Making a WebRadio

deprecated! there should be a newer page

for now, have links:
LiquidSoap audio/video streaming language
Raspberry pi pirate FM radio transmitter
Open Source Speech Synthesis

The setup:

atm i have 2 docker containers, one with liquidsoap running on it and the other with icecast on it, which the liquidsoap is streaming to.

due to a quirk of the old laptop-turned-server im running it on, one docker project won't run while another is, so the webradio is barely up.

one of my main complaints about liquidsoap is that it assumes that you'll only have a collection of singles in a playlist, that then should be randomly selected from, which just isn't how i listen to music. I like to listen through complete albums, and then go from album-to-album depending on the feelings that the albums have. I would like to have a way to play a complete playlist, and THEN choose what to play next, but as far as i know there is no way to do that in liquidsoap.

icecast works well as a webradio host, with the exception that there is no 'play' button (you have to manually navigate to where the stream is to listen to it). I have enjoyed setting one of my raspberry pis up as a FM transmitter and then streaming the audio through that, so i can listem to it on my small portable radio using SoX to convert the stream to a .wav format*.

my next steps are probably to find a different way to choose which audio to output to the icecast server (sorry liquidsoap but ur a bit too simple for me), and to do some experimenting with text-to-speech systems & programs.


*

sox -t ogg "http://[ip_address]:[port]/radio_1.ogg" -r 22050 -c 1 -b 16 -t wav - | sudo ./fm_transmitter -f [frequency] -

Update

I've switched my approach to using a python script and python-vlc outputting to a http stream with the python script controlling what music is play.

at the moment im not working on it too much, life is v. busy >~<