At its heart, Selector is a database program. The director of programming or music at the radio station inputs information for each song programmed, along with “rules” about how that song fits into the format of the radio station.
This can be as complex or simple as it needs to be. As you can see, there are a lot of options. The most important one is the category (“Cat”). This is the song’s basic classification, the role it plays in the station’s hierarchy.
Other rules could include whether the performance is by a group, male vocalist, or female vocalist, or if it’s an instrumental.
Notice how the rules are set, so that there can’t be more than one duet in a row or two female vocalists in a row, and there must be at least four songs between every instrumental.
Then the music director or program director can set the rules for the sound of each song.
In this case, the station could only play four ska tracks over the course of an hour with at least three tracks between them; only two African language tracks with five songs between them, and no more than two cover songs with a minimum of four tracks in between. There are a variety of other rules under the sound code as well – the system can handle as many as 52, twice through the alphabet.
When the PD or MD finishes entering the information – putting new records into the playlist and taking records out or recoding them – he or she presses a couple of keys and Selector goes through the database of songs and rules, programming the station.
Once the program completes this task, it generates a list of songs in the order that the automated system will play them:
On a station with “on-air talent,” that person gets a copy of the list. Unlike during the heyday of progressive rock station some 25 years ago, rarely does the talent actually touch any recorded media on the air. It has all been preprogrammed. The term “disc jockey” has become a nostalgic anachronism in contemporary radio. The on-air talent has little or no say over what goes over the air anymore.
So how do requests get on the air? Well, as a person listening to the station, the requester probably wants to hear a song that’s on the playlist anyway. The talent knows when a song is going to come up, and announces the request when it does. The talent might even record the person making the request, to make it seem like it’s happening at that moment. But the days of “You say it, I’ll play it” are long gone.