March 16 (UPI) -- More than 54 years later, the speech John F. Kennedy was to give on the day he was assassinated is finally being delivered.
The former president was on his way to the Citizen's Council at the Dallas Trade Mart on Nov. 22, 1963, when he was shot and killed. Though it was never orated, Kennedy's speech was recovered and archived.
Now, JFK's words have found new life thanks to voice synthesis technology.
To lend Kennedy's voice to the historic text, researchers with Scottish voice company CereProc stitched together 831 clips, each 0.4 seconds long, mined from 116,777 voice samples. The result is a 2,590-word, 20-minute speech that flows seamlessly.
The process of deconstructing recordings of JFK's speeches and public comments into smaller "phonetic units," and then into the form of a smoothly delivered speech, took eight months.
The project was commissioned by Irish company Rothco, in partnership with The Times newspaper, which published the recording on Friday.
Voice engineers did more than piece words together. They analyzed the former president's speech patterns and replicated what they observed using small bits of sound. Their efforts lend the speech a natural quality.
"We use machine learning and artificial intelligence techniques just to try and figure out how he moves his pitch and the duration of his sounds through his sentences," lead voice engineer Chris Pidcock told Radio Scotland. "That was a big part of the project because we didn't want his intonation to be a boring and standard intonation."