SUMMARY
Along with the development of technology, the technique of identification of human voices based on sex is very necessary for some parties, especially in the field of education. The main parameter in distinguishing male and female voices is the frequency of the same one-word pronunciation sample with the same duration and pressure. To facilitate this, a matlab-based male and female voice differentiator application was designed and the results will be improved by the fuzzy logic method. Based on the application testing, of the 20 human voice samples tested, there was 1 sound sample that was difficult to identify, so the accuracy level was only 95%. However, by using the fuzzy logic method, 1 sound sample can be classified in the form of female voices, so the performance of male and female voice differentiator applications increases to 100%. The final results of this study obtained a range of male voice signal frequencies 81.93 Hz to 233.3 Hz while the frequency range of female voice signals 316.9 Hz to 575.5 Hz.