Should the audio be normed to [-1, 1]? - Githubissues

modelscope / FunCodec

FunCodec is a research-oriented toolkit for audio quantization and downstream applications, such as text-to-speech synthesis, music generation et.al.

https://funcodec.github.io/

MIT License

371 stars 30 forks source link

Should the audio be normed to [-1, 1]? #48

Open FFYYang opened 4 months ago

FFYYang commented 4 months ago

Should the audio be normed to [-1, 1]?

ZhihaoDU commented 4 months ago

There is not need to normalize audio if your audio is a standard .wav file with singed int 16.