Rapid Voice Cloning for Speech-to-Speech Voice Morphing

31 March 2023 | 5 min min

We are proud to announce that we are opening up Altered Studio’s Rapid Voice Cloning to all of our customers. With this ground-breaking technology, you can now recreate someone's voice with just a few seconds of clean audio, opening up a world of creative possibilities in the realms of ADR, storytelling, eLearning, and marketing. 

‍ How it works

Rapid Voice Cloning allows you to create a Rapid Custom Voice with only 5 to 15 seconds of audio. To do so, just open the New Custom Voice tab, as shown in the animation below, and follow the four steps:

Step 1 - Voice Information: First of all fill in your new voice information: add a name, a gender, the vocal age and the accent. This will allow you to search the created voice in your library.

Step 2 - Train Voice: In the second step, choose the files that will be used to train the model for the new voice. Use short audio segments so that the total training time is between 4 to 8 seconds. To gain the best results, the audio provided should be from a single speaker, with the tone and projection you want your voice to have. Also, make sure you use clean, non-compressed recordings, without additional audio in the background such as music, noise or other voices. The User Interface will warn you if your the length of your training file is too long or too short.

Once you’ve provided the training audio, and the related confirmations, you can start the training.

Step 3 - Evaluate: After the training is done, you can try out your newly created voice by uploading test files. To avoid overfitting, we suggest that these files are different from those used for training. Further, we limit the length of test files to 10 seconds as the test generation consumes morph quota. 

The generated preview will contain some background music to prevent misuse.

Step4 - Add Voice to Library: If the tests are successful and you’re happy with the sound of your newly trained voice, add the voice to the library and done! You now have your Rapid Voice Clone ready to be used for morphing!

Please note that the performance of Rapid Cloning depends on the quality of the training audio as well as the proximity of the target speaker to Altered’s dataset. Rare accents and speaker characteristics might not be well represented in our datasets and might not produce convincing results. For assured post-production quality voice cloning, Altered offers the possibility to train a Custom Voice. You can get in touch through this link.

‍ Use Cases

We'll explore four compelling use cases that demonstrate the power and potential of Rapid Voice Cloning.

1. Speeding-up ADR: Automated Dialogue Replacement (ADR) has always been a time-consuming and costly process for creatives in many industries from Film and TV, to Advertising, Video Games and more, requiring actors to return to the recording studio to re-record lines or words that were inaudible or not properly recorded during the shoot. With Rapid Voice Cloning, editors can now make a voice clone of the actor and fix many of these issues themselves, without the need for the actor to come back to the studio. This helps streamline production schedules, and frees up actors to pursue new projects.

2. Personalizing Stories for Children: Imagine the joy of a child listening to their favourite bedtime story narrated in the familiar and comforting voice of a parent, grandparent, or other family members. Rapid Voice Cloning makes this possible by allowing users to clone a loved one's voice and use it for storytelling purposes at a much lower cost than traditional voice cloning. 

3. Revolutionising E-learning: eLearning has become an essential tool for education and skill development in today's fast-paced world. Altered Studio already provides a suite of professional voices to create engaging and diverse educational content, tutorials, or quizzes, enhancing the learning experience for users. Now you can use our Rapid Voice Cloning to personalise this even further to your business, for example you can narrate key safety messages or a regular company update in the voice of your CEO without interrupting their schedule.

4. Create new character voices: The Rapid Voice Cloning opens up a world of possibilities for generating new character voices. Try blending different inputs for creating your voice, and then mix up your morph settings for Gender, Age and Pitch to create new character voices for your NPCs or short form voice needs!

The addition of Rapid Voice Cloning to Altered Studio opens up a world of creative possibilities across multiple industries. By simplifying and streamlining voice cloning, this feature allows users to achieve unparalleled levels of personalization, engagement, and efficiency in their audio projects. With responsible use, Rapid Voice Cloning has the potential to revolutionise the way we create and consume audio content.

Disclaimer: Please note that the use of this technology should be made in accordance with law, and not for illegal or unlawful purposes or actions. Further, use of Altered Studio is subject to our Terms & Conditions.