xiph / rnnoise

Recurrent neural network for audio noise reduction

BSD 3-Clause "New" or "Revised" License

4.05k stars 895 forks source link

how to use the data ? #81

Open ucasiggcas opened 5 years ago

ucasiggcas commented 5 years ago

Dear, Now I have the rnnoise_contributions.tar.gz and TSP48kall.zip , but where should I put them ? Need tar ? how to use ?

ucasiggcas commented 5 years ago

Any one has got success ? Please help me to reproduce the project.

ucasiggcas commented 5 years ago

Now I have the noise_only.wav and the speech_only.wav ——sample rate 48000 What should I do next ? Thx

ucasiggcas commented 5 years ago

you mei you dalao gaoding zhege de ?

ucasiggcas commented 5 years ago

teach me a little

bouderb commented 5 years ago

Hi ! Can you show me your rnnoise_demo.c file in $PROJECT_PATH/examples/ ?

Convert your wav files in pcm because this algorithm only accept 16 bit pcm file:

ffmpeg -i file.wav -ac 1 -ar 48000 -f s16le -acodec pcm_s16le file.pcm

Compile the project and run your executable file named rnnoise_demo in /examples/ (I'm not sure about the name) with arguments used in rnnoise_demo.c file

I did some modification to make this project work so it's normal..

ucasiggcas commented 5 years ago

Hi ! Can you show me your rnnoise_demo.c file in $PROJECT_PATH/examples/ ?

Convert your wav files in pcm because this algorithm only accept 16 bit pcm file:

ffmpeg -i file.wav -ac 1 -ar 48000 -f s16le -acodec pcm_s16le file.pcm

Compile the project and run your executable file named rnnoise_demo in /examples/ (I'm not sure about the name) with arguments used in rnnoise_demo.c file

I did some modification to make this project work so it's normal..

Dear, Thanks a lot Now I have solve the problem, but another ups If My datasets are 16k sample rate ,what parameters should I change or how to modify the codes ?

Thanks for your kind help.

bouderb commented 5 years ago

Hi ! I didn't try but I think you just have to change the value of st->sample_rate in your rnnoise_demo.c file. No prob' !

ucasiggcas commented 5 years ago

Hi ! I didn't try but I think you just have to change the value of st->sample_rate in your rnnoise_demo.c file. No prob' !

No No sample_rate in the script, and other parameters also need change, such as

define PITCH_MIN_PERIOD 60

define PITCH_MAX_PERIOD 768

define PITCH_FRAME_SIZE 960

eband5ms[]

vaishalibhardwaj commented 4 years ago

Hi ! Can you show me your rnnoise_demo.c file in $PROJECT_PATH/examples/ ?

Convert your wav files in pcm because this algorithm only accept 16 bit pcm file:

ffmpeg -i file.wav -ac 1 -ar 48000 -f s16le -acodec pcm_s16le file.pcm

Compile the project and run your executable file named rnnoise_demo in /examples/ (I'm not sure about the name) with arguments used in rnnoise_demo.c file

I did some modification to make this project work so it's normal..

My denoised file is not streaming, I get "couldn't stream this file" sample rate in properties is shown 0. What to do here?

KAWAKO-in-GAYHUB commented 2 years ago

Hi ! I didn't try but I think you just have to change the value of st->sample_rate in your rnnoise_demo.c file. No prob' !

No No sample_rate in the script, and other parameters also need change, such as #define PITCH_MIN_PERIOD 60 #define PITCH_MAX_PERIOD 768 #define PITCH_FRAME_SIZE 960 eband5ms[]

Hi, how to modify these parameters if 16K?