The experimental errors for recording a high-resolution electron micro
graph (HREM) on film are identified and discussed. Both the systematic
and random errors are considered and are found to be of roughly the s
ame magnitude. However, by making the assumption that images of adjace
nt atomic columns in a micrograph are images of identical atomic colum
ns, the random errors can be reduced by averaging. Knowledge of the er
rors associated with quantifying high-resolution images allows for a m
ore rigorous comparison to simulated images.