Creating voiceovers with Automation – CaptionHub

📢

Requirements: Enterprise subscription, Voiceover via Automation enabled for your team

Automation can now create voiceovers for you. Once voiceover is switched on in an Automation template, CaptionHub creates a voiceover for each caption set the automation produces, original or translated, using voices cloned from your video's original audio. It can then publish the voiceover to your video platform, either straight away or once someone has approved it.

⚠️

You may only clone your own voice or a voice you have the rights to use.

Before you start

Voiceover via Automation needs to be enabled for your team. If you can't see the options below, contact your Customer Success Manager.

Voice cloning is powered by ElevenLabs. Make sure the Allow voice cloning box is ticked in the connector. See Voice cloning for more on how clones are made.

Voiceovers use credits, so check your team has enough for the volume of media you expect to process.

If you haven't set up an Automation template before, start with Setting up Automation templates.

Turn on voiceovers in an Automation template

Go to Automation and open the template you want to use, or create a new one.

On the Automation tab, switch on Create voiceovers.

Choose what should happen once the voiceovers are created, and when they should be published: as soon as they're created, or once they've been approved.

Save the template. A Voiceover tab now appears alongside the Original and Translations tabs.

Set up the Voiceover tab

Languages - Tick the languages you want voiced. You can include the original language as well as any translations the template creates.

Provider - If your team has more than one voiceover provider, choose the one to use.

Track - Choose which caption track to voice.

Background audio - Choose whether to keep the original background audio under the voiceover, and set its volume. Save your changes. If a setting can't be carried out by the automation, CaptionHub will tell you when you save or activate the template, so you can fix it before any media comes in.

How voiceovers are created

Each voiceover is created as soon as its caption set is ready, so you don't have to wait for every language to finish. CaptionHub clones the speakers from the original audio and uses those voices for every language you've selected. There's no minimum amount of speech needed in Automation, so every caption set will be voiced. Where a video has very little speech, the cloned voice may be less accurate. We recommend checking these voiceovers in the voiceover editor before they're published. A voiceover won't be created if:

your team doesn't have enough credits

the selected track has no captions with text

the caption set already has a voiceover

💡

Tip: If you'd like to review voiceovers before they go live, set the publish trigger to publish on approval. You can then make any adjustments in the voiceover editor and approve when you're happy.

Publishing voiceovers

Voiceovers are published to the publish target chosen on the template, following its publish trigger. Only platforms that accept audio tracks are available as publish targets. For Brightcove, audio track publishing needs to be turned on for your team before voiceovers can be published.

Viewing voiceover steps in Node view

When voiceover is switched on, the voiceover steps appear in Node view alongside the rest of your workflow, so you can see where voiceovers are created and published.

If a voiceover fails

If a voiceover can't be created, it appears in the automation diagnostics with the option to retry it. Fix the cause, for example by topping up credits, then retry the voiceover from there.