I was experimenting with the Deep Speech model implemented by Mozilla, but the system (Ubuntu 16.04) tells me “Stderr: sox FAIL formats: no handler for detected file type `flac'”, which seems that the Sox do not support ‘flac’ format in the system library.

The default SoX version that we can install through apt-get on Ubuntu 16.04 is 14.4.1.5, which does not support FLAC audio format on my computer.

1. Firstly, we have to install FLAC on Ubuntu:

   sudo apt install flac

2. Then we have to install SoX from source:

   Download the repository from:
   https://sourceforge.net/projects/sox/files/sox/
   
   Then open the folder:
   tar -xvzf sox-14.4.2.tar.gz
   cd sox-14.4.2
   
   Then configure:
   ./configure
   
   Finally, make and install:
   sudo make -s && sudo make install

3. Then we can install pySoX and run the Deep Speech model! 🙂