Further evidence of separate processing of brightness and contours in illusory figures