Before you begin
Before you embark on your translation journey, you will need a couple of things. First, create an account to take advantage of the free translation minutes and tokens. Then prepare a short test clip with clear speech within the supported formats, sizes, and length limits.
Tip: Start with a 30-60 second clip. Short files finish faster and make it easy to compare results when you are testing different workflows. This will also make your free tokens last you longer 😉
Quick start, four steps:
After you have created your account, the free tokens will be added to your balance.
Now you can start translating your first Video or Audio:
Upload your video or audio by selecting upload or by dragging and dropping your file into the upload area.
Confirm the detected source language or change it when needed.
Select the target language and choose your added features, such as Background Sound Retention and Lip-Sync.
Click Translate and wait for the magic to happen 🌟
How it works, end to end
You upload a file
You choose languages
CHAMELAION transcribes, translates and generates audio + new Lip-Movement if you selected Lip-Sync
If needed, you can review and make quick fixes in the Dubbing Studio.
You export and share
Where things are
Once you are in the file you want to translate, navigation is simple. The page shows the player on the left, with the video in the language you selected from the language list on the right. This side includes your Source language and any Generated Languages, each with Audio and Video rows and a download button. Right below the Video Player, you will find the Edit in Studio button. This takes you to our dubbing studio, where you can make quick fixes and changes to the entire video translation. For more details, please see our help centre collection on the Dubbing Studio. At the very top, you will find the Overview button, which will take you back to the file and project overview. You can jump right into another file/project by clicking the New File button next to it.
Tips for a good first run
Use clear, intelligible speech. The active speaker should talk at a normal pace without mumbling or frequent filler words. Avoid people talking at the same time. Alternate between speakers so the transcript stays clean. For Lip-Sync, record the speaker from the front. A face turned toward the camera produces the most accurate mouth movements. Ensure good video and audio quality. Background sounds are fine, but they should not drown out the speaker. If you use Voice Clone, include at least 30 seconds of clean speech for each voice.
Very long files and many speakers increase processing time. Translate into one target first, review and fix issues, then create additional languages. You can rerun a file in different languages or voices without re-uploading.
For more advice, Tips and Tricks, please see the Best Practice collection.
Tokens and minutes, in short
Tokens represent processing, and minutes reflect the time spent processing media. For examples and plans, visit our website or visit our help centre article on Tokens.