__FM

a personal audio content curator – cleaning your local radio from stress-inducing influences in real-time

this is v0.1, supposed to be a proof of concept. there's a lot of work to be done and everyone is invited to improve it!

Full documentation can be found here

Start

brew install libav #file conversion wav to flac

# install all the other libraries
# ...

# export google speech rec credentials
export GOOGLE_APPLICATION_CREDENTIALS="path/...1c7b.json"

# run the stream analyzer
getRadioStream.py

# after 15 seconds run the clean stream player
playCleanedRadio.py

To Do

getRadioStream

• put everything into one script
• improve the audio quality
• better timestamps with different speech rec library
• make everything async

getRadioStream

• improve performance of audio player (and bring it to getRadioStream)
• delete used files

Another Idea

Another idea came to my mind, to pipe the radio stream through super collider and dynamically fade to a second audio track when the word is discovered. (This would not allow for dynamic lengthening to add longer / shorter words though.