All this talk about scaling inspired me to try something with a photograph of a board. I took a picture of a Channel Islands M4 from the CI website and enlarged it until the length was ~17". I then measured the 1/2 width every cm, except for the front and rear 1/3 of the board, which I measured every 0.5 cm. The widths were measured to the nearest 0.5 mm. I plotted the points (solid line) and used 2nd-order polynomial regression in Excel to fit a trendline to the measured points (dashed line). Interestingly (or perhaps not) the fitted line matches the measured curve exactly in the first 1/3 or so of the board. Measurement errors are apparent by the fluctuations in the curve near the tail. These will have to smoothed out before cutting out the blank. I adjusted the chart scale to get the proper aspect ratio to give me a 6’1" x 18.4" template when scaled up. My plan is to scale the picture up and print it to have a rough template. After tracing onto the blank, I’ll “eyeball” the curves and make adjustments where necessary.
hmmm… where’s the picture? I’ll try again.
Looks close enough… I’d go ahead and plot on some masonite as close as possible, cut a little outside the line and true it up with one of those razor planes and a long sanding block. I’ve found I can shape a cleaner line with a longblock than with a pencil. Time spent truing your masonite template will pay off when you start working on the blank.