Test audio sample clipped.

[kchanqvq]

I'm using WSJ0 dataset, and the model convergences. When I use the utility audio_test.py, however, I get only 3 second output when I feed it a 5 second mixed audio. Is that any issue in this python script causing the audio to be clipped?

[zhr1201]

Yeah, the test script just uses 3 chunks of frames. You can change that by modifying the AudioSampleReader.gen() into a python generator and iterate through all the frames in audio_test.py.

[dearwind153]

@zhr1201 Have you ever encountered audio files that cannot be played after voice separation？ I find the separate wav file's format as follow: Input File : 'out_1.wav' Channels : 1 Sample Rate : 16000 Precision : 53-bit Duration : 00:00:01.61 = 25792 samples ~ 120.9 CDDA sectors File Size : 206k Bit Rate : 1.02M Sample Encoding: 64-bit Floating Point PCM

when convert the former wav format as follow, which has't voice. my model is: 链接:https://pan.baidu.com/s/1jWo9fgoaTTN7Akeh5RbupA 密码:ax1s

Can you give me some advice?

Channels : 1 Sample Rate : 8000 Precision : 16-bit Duration : 00:00:03.13 = 25055 samples ~ 234.891 CDDA sectors File Size : 50.2k Bit Rate : 128k Sample Encoding: 16-bit Signed Integer PCM

[zhr1201]

@dearwind153 All the sample point are all 0 in your file. Guess some error happened and I am not very sure. You probably should run that step by step to see what is wrong.

[dearwind153]

@zhr1201 , Ok, thank you for your reply!