Flat Earth Radio

• Very simple server configuration; Icecast just accepts a stream from the show host and spews it back out to the listeners.
• Because the radio stream itself is created on the show host's local computer, they have maximum flexibility to use whichever source client suits their particular show best.

• Potentially higher entry barrier to host a show, as it requires configuring a source client first. It's not hugely involved, but I know we have some non-technical folk here like Thork who might have difficulty just because it's new and unexpected.
• Every new host is a potential source of new problems, if things don't just work for them.
• Following on from the above point, the radio stream isn't guaranteed to maintain a constant format between shows. It could fluctuate between mp3 and Ogg Vorbis, and it will certainly drop out and require a reconnection from all clients when one show ends and the next begins. This creates the potential for one show to just work for a given listener, and the next to not work at all.
• Having multiple people talking together on the same show becomes a nightmare to facilitate, as one of them will need to act as a VoIP server for the other and then feed that data up to Icecast. This also means there is one extra lossy encoding step for one of the speakers, degrading audio quality unnecessarily.

• Consistent, continuous streaming from the listeners' point of view. No breaks or format changes between shows.
• Depending on how the source client accepts audio from the show host, it could be made very easy to configure, with a low entry barrier.
• Multiple speakers on one show becomes trivial; simply accept connections from multiple people.

• Server configuration becomes more complex to set up and maintain.
• An additional lossy encoding step unnecessarily degrades all audio quality.
  
• Limited user flexibility, as the stream encoding is done server-side.

• Very simple server configuration; Icecast just accepts a stream from the show host and spews it back out to the listeners.
• Because the radio stream itself is created on the show host's local computer, they have maximum flexibility to use whichever source client suits their particular show best.

• Potentially higher entry barrier to host a show, as it requires configuring a source client first. It's not hugely involved, but I know we have some non-technical folk here like Thork who might have difficulty just because it's new and unexpected.
• Every new host is a potential source of new problems, if things don't just work for them.
• Following on from the above point, the radio stream isn't guaranteed to maintain a constant format between shows. It could fluctuate between mp3 and Ogg Vorbis, and it will certainly drop out and require a reconnection from all clients when one show ends and the next begins. This creates the potential for one show to just work for a given listener, and the next to not work at all.
• Having multiple people talking together on the same show becomes a nightmare to facilitate, as one of them will need to act as a VoIP server for the other and then feed that data up to Icecast. This also means there is one extra lossy encoding step for one of the speakers, degrading audio quality unnecessarily.

Lol, "butt."

Quote from: Saddam Hussein on August 04, 2012, 02:07:52 PM

Lol, "butt."

Steve, would it be possible to setup a very small linux distro virtual system?  That way each speaker could have the same system setup as everyone else without having to worry about configuring it themselves.

Steve, would it be possible to setup a very small linux distro virtual system?  That way each speaker could have the same system setup as everyone else without having to worry about configuring it themselves.

Quote from: Lorddave on August 04, 2012, 08:53:00 PM

Steve, would it be possible to setup a very small linux distro virtual system?  That way each speaker could have the same system setup as everyone else without having to worry about configuring it themselves.

It's possible, but that's a retarded idea. It's overkill; you don't need the entire OS to be identical, some people's computers may not be capable of running a VM efficiently (not to mention the lack of decent cross-platform virtualisation tools), and you still have the uncertainty of how audio is fed into the VM. If people want to play music in the VM, they need to figure out a way of copying it there from the host OS.

No, if I was going to make something that's "identical" for everybody, I'd write a simple web form to allow uploading an audio stream or something. But I'm not convinced that's the best way to go, as it has its own limitations.

Then I'd go with option 2.  The degradation of audio quality probably won't be noticeable if it's only speech.  You really only need radio quality, which I believe is 64 kbps.

Posted by: Parsifal
« on: August 05, 2012, 03:07:59 AM »

Posted by: cmdshft
« on: August 19, 2012 »

Looks like free software has made this happen real fast.

Just use SHOUTcast.

Jesus Christ, fuck off. The freeness or non-freeness of my software has nothing to do with my limited free time.

I did get some time to look at this again about a week ago, but there's still the problem of how to mix streams together from multiple users in a way that makes it easy for them to contribute.

If you think you can do a better job of it, go right ahead, but don't come in here just to criticise others.

SHOUTcast takes maybe an hour to setup properly.

Why would you want to do that anyway? Radio stations don't accept streams from other people to listen to, otherwise they'd have their own streams being broadcasted. Sounds dumb and a waste of time, like you.

Anyway, if you have a stream server already setup, why are you bothering asking for streams? Why not just broadcast and then figure out all the hard shit along the way?

Besides, I don't recall seeing what you intended to broadcast unless it's somewhere in another thread. Music? Talk?

If it's talk, just use a VoIP app and find a plugin that can take the audio stream from that and route it into Icecast. I know for sure SHOUTcast offers such a plugin, I used to use it when I was going to school. Or maybe Ventrillo has some kind of solution and setup a virtual audio card to stream from.

I will just play music nobody likes for an hour or two.

I will broadcast the sound of me rubbing a microphone against my beard for two hours.