Two prior approaches to size processing are discussed in this paper. T
he first approach is based on measurements of mental size transformati
ons, the second on measurements of thresholds for size and separation.
We first analyze these prior approaches and point out differences amo
ng prior models and similarities among prior results. This analysis le
d to new psychophysical experiments that tested the effect of size, re
lative precision, and eccentricity on the speed of perceptual processi
ng. Speed was not affected by size, but was affected by relative preci
sion and eccentricity. These new results, along with prior results, ar
e then used to formulate a new model based on an exponential pyramid a
lgorithm. This new model, which uses elements of both traditional appr
oaches, can better account for prior, as well as our new results, on t
he time-course of size processing.