Slashdot Mirror
Linux Radio Station Automation?
miazmatic asks: "I am one of the tech managers for my high school's FM radio station. We have been using Rhythmbox on Debian to play music after hours when no one is broadcasting. However, it took some pretty ugly hacks to get it to transmit the station ID every hour. We are adding a 600GB RAID 0 VG to our PC (P4 2.4/512MB), to which we plan to encode all our CDs losslessly. Along with this upgrade I would also like to find a permanent solution for broadcasting the station ID hourly. Has anyone used Linux to run a radio station before? Can anyone suggest a F/OSS software package or solution? Any help is appreciated."
Skip RAID 0 like the plague if this is your main storage area. Much much better to go with RAID 1+0 or RAID 5. RAID 0 is like tempting the HDD Gods for HDD failure.
Otto Audio jukebox.
Mind Booster Noori
They have this tool called 'cron', maybe... ;-)
Paul B.
http://freshmeat.net/projects/livesupport
Why would you encode the CDs losslessly? Correct me if I'm wrong but I do believe that FM radio is about half the quality of a normal CD, so you could encode the songs in something like MP3 or Ogg and still not have your listeners notice a difference in quality. This would save you disk space so you could then run a redundant disk array to protect against hd failure.
Also, contact http://www.radioparadise.com./ The owner/operator was spotlighted in Linux Journal a few years ago about his control system, 100% homebrew, FOSS product. He might have quite a few tips for you.
Rivendell aims to be a complete radio broadcast automation solution, with the facilities for the acquisition, management, scheduling and playout of audio content. As a robust, functionally complete digital audio system for broadcast radio applications, Rivendell uses industry standard components like the GNU/Linux Operating System, the AudioScience HPI Driver Architecture and the MySQL Database Engine. Rivendell is being developed under the GNU Public License.
Ethics II Axiom 2. "Man thinks." B. Spinoza
Wow, overrated. You seem to assume that high school kids (or anyone else) working for a radio station are paid. And your average educational station puts out at a paltry 10W. I have a handheld CB with more output than that.
Some manage to do more than that, and they manage to be mostly member-supported, thanks to their pre-existing popularity. I only know of one, though.
So, in short, you're an idiot.
Terrorists can attack freedom, but only Congress can destroy it.
RAID 0, also known as striping, is useful for obtaining very high I/O bandwidth at the expense of making every one of the constituent disks a single point of failure. Lose any one of the drives, and you lose all of your data.
I can see that if you lose all of the data you won't really have lost much, since you can always re-rip it, but why take the risk? You don't need high I/O bandwidth for CD-quality audio... it's only 86KBps uncompressed, and an average ATA-100 hard drive can easily sustain over 200 times that data rate. Since you plan to encode the data, you'll need even less bandwidth, probably half as much if you're using a lossless codec.
I wouldn't use RAID-0 for this. Use linear RAID or, since you mentioned volume groups, just let the volume manager append the volumes. The result will allow you to keep most of your data in the event a drive fails and will be plenty fast for your application. If you're not tight on space, I'd say you might as well go for RAID-5 and get yourself a little insurance against data loss. If your 600GB consists of three 200GB disks, you'd still have 400GB of RAID-5 storage, which is a lot of audio CDs. About 800 of them with no compresssion, double that with lossless compression, or quadruple that with high-quality lossy compression.
Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
FCC rules state that the annoucement has to be at the top of the hour, or as close as possible to that time, every hour. For example, if your song finishes at 09:59:47, you're golden. If it finishes at 10:01:05, you can fade the song out before the top of the hour, and play the ID. That works well with instrumentals, but not so good with the vocal stuff.
On the other hand, you can wait till the song is finished, and THEN play the ID. Problem is, if you're running satellite or rebroadcasting a larger or distant station, you're a slave to their timing. If their program starts at 1:00:00, then you almost have to fade the song before the top of the hour, before you cut back to the satellite feed.
You have a constitutionally protected right to be wrong, and I the right to ignore you.
Having worked at a high-power FM station in the Atlanta area within the last year, you have to do a legal ID within 5 minutes of :00, although if it's 5 minutes before, or 5 minutes after, its okay.