in the last part of the paper, he mentioned the slow response time
of his FET. The reason, as seems to me, is obvious. The gap between
two transparencies is too big for any effective diffusion. But the idea
of making grids by using a laser printer is pretty cool.
