Bald eagle cam showed what appeared to be the remanes of the youngest eagle, frozen in the snow, perpetually staring back at the nest. Being a public cam, they couldn't zoom in to confirm the identity & risk shocking the masses. The cam stayed wide from then on.
It must have wandered out during the snow storm & couldn't get back under the adult before the snow piled up. It froze, staring back. That was the last to hatch. The gene for wanderlust was strong in that one. Maybe wanderlust is a consequence of not being born simultaneously. That's how a lot of them go.
Kind of macabre for them to go about their routine like nothing ever happened, with their offspring's corpse staring at them. Such is the necessity of survival in the wild. The ravens would eventually take it away.
Another view showed a more likely corpse. It was still under the adult. #3 was still alive but looking a bit ragged. Snowstorms seem to be stressful events, despite millions of years of evolution.
The male baldy moved the corpse to the food pile, rather than farther away or throwing it off the nest. No undertaker in the wild. Then stared at it while brooding for 25 minutes before going back to scouting for ravens. Noted the camera zoomed out when the corpse came out.
Female bald eagle pulled the corpse out later, started eating it, then decided it wasn't food & flew off with it.
The story plays over in bird nests around the world. The live cam reveals unedited life in nature while also revealing the unedited brutality of nature.
------------------------------------------------------------------------------------------------------------------
Set to the task of trying to convert some exposure brackets to HDR. The common practice nowadays seems to be just pasting python commands in the command line interpreter. The libraries are high level enough to make the python command line as good as any dedicated image processing program.
Pretty good results with some 8 bit mp4 eclipse footage.
cv::MergeMertens is an HDR merging tool which doesn't need exposure knowledge. It processes images in floating point. The problem is opencv doesn't have any file I/O for floating point. It would entail writing a C++ program with custom file I/O to do it right.
The 8 bit version did worse than previous lightening of a single exposure.
The floating point C++ version was the 1st one clearly showing the eclipse & the foreground without a lot of noise. Helas, the best one continued to be the very 1st histogram correction in 2017.
There was no HDR through the T4I. That was a timelapse at constant exposure.
HDR from the mark 3 was once again a fail. Might have already tried an HDR hack years ago & gotten the same banding. The gamma in the raw CR2's might need correction.
Pushing the gamma way up got rid of the banding & revealed more of the corona than any previous effort.
Besides the lion motion blur, the corona in this one also looked better.
Reducing the maximum & using floating point got a hair more information from the lunar lander. The Debevec & Robertson algorithms only take 8 bits so those would be expected to fail just like 8 bit Mertens.
-------------------------------------------------------------------------------------------------------------------------
Liminal space
---------------------------------------------------------------------------------------------------------------
Looks like Retire Early Guy has the equivalent of $2 mil in his house, pension, investments & he lives pretty minimally in what looks like 600 sq/ft. If he hits 85 & the pension tracks inflation, the pension is equivalent to $1/2 mil of the total. A pension can only be compared to the risk free return on the equivalent cash, which is 0 to negative.
That puts things in perspective. Anyone bragging about retiring on less than $1 mil has a large sum in home equity, a pension, or a bank of marriage.
A typical Calif* teacher gets $3k/month from CAL STRS, equivalent to $1.7 mil over 30 years.
The required fund decreases with age. Retire Early Guy with $2 mil actually has a lower hill to climb than lions.
Comments
Post a Comment