OpenAI Voice Creation Step-by-Step Walkthrough
*This feature is currently available for Enterprise plans only*
HeyGen Voice is available for customization only to users in select regions (the United States, excluding Washington, Texas, and Illinois; European countries; or Canada). Before production commences, we need to sign the Voice Actor Services Agreement to ensure that we have the rights to produce the voice clone.
Building a voice engine with HeyGen is designed to be straightforward. To capture the full range of your vocal expressions, we've prepared scripts representing five distinct emotions: Excited, Friendly, Serious, Soothing, and Broadcaster. Each script is approximately 30 seconds long. You'll record them separately, creating a file for each emotion. It's crucial to ensure that the recording volume and environment are consistent across all recordings to maintain uniform quality and performance of the voice engine.
Step 1: Sign the Voice Clone Consent
Before beginning the voice recording process, you must sign the Voice Clone Consent form. This ensures that you give permission for your voice to be used in the creation of the voice clone and that you understand the terms and conditions associated with the process.
Step 2: Record Your Voice Sample
Equipment & Environment Setup
Microphone: Invest in a high-quality, noise-canceling microphone. This is vital for capturing your voice clearly across different emotional tones.
Recording Space: Choose a quiet, controlled environment for all your recordings. Consistency in your recording space is key to ensuring uniform sound quality.
Mic Position and Settings: Position the microphone consistently about 6-12 inches from your mouth for all recordings. Adjust input levels to ensure your recordings have a consistent volume, avoiding clipping for louder tones and ensuring clarity for softer tones.
Script Recording Guidelines
Preparation:
Familiarize yourself with each script, understanding the emotion behind it.
Perform vocal warm-ups, especially focusing on exercises that enhance emotional expression.
Recording Approach for Each Emotion:
Excited: Convey enthusiasm and energy. Your voice should be lively and engaging.
Friendly: Aim for a warm, welcoming tone. Imagine speaking to a friend.
Serious: Maintain a steady, authoritative tone. Keep your voice even and controlled.
Soothing: Use a calm, gentle voice. Aim for a comforting and reassuring tone.
Broadcaster: Adopt a clear, confident tone. Imagine youโre speaking to a large audience.
Recording Sessions:
Record each script in a single session to maintain vocal consistency. Take breaks between sessions if needed, especially when switching between contrasting emotions like "Excited" to "Soothing".
Make sure to have one second of silence ๐ at the beginning and end of the footage.
Record multiple takes for each script to ensure you capture the best performance.
Post-Recording and Submission
Review and Select: Listen to all takes and select the best one for each emotion. Ensure the tone is consistent and true to the script's intended emotion.
Edit with Care: Perform light editing as needed to remove any unwanted sounds or long pauses. Strive to maintain the natural dynamics of your voice.
Finalize Formats: Export each recording as a separate file in a lossless audio format (WAV or FLAC). Ensure the volume levels are balanced across all files.
File Naming: Name each file clearly according to the emotion it represents. For instance, use naming conventions like "Excited.wav" for the excited tone, "Friendly.wav" for the friendly tone, etc. This approach ensures that each file is easily identifiable and corresponds directly with the intended emotional tone of the script.
Uploading to Google Drive:
Prepare for Upload: Once you have your files named correctly and ready, open Google Drive in your web browser. If you don't have an account, you will need to create one.
Create a New Folder: In Google Drive, create a new folder for these recordings to keep them organized. You might name it something like โHeyGen Voice Engine Recordings - [Your Name]โ.
Upload Files: Drag and drop your files into this folder or use the "New" button to select and upload your files manually. Also, please add your Voice Clone Consent into the folder.
Share the Folder: Once all files are uploaded, right-click the folder and select "Share". You can then enter the email provided by HeyGen or generate a shareable link. Ensure the link is set to โAnyone with the link can viewโ to allow HeyGen access to the files.
Send the Link: Send the link to your HeyGen main contact person.
Consistency Tips
To achieve consistent volume and sound quality, set your recording software to the same input level for each session.
Record at the same time of day, if possible, to minimize variations in your voice due to fatigue or environmental factors.
Use the same microphone setup and positioning for each recording to ensure uniform audio quality.
By following these detailed instructions, you'll provide high-quality, emotionally diverse recordings that will enable the HeyGen voice engine to capture the rich nuances of your voice, ensuring a versatile and realistic voice clone.
Scripts
Friendly:
๐ Hi there, friend! It's so good to talk with you today. I always enjoy our chats - you just have a way of making me smile and brightening my day. Want to hear a funny story from my day? I was walking down the street earlier, lost in thought, when suddenly I tripped over an uneven piece of sidewalk. As I went tumbling forward, my bag went flying, and all my belongings scattered everywhere! I must have looked like quite a sight, sprawled out on the ground surrounded by a mess of dropped items. Luckily, a kind passerby rushed over to help me up and gather my things. We both had a good laugh about my clumsiness! ๐
Serious:
๐ Listen carefully, because what I'm about to say is of utmost importance. We face a critical situation that demands our full attention and decisive action. No more delays, no more excuses. Time is rapidly running out to address the escalating climate crisis before lasting, catastrophic damage is done to our planet. The scientific data makes it plain - atmospheric greenhouse gas levels are higher than they've been in millions of years due to human activities like burning fossil fuels. If we don't take bold, immediate steps to transition away from carbon-emitting energy sources and enact other mitigation efforts, rising temperatures will cause devastating impacts worldwide.๐
Soothing:
๐ Take a deep breath and allow the tension to leave your body. You are safe and cared for. Imagine yourself in a tranquil meadow on a warm summer day, gently swaying wildflowers all around you. The sweet fragrance of blooms carried on a calm breeze soothes your soul. Look up at the brilliant blue sky, with just a few wispy white clouds lazily drifting by. Feel the sun's rays providing a gentle, comforting warmth across your skin. The only sounds are the occasional buzz of a bee and the melodic trill of birds singing their joyful songs. Allow yourself to fully relax and release any worries or stress in this peaceful haven.๐
Excited:
๐ Oh my gosh, I can hardly contain my excitement! You'll never believe the amazing news I just received! This is truly a dream come true for me that I've been hoping and wishing for so desperately for years. Ever since I was a little kid, I've had this crazy ambitious goal of traveling to outer space. I used to gaze up at the stars every night and dream of the day I could finally experience the wonders of the cosmos firsthand instead of just reading about them. Well, I was just selected to join the crew for an upcoming commercial space tourism flight! Can you even imagine how incredible that's going to be?๐
Broadcaster:
๐ Attention now to our breaking news alert! We're interrupting your regularly scheduled programming to bring you this critical update. Eyewitnesses are reporting a multi-vehicle collision on the interstate involving several semi-trucks and passenger vehicles. Preliminary details are still emerging, but it appears a tractor-trailer lost control and jackknifed, triggering the chain-reaction pileup. Footage from traffic cameras shows thick, black smoke billowing across the highway as fire crews rush to the chaotic scene. Authorities are urging all motorists to avoid the area as multiple lanes in both directions will likely remain blocked for an extended period while emergency responders triage the situation and clear the wreckage.๐