De Juan Loquendo | Voz

Founded in 2001 as a spin-off from the prestigious Centro Studi e Laboratori Telecomunicazioni (CSELT) in Turin, Italy, Loquendo was a cutting-edge text-to-speech (TTS) engine. Unlike the robotic voices of the 1980s, Loquendo used concatenative synthesis—recording hundreds of thousands of phonemes (the smallest units of sound) from a real human voice and reassembling them to form any word or sentence.

Do you have a favorite memory of the Voz de Juan Loquendo? Share your story in the comments below (or, if you’re feeling nostalgic, type it into a TTS engine and let Juan read it back to you). voz de juan loquendo

In the early 2000s, radio stations faced a problem. They needed to produce imaging (promos, IDs, time checks) quickly, cheaply, and consistently. Hiring a human voice actor for every 5-second bumper was expensive and slow. Founded in 2001 as a spin-off from the

Radio producers discovered that by typing a script into Loquendo and selecting the "Juan" voice, they could generate a professional-sounding drop in seconds. It was a revolution. Suddenly, small community radio stations in rural Mexico could sound as polished as a major network in Madrid. Share your story in the comments below (or,

If you grew up listening to radio in the Spanish-speaking world during the 1990s and 2000s, you have heard this voice. It is deep, warm, slightly theatrical, and absolutely unmistakable. It is the voice that introduced songs, announced contest winners, and gave life to thousands of radio jingles across Latin America and Spain.

But "Juan" was special. The specific vocal model—the —had a unique timbre. It was crisp, authoritative, and possessed a natural cadence that felt almost human. This wasn't a glitchy robot; this was a virtual radio announcer. Part 2: From Italian Lab to Latin American Airwaves How did an Italian text-to-speech engine become the king of Spanish radio?