Disrupted mel with Tacotron2 model using studio recorded wav files.

TensorSpeech / TensorFlowTTS

:stuck_out_tongue_closed_eyes: TensorFlowTTS: Real-Time State-of-the-art Speech Synthesis for Tensorflow 2 (supported including English, French, Korean, Chinese, German and Easy to adapt for other languages)

https://tensorspeech.github.io/TensorFlowTTS/

Apache License 2.0

3.8k stars 810 forks source link

Disrupted mel with Tacotron2 model using studio recorded wav files. #741

Closed rashimihup closed 2 years ago

rashimihup commented 2 years ago

@dathudeptrai I have been trying to use Tacotron2 + fs2 model architecture to create audio from LJSpeech data format. The wav file used for the same is recorded in studio and has a very nice quality to it , but the mels generated are completely disrupted [image attached below]. Screenshot from 2022-02-08 08-09-10

stale[bot] commented 2 years ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs.