How to Use Talking Avatar in PowerDirector

Q: How long can the audio be?

Between 5 seconds and 3 minutes , both printed on the audio drop zone. If your recording is longer, load it anyway and use Trim in the same panel rather than cutting it elsewhere first.

Q: Do I need a recording at all?

No. The Text to Audio tab takes a typed script of up to 1000 characters and speaks it in a voice you choose, so the whole thing can be made from a portrait and a paragraph. Use Import if the script already exists as a file.

Q: Why does my avatar speak in the wrong language?

Because the voice, not the script, decides that. Select a voice at the bottom of the Text to Audio tab shows the loaded voice with its language beside the name, and one is already selected when you arrive. Click Change and pick one that matches what you typed.

Q: What does it cost?

It varies with what you give it, so read the badge on Generate rather than working from a fixed figure. The cost is shown there before you commit, and the confirmation that appears afterwards states the number of credits that will be deducted.

Give a still portrait a voice, from an audio file you have or a script you type.

Updated Aug. 27, 2026 · 4 min read

In a hurry? Use the Talking Avatar quick guide instead.

Talking Avatar in PowerDirector takes a still portrait and a piece of audio and returns a video of that face speaking it. Two things go in and one comes out, and the audio half can come from either direction: a recording you already have, or a script you type and hand to a synthetic voice. No camera, no presenter, no studio time.

The still

One portrait from the sample gallery.

The video

⚡ Quick Answer: How to Make a Talking Avatar in 3 Steps

Open the GenAI menu, go to Video, and choose Talking Avatar.
Import a portrait on the left, or pick one from Sample.
On the right, import audio or switch to Text to Audio and type a script, then click Generate.

Steps to Make a Talking Avatar

Step 1: Open It and Choose the Face

Open the GenAI menu at the top of the editor, go to Video, and choose Talking Avatar. It is at the bottom of that section.

GenAI, then Video, then Talking Avatar.

The left half of the window is Import Avatar Image. Click or drag a reference image in, or take one from the Sample gallery underneath, which is the quicker way to see what the feature does before you supply a face of your own.

Background, for what sits behind the subject, becomes available as soon as an image is in. Crop / Rotate sits beside it but stays grayed out with a Sample portrait loaded.

Your own portrait, or one from the sample gallery.

Step 2a: Bring Your Own Audio

On the Import Audio tab, drag in a recording. The drop zone states the range it will accept: Maximum duration: 3 min and Minimum duration: 5 sec.

Once it is loaded you get a player, plus Trim and Replace. Use Trim rather than going back to an audio editor, particularly if your recording runs past three minutes.

Five seconds to three minutes, with Trim built in.

Step 2b: Or Type a Script Instead

Switch to Text to Audio and type into Text Script, up to 1000 characters. Import at the top right pulls a script in from a file, which beats pasting a long one.

Select a voice at the bottom shows which voice is loaded, with its language and its delivery, and Change swaps it. Check this before generating: the voice is preselected, and its language will not necessarily be the language you just typed.

Type a script, then check which voice is actually loaded.

Step 3: Generate and Walk Away

Generate shows the cost on the button before you click it. It is not a flat price, so read the badge rather than assuming one. A confirmation then restates the figure that will be deducted.

The useful part is the rest of that notice: you can close this window and continue to explore other features. You will be notified in PowerDirector and we'll send you an email when it's done. Click OK and get on with the edit. There is nothing to sit and watch.

Close it. You get told in the app and by email when it lands.

Haven't got PowerDirector yet? One portrait and one piece of audio is the whole input, and you can type the script instead if you have no recording.

Download PowerDirector Free →

FAQs

1. How long can the audio be?

Between 5 seconds and 3 minutes, both printed on the audio drop zone. If your recording is longer, load it anyway and use Trim in the same panel rather than cutting it elsewhere first.

2. Do I need a recording at all?

No. The Text to Audio tab takes a typed script of up to 1000 characters and speaks it in a voice you choose, so the whole thing can be made from a portrait and a paragraph. Use Import if the script already exists as a file.

3. Why does my avatar speak in the wrong language?

Because the voice, not the script, decides that. Select a voice at the bottom of the Text to Audio tab shows the loaded voice with its language beside the name, and one is already selected when you arrive. Click Change and pick one that matches what you typed.

4. What does it cost?

It varies with what you give it, so read the badge on Generate rather than working from a fixed figure. The cost is shown there before you commit, and the confirmation that appears afterwards states the number of credits that will be deducted.

5. Do I have to wait with the window open?

No, and the confirmation says so: you can close the window and carry on with something else. PowerDirector notifies you in the app when the video is ready, and sends an email as well, so a long script is not a reason to sit and watch a progress bar.