wow, lots of posts I've not read since I was on yesterday.
Looking at the original images with my improvised color filters, there seems to be some extra structures that need to be explained. The structures are clearly there, but they are also mixed with JPG artifacts. I will describe the structures in relation to the following image which I posted earlier.
I notice three additional features in my images:
1- I am finding that the lower "wing" turns upward at exactly the point where the image intersects the pole, and continues upward until about the level of the bottom of the "abdomen." I'll call this the upswing.
2- It seems to cross under a dark appendage which extends diagonally downward from the "abdomen." I'll call this the dark appendage.
3- The "abdomen" has a large dark area associated with it, centered almost exactly on the top of the pole. I'll call this the dark area.
(I use the word "exactly" for the benefit of the pole theorists)
In filtering for a more bluish wavelength, the dark area becomes apparent, the upswing become highlighted, and the dark appendage appears as a shadow. In filtering for a more reddish wavelength, the dark appendage becomes more highlighted and the upswing and also the initial downward "wing" become a shadow.
The top left image is an inverse of the usual diff image. The top right image is a pre filtered diff which highlights the upswing very clearly, as well as the shadow of the dark appendage. The middle left image shows the dark area, which I diff against a filtered (before + After)/2 to try to show the size of the dark area more clearly in the middle right image. Note that the middle right image is intended to show pixels getting darker, and clip out pixels getting lighter compared with the BefAft image. The middle left image also shows the upswing and shadow of the dark appendage with only color filtering and no contrast enhancement. The bottom image is in a more reddish wavelength as mentioned with no contrast enhancement.
Don't ask me what it means.
BefAft = (before + after)/2;
filter1 is on color vector 60/53/49
"Pos" means negative results clipped out.
Top row images have enhanced contrast, also the second middle row image.
Top Row
1- invert(Pos(BefAft - Event)
2- Invert Pos((Event - Filter1(Event)) - (BefAft - Filter1(BefAft)))
The second frame of the top row
MiddleRow
1- BluishFilter(Event).
2- BluishFilter(BefAft) - BluishFilter(Event) -- as I recall.
The middle row first frame shows a larger dark area with its center at the top of the pole. The second frame is a contrast enhanced Pos Diff with the same wavelength filter of BefAft, with the intent to highlight the size of the dark area.
Bottom row:
ReddishFilter(Event)
I vote for bug theory! I don't see anyone who has shown a flaw in the theory that cannot be explained to the point where we must dislcude the bug theory. Also about that length of the streak, hazeii's experiment showed it is possible to obtain such a lenght:
