hehehwang / voice-identicon-web

voice-analysis

• praat
• my-voice-analysis
• jdenticon

목소리의 특성을 반영한 아이덴티콘을 생성해준다. (목소리가 저음일수록 차가운 색으로 생성)

사용법

1. pip install praat-parselmouth
2. node app.js
3. 목소리.wav를 업로드
4. 짜잔

예시

hello World! ①

설명: 파파고 TTS, 여성, 3번 반복

파일명: helloWorldx3.wav

결과물:

{"number_of_syllables":"11","number_of_pauses":"2","rate_of_speech":"3","articulation_rate":"4","speaking_duration":"2.9","original_duration":"3.8","balance":"0.8","f0_mean":"239.72","f0_std":"68.75","f0_median":"245.7","f0_min":"88","f0_max":"336","f0_quantile 25":"208","f0_quantile 75":"301","prob_pron":"0.8","gender_analysis":{"Success":true,"gender":"Female","mood of speech":"speaking passionately"}}

hello World! ②

설명: 파파고 TTS, 남성, 5번 반복

파일명: helloWorldx5.wav

결과물:

{"number_of_syllables":"19","number_of_pauses":"4","rate_of_speech":"3","articulation_rate":"4","speaking_duration":"4.7","original_duration":"6.6","balance":"0.7","f0_mean":"133.11","f0_std":"22.03","f0_median":"134.6","f0_min":"82","f0_max":"168","f0_quantile 25":"118","f0_quantile 75":"153","prob_pron":"0.8","gender_analysis":{"Success":true,"gender":"Male","mood of speech":"Reading"}}

further discussion