The Desk appreciates the support of readers who purchase products or services through links on our website. Learn more...

AI software allows radio stations to create fake listener calls

The software, released by Benztown this week, allows radio stations to create realistic "listener audio" using artificial intelligence tools.

The software, released by Benztown this week, allows radio stations to create realistic "listener audio" using artificial intelligence tools.

(Stock image)
(Stock image)

The next time you listen to a radio call-in show, “Dave on Line 4” might actually be a robot.

This week, audio imaging provider Benztown debuted a new product that allows radio stations to create so-called “listener audio” for their promotional and programming needs.

The tool, called the “AI Listener Voice Generator,” allows anyone at the station to input 1,000 characters of text to output as much as 50 seconds of speech that can be inserted into commercial spots or live radio programming.

The voice generator includes a multitude of pitches and tones, including female and male voices that range from “normal” to “hyped.” Radio broadcasters can also switch between “listener” and “caller” audio, which changes the output to match audio recorded on a conventional microphone or telephone.

The end result is an audio file that radio stations can use to simulate actual listeners or callers. It comes at a time when radio broadcasters are facing challenges in collecting listener audio from the public, according to Benztown.

“We are excited to provide our library users with the most compelling listener material possible — the listener audio they actually need for every specific piece of production,” Benztown CEO Andreas Sannemann said in a statement emailed to The Desk this week. “Benztown’s AI Listener Voice Generator is another giant step forward in creating great-sounding radio with next-generation technology through AI-generated listeners on demand.”

In addition to text-to-speech, the software also allows radio stations to import audio recorded from a listener or caller, then modify it based on certain parameters, known as “speech-to-speech.” As with the text-to-speech function, the speech-to-speech perk can also be used in radio promos or on-air to simulate listener calls or man-on-the-street interviews.

“The text-to-speech and speech-to-speech capabilities of this new feature make an audio producer’s job easier and more-efficient than ever, and it sounds incredible!” Sannemann exclaimed.

The audio imaging software allows radio stations to create listener spots and calls using artificial intelligence. (Inlay photo courtesy Benztown, Graphic by The Desk)
The audio imaging software allows radio stations to create listener spots and calls using artificial intelligence. (Inlay photo courtesy Benztown, Graphic by The Desk)

Benztown said the AI tool is available only for broadcast radio stations, though it said online stations who are also interested in using the software should contact a representative.

The new tool comes at a time when commercial radio broadcasters are increasingly looking to artificial intelligence to power different aspects of their stations.

Last February, technology firm Futuri Media unveiled a new machine learning tool called RadioGPT that simulates the on-air presentation of a human presenter. RadioGPT is fed information from a variety of sources, including social media posts and newspaper stories, to produce community bulletins, traffic reports, weather forecasts and other information on-air.

While it isn’t clear how RadioGPT verifies the information, it scrapes from the Internet, the tool was considered reliable enough for radio broadcaster Alpha Media to begin testing it in Portland last July, putting it to use on its Top 40 station KBFF (95.5 FM), where it simulates the presentation style of an actual DJ named Ashley Elzinga.

Earlier this month, Cumulus Media said it was working with radio software company Super Hi-Fi to launch two AI-powered stations in Nashville. The stations, called “The Hill” and “Nashville Songwriter Radio,” are broadcasting on HD Radio sub-carriers of WGFX (104.5 FM, “The Zone”).

Get stories like these in your inbox, plus free breaking news alerts on business and policy matters involving media and tech.

Get stories like these in your inbox, plus free breaking news alerts on business and policy matters involving media and tech.

Photo of author

About the Author:

Matthew Keys

Matthew Keys is a nationally-recognized, award-winning journalist who has covered the business of media, technology, radio and television for more than 11 years. He is the publisher of The Desk and contributes to Know Techie, Digital Content Next and StreamTV Insider. He previously worked for Thomson Reuters, the Walt Disney Company, McNaughton Newspapers and Tribune Broadcasting.
Home » News » Industries » Radio » AI software allows radio stations to create fake listener calls