Home SEEE Student Project Page
Separation of Voice from the Song

The project uses a C program to separate vocal component from the mixture of other sound (song). The computer or the system is not previously trained to identify the voice or the words of the singer. It is not always possible to train computer to recognize words as different singers may have different voice spectrum and harmonics. Thus the audio or the wave file was first divided into the header and the waveform and separately saved. The header file was not modified. The waveform was analyzed using the FIR filters, band pass and high pass filter. The filters were used to separate the voice from the rest of the spectrum. The output of the filter was separately saved and it was merged with the header file to create new audio files. Thus the voice and the rest of the song was separated and saved as audio files. These files can easily be played by any programs which support wave files. SIMULINK can be also used to filter out the voice directly from the song. The filter bank which passes required frequencies only can be used to filter out the voice. Similarly filters can be used to filter out the orchestral component. The outputs of the filters are separately saved as audio file to get the voice song and the orchestral song.


  •     Rujuna Tamrakar
 

Course of Study

Course of Study

Project Page

Project Page

Feedback

Feedback

Click HERE to give feedbacks and comments on the site!

Encipher

Encipher 2009Download the technical journal, Encipher, published annually by SEEE.