#SameFaceNewVoice — Your Voice, a Song You Would Never Sing

The challenge: generate a cover of a song you would never get up to sing — wrong genre, wrong decade energy, wrong amount of screaming — in your own recorded voice. Post it with #SameFaceNewVoice and the line "I would never sing this. That's still me."

This is not "sound like Lady Gaga." There is no celebrity library. If your joke requires someone else's throat, you are in the wrong challenge and the wrong app. You cannot clone someone else. The punchline is your timbre on a track your friends cannot picture you attempting.

The rules (there are barely any)

1. Own voice only. Three English samples in the studio. You. Not a friend "for the bit," not a pulled clip from their story.
2. A song you would not pick at karaoke. If you already belt Cruel Summer in the car, pick something you would skip. bad guy, Drama, luther, Supernova — contrast is the point. Demo, or an upload you have rights to.
3. Same face. If you show yourself on camera, it should still be you. No disguise that implies a different singer. The mismatch is repertoire, not identity theft.
4. One hashtag. #SameFaceNewVoice so other people's impossible setlists can find yours.
5. Pass it on. Share page → Make your version. Dare them to pick their "never."

Why this works

Everyone has a public musical identity: the Swiftie, the K-pop stan, the "I only like lyrics." Hearing that same speaking voice ride a NewJeans hook or a Kendrick Lamar groove is a double take. You do not need to perform it. You do not need to know how to sing. The model follows the source melody; you supplied the timbre.

It also fails in an honest way if you pick a belt-heavy ballad as your first "never." A speaking clone on Beautiful Things is a hard test. Start with a hook you can hum even if you would never choose it — Hype Boy, Levitating — then go weirder.

How to enter

1. Open the studio and record three short English samples (about 30 seconds total). Quiet room. Speak. Clean samples if last time sounded like a hallway.
2. Pick the song you would never put on your story — demo picker, or upload only if you have rights.
3. Hit Generate. Listen once. If it sounds like you, you win even if it is ugly. Ugly is allowed.
4. Post the share page (or a screen recording of it) with #SameFaceNewVoice. Video export exists; do not wait on it. Send one friend the link with "your never."

Free: 100 credits a month (~10 songs), one lifetime training. Train 10, clip 10. One never per month still counts.

Featured versions

The best versions get featured here and reposted by the VibeSing account — the ones where the comments are "wait that's YOUR speaking voice?" not "which artist model is that." Wrong answers: celebrity names in the caption as if you cloned them.

Make your version →

AI voice covers on VibeSing use only your own recorded voice, with your consent, and exports carry an AI-cover disclosure. Confirm you have rights to any audio you upload.