Introduction
Creating an AI cover is one of the most exciting ways to use modern AI technology. Whether you want to hear yourself singing a famous pop hit or see how a legendary voice sounds on a new track, OG Voice makes it simple.
Step 1: Prepare Your Voice Model
Before you can create a cover, the AI needs to know what the target voice sounds like.
- Navigate to the "Voices" tab.
- Click "New Profile".
- Upload 3-5 minutes of clean, isolated vocal recordings of the voice you want to clone.
- Click "Train Model".
Step 2: Choose Your Source Song
The "source" is the song you want to transform.
- Find a high-quality version of the song.
- Ideally, use a version where the vocals are clear.
- If you have the "Acapella" and "Instrumental" versions separately, that's even better!
Step 3: Use the Generate Form
Now head over to the "Generate" tab:
- Select Voice: Choose the profile you trained in Step 1.
- Upload Song: Drag and drop your source audio.
- Configure Settings:
- Pitch: If you're switching between male and female voices, you might need an octave jump (±12).
- Voice Strength: 0.75 is a great starting point for most voices.
- Add Back Vocals: Turn this on if you want to keep the original backing harmonies.
Step 4: Generate and Refine
Click "Create Cover". The process usually takes 2-3 minutes.
- Once done, listen to the result in the "Library" or the result panel.
- If it sounds too high or low, try adjusting the Pitch and generate again.
- If the AI "struggles" with certain parts, check if your training data was clean enough.
Conclusion
Your first AI cover is just the beginning. As you experiment with different voices and songs, you'll learn the nuances of pitch shifting and vocal dynamics. Have fun creating!