It has been in the back of my mind to write this up for MT since I retired from a certain large computer company. Inasmuch as Paul's article above is a perfect lead-in, there is no time like the present. Our lab supported a semiconductor integrated circuit and a ceramic substrate manufacturing facility. We were continually required to measure circuit line widths on plan-view specimens and layer thicknesses on cross-section specimens for both semiconductor and ceramic substrate specimens and we were often asked to determine thin film grain size and ceramic raw material particle size data. A large number of measurements were required for each specimen to guarantee statistically sound data. We had image analysis software available that we used whenever we could, but often found that measuring things on a system using grey-level image analysis as input simply did not work. This is especially true for thin film grain size determination when using diffraction contrast TEM images for input.