Metal opacity shapes the near-UV spectrum of late-type stars, which dominate intermediate and old stellar populations. Learning the details of how metal opacity blocks the light in this spectral region is of capital importance to understanding the energy balance in the atmosphere of these stars and, ultimately, building reliable models to interpret observed fluxes. The model atmospheres most used in spectroscopic analyses of individual stars and at the core of population synthesis codes are based on calculations of photoionization cross-sections from the 70's, when better data have been available for a long time. We implement modern cross-sections in our calculations of synthetic fluxes and model atmospheres, but the models need to be confronted with observations. Detailed absolute fluxes for stars of known effective temperatures and angular diameters can constrain the opacities directly from observations. So far, such high-quality UV observations are available only for the Sun, and this leaves some room for ambiguity between line and continuum opacity. Observations with identical quality are possible with STIS for a second nearby late-type star: Procyon A. This star is indeed the only relatively unevolved late-type star for which an extremely precise determination of its angular diameter is available.