The human eye is a great thing (at least when its working in conjunction with an great occipital cortex.

For at least 2 years I have been trying to get the most information out of micrographs (AFM, TEM) of surfactant protein d just to determine how many peaks in brightness (luma, luminescence, lightness, brightness, whatever!!) and going through the whole lot of grayscale conversions and assessing the LUT tables for peaks and valleys in SP-D, i have come to the startling conclusion that the best plots are obtained when “I” that is my “eye” tells me that i have found a good balance between contrast and brightness, reduced the background to a tolerable point and have not lost any detail.  Out of four programs, I think my eye is best.  LOL.  That sounds totally unscientific, and I will provide the plots to demonstrate that the quality plots are when i determine the brightness and contrast…. next best might be using “luma” to convert RGB to grayscale.  The typical method (averaging RGB) is really not that great.  Just saying…  i am providing dozens of plot to verify…  but the best plots come when I adjust the parameters.

This sounds like bias….  nah…. its the same as training AI.