have been doing a lot of research regarding Poisson noise filtering. Poisson noise is the dominant type of noise encountered in the dimmest parts of our astrophotos. I read lots of scholarly papers which really gave my brain a work-out. I experimented with a few Poisson noise filtering algorithms and found that the existing filtering logic in CaLIGHTs is pretty good. I spent a week or so tweaking the CaLIGHTs noise filter and manage to get it to work even better. The noise reduction is now stronger. I also found that the existing CaLIGHTs noise filter tended to have horizontal and vertical artifacts that were small but visible at very high magnification. I was able to eliminate these artifacts.
I like the noise filter improvements so much that I now allow the noise filter to be applied up to 6 times instead of the previous 3 times. Because the noise filter uses a Pixel Threshold(SD) to decide what pixel need to be filtered there can be a dramatic change in noise where filtered pixels are right up against slightly brighter pixels that are not filtered. This is most noticeable when using the x4, x5 and x6 settings. To smooth out this noise transition, CaLIGHTs will add 0.1 SD to the first iteration of the noise filter when using the x4, x5 and x6 settings.
-Modified the Create Average Master Histogram so that it displays two different standard deviation calculations. The StdDeva values are the original standard deviation values. I was noticing that these values can vary widely when there is lots of hot pixels or Amp Glow. I suspect that this is happening because hot pixels and Amp Glow are not valid for a “normal distributions of values” or a classic bell curve. I added StdDevg values which uses the area under the curve to determine standard deviations. This method seems to be pretty much immune to hot pixels and Amp Glow. StdDeva means “Standard Deviation using the traditional algebraic equation”. StdDevg means “Standard Deviation using a graphical method.
-Added a Re-Load feature so that CaLIGHTs can take advantage of the HTML Calibration files that can optionally be created during a calibration. If you use the Preferences window to enable creation of these HTML Calibration files you can now go back and quickly reload all of the Master files, LIGHT frames and settings from a previous calibration. I do this a lot when I am testing CaLIGHTs or when calibrating LIGHT frames that I need to experiment with to get the best results. Sometimes it can take a day or so to tweak the calibration to eliminate hot pixels or apply the right amount of filtering.
-The ExposureComp feature has been eliminate and replaced with…
-Added a Star Saturation Guard [StarSatGuard] feature which is a user request. FLAT compensation typically causes the stars at the edges of a LIGHT frame to be brightened so that the vignetting effect can be eliminated. This can easily brighten these stars more than the maximum 65,535 data value which causes these stars to be clipped. This causes these stars to appear bloated and lose their unique colour. StarSatGuard allows the user to specify a limit value which is used to identify stars that may be clipped. CaLIGHTs will identify these stars and reduce their brightness so that they are not clipped and they retain their colour. I typically set the StarSatGuard limit to 65,000. This feature seems to work very well and I have been enjoying seeing how it improves my astrophotos.
To download CaLIGHTS click on this link