Notes. We fixed or released effective temperature T₁ of the primary and the third light l₃. We prefer the model with fixed T₁ and l₃ (bold). The uncertainty σ is estimated the same for all models. The following numbers of data points were used: N_total = 321, N_LCB = N_LCR = 100, N_RV1 = 71, and N_RV2 = 50. T₀ denotes the time of periastron passage (^( * )means HJD−2457733); T₁ and T₂, the effective temperatures of the primary and secondary, respectively; R₁ and R₂, the radii of the primary and secondary, respectively; i, inclination; S_B and S_R, the coefficients adjusting normalisation of the flux defined in Eq. (4); m₁ and m₂, the masses of the primary and secondary, respectively; e₁, eccentricity; ω₁, the argument of periastron; γ, the systemic velocity; l_3R and l_3B, the third light in the blue and red filters, respectively; , the total value of χ²; , , , and , the contributions to χ² from the LCs in the red and blue filters and radial velocities of the primary and secondary; N_total, the total number of data points; N_LCB, N_LCR, N_RV1, and N_RV2, the corresponding numbers of data points. Uncertainties were estimated from Figs. 11 and 13.