You are clearly having trouble understanding me. If KBLB took a great deal of care to disguise the true quantity of silk they had produced, one would have to conclude that this was done deliberately. "Instead of showing the whole pile of silk we produced, let's make it look smaller by artfully arranging the pile and the camera's viewpoint to minimize the apparent result." Suppose KBLB had actually done that. Why? My assumption: they would only engage in that effort as a way to deliberately mislead people who looked at the picture. What's your explanation for why they did all of that work?
I did not use 300 dpi as a conversion factor in my calculations. DPI is irrelevant in my calculations. I measured the diameter of cocoons in pixels and the known distribution of cocoon widths (in mm) to perform the size computations. The use of diameters is important, because diameter is far less variable than length if objects are partially occluded and are not parallel to the image plane. It is also important to use cocoons at the bottom edge of the image. Those measurements give me the scale for estimating the width of the image. If KBLB had posted a much larger image, the measurements would be off by, at most, a pixel or two. Often the differences would be sub-pixel. That's not going to change the calculations appreciably.
What you seem to be postulating is that, coincidentally, KBLB chose a one-point perspective view of their pile of cocoons, even though that is a difficult photograph to take and disguises the true volume of their pile of silkworms. And they did this for no particular reason at all. Unfortunately, I find your explanation to be nonsensical.
Market Data 
Markets 
AI
