The Voice Studio allows you to transform the timbre of an original vocal recording to mirror the style of one of our AI Voice Models, offering a flexible way to customize vocals just like you would with other instruments.
Consider yourself free to utilize our royalty-free voice models in various applications such as music production, advertising, social media, live-streaming, live performances, and more. However, please note that the resale of the voice model itself, any portion of its generated output in isolation, or its provision for download is strictly prohibited. For a detailed insight into our policies, we encourage you to review our Terms and Conditions.
Voice Studio gives you two ways to upload your files: on the main page or directly on a selected voice. Now you wonder, what's the difference?
Uploading your file on the main page:
When you upload your vocal file in the general upload, our AI will identify and suggest the most suitable voices for your project. You'll see a list ranked from most to least likely to work with your file. Please note that you will only get this voice if you upload a file using a specific voice. You have to use the general upload at the top right of the Voice Studio main page.
Uploading your file on a specific voice:
In this upload option, you will not receive suggestions from our AI; your upload will happen directly in the selected voice. You can check a demo of the voice, optimal range, and the equipment used for recording.
We constantly provide upgrades across time. You will see improved clarity, resolution, bitrate, and microphone options. During studio recordings, we used resolutions equal to or higher than 48kHz and 96kHz. We also use state-of-the-art preamps and microphones.
Extra tips:
-
Does the microphone I record my input audio make a difference?
Our models are resilient enough to imprint the characteristics of the voice model. Even recordings that use mobile phone microphones might generate surprising results with high resolution. -
Are there any best practices?
We recommend using dry vocals and avoiding adding tracks with reverb, delays, leaks, or instruments. Tracks with reverb might still present great results due to our voice model resilience. -
Does it support any language?
We are constantly improving the models to support a diverse set of languages. Although we strive to make our models work in any language, some models may perform better in English, Spanish, Portuguese, and Italian.