Tonal data of one frame of a video signal is subject to time division into dark subframe tonal data at a low tone and bright subframe tonal data at a high tone to compensate for a lowered luminance in the dark subframe by the bright subframe so as not to change an integrated luminance per one ordinary frame. Subframe tonal data read from a memory is converted into output tonal data constituted of dark subframe tonal data at a low tone and bright subframe tonal data at a high tone, by four arithmetic calculation operations using calculation parameters supplied from a register.