We investigate the best method of characterizing high-divergence lasers, su
ch as excimer lasers, to suppress fine-scale intensity nonuniformity that i
s due to coherence effects of lenslet homogenizers. We show by a detailed a
nalysis of the lenslet homogenizer that, for highest accuracy, a direct mea
surement of the value of the autocorrelation function should be made at the
separation p of the lenslet elements, identified as the critical spatial p
eriod. We show that the commonly used characterization of lasers by the 1/e
(2) width of the angular divergence is not the most accurate test and may o
verstate or understate the effectiveness of a given laser. (C) 2001 Optical
Society of America.