To make a calling satisfactory in quality with a clear voice by outputting original data for voice synthesis in which a component of ambient noise is eliminated from a voice signal and converting the original data into an audible synthetic voice and outputting it.

A voice recognizing part 102 outputs original data for voice synthesis in which the ambient noise is eliminated by performing the voice recognition of the voice signal of a sound in which the voice of a speaker and the ambient noise are included. A voice synthesizing part 103 outputs a synthetic voice as an audible synthetic voice which a human being can hear from the original data for voice synthesis. A transmitting part 104 transmits the synthetic voice outputted from the voice synthesizing part 103 to the outside. Moreover, it is possible to make this voice processor have an ambient noise eliminating function by making the processor recognize tip shapes. That is, a lip shape recognizing part outputs original data for voice synthesis based on the shapes of the lip of the human being. In this case, since a voice signal in which the ambient noise is not included can be transmitted and also it is not necessary to perform a conversation by speaking with a voice, the noise problem of portable terminals in a public place can be dissolved.