Earlier this year, Dave Grohl, frontman of the Foo Fighters, made a lot of headlines by suggesting that Taylor Swift doesn’t perform her songs live. It seems he may have been onto something after all.
While Taylor Swift has shattered records with her Eras Tour, which is set to wrap up next month, a vocal analyst has presented some strong evidence proving that she is not singing her songs live.
In a YouTube video shared this week with his 434,000 subscribers, well-known vocal analyst Fil Henley demonstrated through objective audio analysis that Taylor Swift lip-syncs —or mimes — during her Eras Tour performances.
The proof, Henley suggests, is that the vocal performance is the exact same at every single show throughout the tour.
“You can hear that across all of these shows, [her vocals] are perfectly in sync,” he said. “Her vocal is starting at the exact same time, finishing at the exact same time.”
As Henley notes, this is physically impossible for a human to achieve.
“If we get vocal lines that match each other… It can’t happen,” he said. “Even with a clear vocal of a singer, doing one performance and even just singing a few notes and singing a few notes again, those vocal lines will not overlap, in terms of the detail – your hundredths of a second of your vocal cord vibrations.”
“It’s impossible to sing over your vocal lines that you’ve just done – it very much is like a fingerprint,” he concluded.
He also observed that even the “plosives” — the bursts of air hitting the microphone — are identical from performance to performance.
Using pitch detection software, Henley compared the pitches from two of Swift’s performances. The software clearly showed visually that the pitch and timing were identical from one show to the next, which he argued would be physically impossible for a human singer.
“This shouldn’t overlap at all,” he said. “It shouldn’t be anywhere close. But then when we line up the pitch, we see that it is a perfect replication.”
“That is something that is simply impossible to do with human vocal cords – to sing to that detail,” he continued. “We’re looking at hundredths of a second here and [there is] a pitch variation of zero cents.”
“So, this is the same audio file,” he concluded.
He performed the same analysis on two additional shows and reached the same conclusion.
Henley emphasized that this evidence becomes even more compelling when you consider the varying ambient crowd noise present in each performance, which should make the audio appear different. However, the pitch detection software still shows the same exact audio file being played.
“And just to clarify, when I say that it’s the same audio, I mean the same audio file,” Henley explained. “It’s the same audio that’s been recorded in the studio to sound live – because it’s got the plosives left on there – but it’s that same audio that’s being mimed to. It is one vocal that’s been pre-recorded before all of these performances that Taylor is then miming to.”
Additionally, Henley argues that the pre-recorded track isn’t even exactly “live.” Instead, it has been heavily pitch-corrected to the point that he describes it as “borderline autotuned.”
“Something that I will say about the live performance and this vocal is that it is heavily edited from a pitch perspective,” he said. “I mean, it is borderline autotuned – it might be autotuned – because it is so stuck on these lines.”
“This is a pre-recorded vocal that has been heavily pitch-corrected, borderline autotuned,” he concluded.
And Henley is willing to stand by his conclusion.
“That is fact,” he said of his findings. “This is objective data that we are looking at. It’s undeniable.”
It’s certainly some compelling evidence.
