Raw contrast as function of iterations for the perfect system with instrumental polarization (p) effects. At the last iteration, the RH is blocked to compare the raw contrast in the two channels. Panel a: during the acquisition of the response matrix and the closed loop tests the effects of p were not corrected. Panel b: during all steps of the response matrix acquisition and WFC tests, p was corrected. All three tested levels of p now overlap and reach the performance level of the benchmark tests.