Text behind person

Put your big caption words behind the speaker

Turn on Behind person in the editor to place selected caption words behind a detected person. Capflair prepares the cutout before playback, shows its progress and reuses that prepared result when exporting.

Try it free →See plans
Free plan: 2 minutes of transcription a month, watermarked exports, all styles.
Caption sample rendered with Capflair’s export engineEdit sample text and scrub its word highlights. This preview does not transcribe audio.

Add the depth effect in three steps

  1. STEP 1

    Caption the clip

    Upload a video, generate or paste your words, and choose a style with large words or a short caption line.

  2. STEP 2

    Turn on Behind person

    Let the editor prepare the person cutout. Wait for the progress indicator to finish before judging the effect.

  3. STEP 3

    Check the overlap

    Replay the scene, adjust the word placement if needed and export. Check the downloaded video as well as the preview.

A person cutout prepared before the effect plays

The editor analyses the video frames and prepares a person track. It then draws the visible person over caption words marked for the behind layer. The progress indicator lets you see when preparation is complete instead of judging a partially processed preview.

The person frames are processed in your browser for this effect. The caption engine and export reuse the prepared mattes. Audio sent for transcription is a separate part of the caption workflow; local person processing is not a promise that every service operation stays on the device.

Which words go behind the speaker?

In a longer caption, the larger words are candidates for the behind layer. The current engine treats words at least 1.4 times the line’s median size as large. A one- or two-word caption can go behind as a whole. Smaller supporting words remain in front to keep the line readable.

The effect also checks visibility. If nobody is detected, it keeps the words in front. If the person would hide too much of the behind words, the readability guard keeps them in front instead. A word staying visible can be an intentional safeguard rather than a failed cutout.

Choose footage with a clear subject

Start with one visible speaker and a reasonably distinct background. Fine hair, fast movement, motion blur, hands crossing the face and another person entering the shot can make a cutout harder to judge. Replay those moments and decide whether the effect helps the clip.

Place the large word where it crosses the person enough to show depth, while leaving the phrase understandable. Moving a word far behind a face can conceal the point of the caption. Use the per-word move, resize and tilt controls to balance the composition.

Review the exported overlap

The style preview beside this page demonstrates caption lettering; it does not simulate person segmentation. Use your video in the editor to assess the actual depth effect. Turn the feature off if a difficult scene would be clearer with captions in front.

Export and play a short result before committing to a long sequence. Look for unstable edges, a hidden key word and abrupt changes at a cut. Keeping the original project makes it easy to adjust placement or disable the effect for a different creative direction.

Limits, stated plainly

  • Person cutouts can struggle with blur, fine hair, occlusion, similar backgrounds or multiple subjects.
  • Preparation takes time and uses the browser’s resources.
  • Readability safeguards can keep words in front.
  • The free plan includes 2 minutes of transcription a month, all caption styles and watermarked video exports. Online checkout is opening soon; ask support about a manual upgrade.

Questions

Does the feature upload person frames?

The person effect processes the video frames in your browser. Transcription sends extracted audio through its own service workflow.

Why is a word still in front?

Small supporting words, missing person detection or the readability safeguard can keep words in the front layer.

Does it work with Urdu captions?

Yes. The behind layer operates on caption words, including Urdu-script and Latin-letter words.

Does the page sample show the cutout?

No. It shows caption type. Open your video in the editor and enable Behind person to assess the cutout.

Related

Caption your next video in minutes

Upload a clip, check the words, pick a style and export. No card needed to start.

Start free →