The objective of extracting the spatial location of a reflector, and its local angle-dependent reflection coefficient, from seismic data, depends on the ability to identify and to remove the effect on primary amplitudes of propagation down to and back from the reflector. All conventional methods that seek to correct for such transmission loss require estimates of the properties of the overburden. In this paper we propose a fundamentally new approach that will in principle permit correction of primaries for such transmission loss without requiring overburden properties as input. The approach is based on the amplitude of the first term of the inverse scattering series internal multiple attenuation algorithm, which predicts the correct phase and approximate amplitude of first order internal multiples. The amplitude is estimated to within a factor determined by plane wave transmission loss down to and across the reflector producing the event’s shallowest downward reflection. Hence, the amplitude difference between a given predicted and actual multiple, both of which are directly available from the data and the algorithm output, in principle contain all necessary information to correct specific primary reflections for their overburden transmission losses. We identify absorptive overburdens/media as requiring particular focus, so as a first step, previous amplitude analysis of the internal multiple attenuation algorithm is here extended to include stratified absorptive media. Using this newly derived relationship between predicted and actual internal multiples, and existing results for acoustic/elastic media, correction operators, to be applied to specific, isolated primaries in both types of media, are then computed using combinations of multiples and their respective predictions. We illustrate the approach on synthetic data for the absorptive case with three earth models with different Q profiles. Further research into the ampli- tudes of the plane wave internal multiple predictions in 2D and 3D media is a likely pre-requisite to field data application of this concept-level algorithm.