We study the effect of separate chemical and kinetic freeze-outs to the ideal hydrodynamical flow in Au + Au collisions at RHIC ( energy). Unlike earlier studies we explore how these effects can be counteracted by changes in the initial state of the hydrodynamical evolution. We conclude that the reproduction of pion, proton and antiproton yields necessitates a chemical freeze-out temperature of T ≈ 150MeV instead of T = 160–170 MeV motivated by thermal models. Contrary to previous reports, this lower temperature makes it possible to reproduce the p T spectra of hadrons if one assumes very small initial time, τ ₀ = 0.2 fm/c. However, the p T differential elliptic flow, v ₂(p T) remains badly reproduced. This points to the need to include dissipative effects (viscosity) or some other refinement to the model.