Speaker review is where you check who speaks each line of your video and choose a voice for each speaker. Dubbing does not start generating speech until you confirm. This article shows how to do it step by step.
Open speaker review
Open the dubbing project. Under Voice Generation Jobs, select Speaker Review on the job. The top bar has Back to project and the file name with its language.
[[IMAGE 1 - the voice generation job with the Speaker Review button, from kb-screenshots/dubbing. Delete this line after you add the image.]]
What you see
- Detected speakers on the left, with views for All dialogue and Unassigned, and Add speaker manually for a speaker Feltiv missed. A manual speaker needs a name and Male or Female.
- A Video preview on the right. Select any dialogue line to jump the video to it and check who is speaking.
- Speakers ready and Segments assigned progress bars.
- A search box, Search dialogue, speaker or timecode....
[[IMAGE 2 - the Speaker Review screen. Capture it from a project at this stage, using a video you own. Delete this line afterwards.]]
Set up each speaker
- Enter a name under Enter a name. Duplicate names are not accepted.
- Choose the voice. Feltiv compares each speaker with the voice library and suggests the closest match. You can see the voice that was applied automatically and the top 3 matches. Accept one of them, or choose your own.
- To choose your own, use the Library Voice tab, which shows the Engine, Gender and Voice, and select Browse voice library.
- Assign lines to the speaker with each line's Assign to... selector. If you remove a speaker, their lines go back to Unassigned.
Browse the voice library
In Browse Voice Library, search with Search Voices, filter by Language, Gender, Engine and Style, and select the play button to hear a preview. Mark your favourites, then select Use this voice. Your changes save automatically.
Confirm
Select Confirm & proceed. The button stays disabled while there are no speakers, a speaker has no voice, lines are unassigned, or a save is pending. The reason shows beside the button, for example (2 segments unassigned).
When it works, you see Voices confirmed - dubbing voice generation started. When the speech has been rendered, open the job with Review. See the article on reviewing and exporting your dub.
Speakers you can merge later
If the same person was detected as two speakers, you can merge them later on the dubbing review page. See the article on reviewing and exporting your dub.
Need help?
If Confirm & proceed stays disabled, read the reason beside it. If it is still unclear, write to support@feltiv.com with the project name.