How to place captions under the speaker's face in Premiere Pro (beta)

Step by step

1. Create the video background — In step 3 Layout, click the orange Create background button and wait until it turns green and reads Video background.

2. Tick Below the face — In the Layout section, just under Height, tick Below the face (marked beta).

3. Analyze faces — Click Analyze faces and wait for the result line, for example Blocks with a face: 12 of 18.

4. Adjust the gap — Move Gap below the face: 0 = the text touches the bottom of the face, higher = lower down, negative = moves up onto the face.

5. Check every block — Go through the blocks with ‹ › and play them with ▶ before generating: the feature is in beta.

6. Generate — Click ⚡ Generate, or ⚡ Update if the captions are already on your timeline.

What does Below the face do?

On talking-head videos, captions often land on the chin or the mouth. Below the face moves each caption block to just below the speaker’s face:

• for each block, the plugin looks at the face during that block and puts the top of the text just under the lowest point the face reaches;
• if several faces are on screen, the text goes under the lowest one;
• a block with no face on screen keeps the normal Height setting;
• blocks you moved by hand stay where you put them;
• the setting applies to the whole sequence, not to one block: it’s the same in the All blocks and Block N tabs.

Why it matters: doing this by hand means scrubbing every block, checking where the face is and nudging the caption up or down, shot after shot. Here one analysis places every block, and you only check the result.

How do I turn on Below the face?

1. Go to step 3 Layout and create the video background: click the orange Create background button and wait until it turns green (Video background). The analysis reads this background — see Layout and word placement.
2. In the Layout section, just under Height, tick Below the face. A beta badge sits next to it.
3. Two settings appear: the Gap below the face slider and the Analyze faces button.
4. Click Analyze faces. The line under the button shows Analyzing… N%, and the button becomes Stop the analysis in case you want to stop.
5. When it’s done, the line shows Blocks with a face: N of T. — or No face found: every block keeps the normal Height.
6. The preview updates right away: go through the blocks with ‹ › and ▶ to check.

Beta: the feature works, but we’re still refining it. Check the placement of each block before generating.

How do I adjust the gap below the face?

Gap below the face goes from −100 to 100, and is set to 10 by default.

Value	Where the text goes
0	The text touches the bottom of the face.
Higher (up to 100)	Lower down, further from the face.
Negative (down to −100)	Moves up onto the face.

The full range moves the text by up to about 30 % of the frame height either way. When a face is very low in the frame, the text stops at the bottom edge so it never leaves the image — it can then cover the chin. In that case, lower the gap for this sequence, or move that block by hand: a block placed by hand is never moved by the analysis.

When do I need to run the analysis again?

The analysis is a record of where faces are over time, not a result per block: you can split or merge blocks without redoing it. Run it again when you add text over a part of the video that had none during the analysis, and whenever the video itself changes:

What you see	What to do
The video background was recreated after the analysis: click “Run the analysis again” to place the text below the faces again.	You rebuilt the background (after a re-edit, or in another quality). Click Run the analysis again.
Create the video background first (orange “Create background” button).	No background yet for this sequence: create it, then analyze.
The video background is being created: start the analysis as soon as it’s ready.	Wait for the green Video background button.
Video background too large (over 1 GB): click “Background too large” to recreate it lighter, then analyze faces.	Recreate it in Low quality, then analyze.
The analysis did not complete.	Recreate the video background and try again. If it happens again, write to us from ⚙ › ✉ Contact with the technical diagnostic.

If the background only covers part of the sequence (created between in/out points), the blocks outside that part keep the normal Height.

The analysis is saved next to your project, in the Subkick Caption Save folder, with the video background: it’s still there when you reopen the project. Unticking Below the face stops an analysis in progress and puts every block back on the normal Height.

Is my video sent anywhere?

No. Face detection runs on your computer, on the video background already rendered for the preview. Nothing is uploaded, nothing is billed, and no transcription credits are used.

Is it saved in my preset?

No, on purpose. Below the face and Gap below the face are saved with the sequence you’re working on, never in a preset: a preset is a look, while faces depend on the footage. A new sequence starts with the option off. The rest of your layout (Margins, Height, spacing, Global size) does travel with the preset — handy to keep one preset per format.

Which plans include Below the face?

All of them. Below the face is open to every plan: Free, including simple mode, Pro and Max. It doesn’t count as custom placement, so it never triggers the simple-mode limits described in Free, Pro and Max. When the layout is right, send it to the timeline as explained in Generate your captions.

Frequently asked questions

How do I keep captions from covering the speaker's face in Premiere Pro?

With Subkick Caption, open step 3 Layout, create the video background, tick Below the face under Height and click Analyze faces. Each block with a face on screen is placed just below that face; the other blocks keep the normal Height.

Is Below the face included in the free plan of Subkick Caption?

Yes. Below the face is open to every plan, including the Free plan and simple mode. It is marked beta: check the placement of each block before generating.

Does Subkick Caption upload my video to detect faces?

No. Face detection runs on your computer, on the video background that Premiere Pro rendered for the preview. Nothing is uploaded and no transcription credits are used.

Why didn't some blocks move under the face?

Subkick Caption leaves a block at the normal Height when no face was found during enough of that block, when you moved the block by hand, or when the block is outside the part of the sequence covered by the video background.

Does Below the face work on horizontal videos?

Yes. Vertical, square and horizontal footage can be analyzed; on horizontal shots the analysis also looks for smaller faces, such as wide shots or two-person podcasts. When several faces are on screen, the text goes below the lowest one.

What happens if I re-edit my video after the analysis?

Rebuild the video background so it matches your new edit: Subkick Caption then stops using the old analysis and asks you to click Run the analysis again. Splitting or merging blocks doesn't require a new analysis.