Euryka uses speech-to-speech technology to clone a voice, allowing you to generate audio that mimics an original recorded voice, an uploaded voice, or a voice from our Character library.
Our AI analyses the original voice, extracts its characteristics, and creates a digital clone that retains the essence of the original. It learns nuances like pitch, tone, and cadence, then synthesises the voice to match the original profile while maintaining high realism and authenticity.
Using speech-to-speech to instantly record and generate a cloned voiceover
Click Voiceovers on the sidebar.
Click the voiceover project for which you want to generate a new VO.
Click New Row.
Click in Type, and select STS.
Click the Record icon and speak into your device's mic. Press the Stop icon when finished.
You can record up to 300 seconds of audio in the language of your choice.
Euryka generates high-quality audio output. To achieve the best quality audio, we recommend recording in a professional sound studio, or recording from home using high-quality microphones and equipment, and ensuring the environment is noise-free.
Keep recording until you are satisfied.
The voiceover appears in the Preview column. Click Preview voice to play the VO.
Step 6 completes the creation of an instantly recorded VO. If you are satisfied with the output and do not want to clone it, skip the next steps.
To generate a cloned STS VO from your recording, click in the Character column, scroll down or enter the Voice ID in the Search bar, and then select the voice.
Click the three dots to display the Action menu, and click Generate. This may take a few seconds to a few minutes depending on the length of the recording.
The generated voiceover appears in the Preview column. Click Preview voice to play.
To upload more instant recordings, repeat steps 3-6.
To clone another instant recording, repeat steps 3-9.
Using speech-to-speech to upload pre-recorded audio and generate a cloned voiceover
Click Voiceovers on the sidebar.
Click the voiceover project for which you want to generate a new VO.
Click New Row.
Click in Type, and select STS.
Click the Upload icon, and select the audio file to upload. See Supported file formats.
You can upload up to 300 seconds of audio.
The voiceover appears in the Preview column. Click Preview voice to play the VO.
Step 6 completes the creation of a pre-recorded VO. If you are satisfied with the output and do not want to clone it, skip the next steps.
To generate a cloned STS VO from the pre-recording, click in the Character column, scroll down or enter the Voice ID in the Search bar, and then select the voice.
Click the three dots, and click Generate. This may take a few seconds to a few minutes depending on the length of the recording.
The generated voiceover appears in the Preview column. Click Preview voice to play.
To upload more pre-recordings, repeat steps 3-6.
To clone another pre-recording, repeat steps 3-9.
Note:
Content generation and visibility is based on the current workspace and your roles within the Organisation and Project teams. Your personal workspace displays only the content you have generated, while Organisation workspaces show all content you have generated across projects.
Your roles in the Organisation Team take precedence over your roles in the Project Team. For instance, if you are designated as 'Owner' in the Organisation Team, you can view, edit, and archive project content, regardless of your role in Projects.
If you are removed from a project, you can view but not edit assets you have created.
While archived projects are inaccessible to you, the content remains visible in the Dashboard and modules (Threads, Imaginations, Voiceovers, etc.) where it was created.
For Beta phase users:
As we continue to refine and enhance our platform, you might encounter occasional hiccups during the Beta phase. Should you face any of the following issues - screens not displaying properly, tools delaying output generation, or processes appearing frozen - we recommend you refresh the screen, or exit and re-enter the problematic screen or task. If you are still stuck, log out and log back in.