Localization or classical diffusion of light?

Wiersma et al.¹ have reported near-infrared optical transmission and coherent backscattering data from strongly scattering slabs of micrometre-sized semiconductor particles. Their optical transmission was much weaker, and the angular shape of their coherent backscattering more rounded, than would be expected for classical diffusive light propagation without absorption. The authors interpret this as evidence for the onset of strong localization of light, but we find that their data can be explained by classical diffusion combined with reasonable amounts of absorption. Moreover, the turbidities of their samples are much lower than those given in ref. 1 and are comparable to samples with classical transport properties. We therefore question whether their samples are in fact close to the proposed localization transition.