From: blockedofcourse@foo.invalid   
      
   On 11/26/2025 3:43 AM, Mikko S wrote:   
   > Don Y  writes:   
   >   
   >> On 11/24/2025 12:04 AM, Mikko S wrote:   
   >>   
   >> Alignment is important because we want to know how cloud cover will   
   >> move to give us an idea of "what's coming".  *Watching* it move doesn't   
   >> tell us if it is moving the way we expect.   
   >   
   > I was referring to measuring the brightness of a certain star at given   
   > time, which required careful alignment and distortion correction. Your   
   > needed level of alignment is easy to do.   
      
   Actually, I was expecting to point at a specific area of the sky where   
   I would have expected a particular "object" to appear at a particular   
   (time,date).  Knowing the motions of the heavens, to be able to ascertain   
   what a particular part of the sky looked like in terms of cloud cover/density   
   by noting whether or not "something" appeared where I expected it to appear.   
      
   [I don't need to "see" the object; just register it as a light source   
   originating above the clouds.  E.g., a *cluster* of stars is a valid   
   "object", regardless of their individual appearances]   
      
   E.g., if the moon is "out" (even if during daylight hours!), then I should   
   be able to see reflected light from it wherever it is supposed to reside in   
   the sky.   
      
   > The suggested cameras and fisheye lenses in the Indi Allsky project are   
   > very sensitive and also sunlight tolerant, so this should be easy - if   
   > you can do some simple image analysis (AI can help you to code this!)   
      
   I already have scene analysis software that works.  But, is intended for   
   scenes with lots of contrasting detail (i.e., easily recognizable "features"   
   to align the algorithm with the default scene).  I don't know how well it will   
   fare for "varying degrees of light intensity" devoid of real "features".   
      
   But, I also have "forever" to keep correlating the observations with the   
   actual phenomenon:  "the last time I saw stuff like this, solar panel   
   output was down 23% from nominal" or "it rained the next day".   
      
   It's a discomforting problem space when you don't "need" an exact answer but,   
   rather, are just looking for some sort of reliable bias.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)