In spontaneous conversational speech, all portions of speech do not always have high clarity. For example, the portions not having important information or the end of a sentence are not very clear. We consider that clarity of speech is controlled by F0, power, speech rate, place of articulation and so on. We consider that the clarity changes continuously, and change of clarity of speech produce a fluent rhythm in human speech. The purpose of our research is introducing the change of clarity into synthesized speech. In this paper, we try to control clarity of synthesized speech by post-processing of F0, power and formants. We evaluate the synthesized speech by auditory tests using SD method. The synthesized speech with control of clarity is better than the synthesized speech without control of clarity in several speech properties, e.g., calmness and smoothness.