🎚 Vocal Stitch

Prepares a dataset for voice training: upload one or more vocal WAVs — the tool cuts the pauses (where there is no voice) and joins everything into one continuous file. Runs in your browser, files never leave your device.

🎙 The result is a ready dataset for “Train a voice”. Clean, pause-free, 5–10 minutes of voice = a better model with fewer artifacts.

📘 How to build a great dataset — guide →

🎁 The tool is free. What about the full catalog?

811 ready Suno prompts in artist styles. Forever, no subscription.

Open catalog →

🎙

Drop WAV files here or click

wav / mp3 / m4a / flac — multiple allowed

Silence threshold -40 dB

Below this level = “no voice”. Lower (−50) cuts harder, higher (−30) softer.

Min pause to cut 400 ms

Pauses shorter than this are kept (breaths, in-phrase rhythm).

Padding around phrases 120 ms

“Air” around phrases so attacks and tails are not clipped.

🔒 Files never leave your device. Output: mono WAV 48 kHz (matches voice training and Suno). Many large files may use device memory — process in smaller batches if it stalls.

Like the tools? Unlock the full catalog

Ready style prompts for 811 artists · 🧪 Lab (12 tools) · 50 𝄞 monthly. One-time payment, no subscription.

Unlock all · 22,69 € I have a code