CaptionAgent Step by Step

Quick Start

Go from video to finished clip with captions in three steps. Everything runs right on your device – no account and no upload.

Step by step

1. On the home screen, tap “Record” to record a video or audio directly in the app, or “Import” to open a video from Photos or Files.
2. CaptionAgent creates the captions automatically. The first time, Apple downloads the language pack for your language once.
3. Check the text, choose a style and tap “Export”. Your clip is done – as a video with captions in the picture or as a subtitle file.

Tips

• With “View Example” you can try out all the tools on a short sample video, without any footage of your own.
• Several files at once? They’re processed one after another, and you can keep working in the meantime.
• In the ⋯ menu at the top right you’ll find Help, “How it works”, Feedback and Settings.

If something doesn’t work

↑ Back to top

Recording Well

Good sound matters more than a perfect picture – it also makes the captions more accurate. A few simple steps make every recording noticeably better.

Step by step

1. Tap “Record” and choose “Video”, “Audio” or “Interview”.
2. Turn on the camera and microphone. The frame in the preview shows where the captions and the social media controls will appear later – keep that area clear.
3. Watch the level meter while you speak and start recording, with a 3-second countdown if you like.

Tips

• Light: Face a window or a lamp. Light from behind makes your face look dark.
• Sound: Record in a quiet room. Rugs, curtains and furniture reduce echo. A clip-on microphone helps a lot.
• Distance: about an arm’s length, with your eyes in the upper third. For Reels, hold the iPhone upright.
• With “Mirror” you see yourself in the preview as in a mirror; the finished video stays the right way round.

If something doesn’t work

↑ Back to top

Teleprompter

Read your script straight from the screen while you record. The prompter follows your voice: word by word with on-device speech recognition, and in every other language at the pace of your voice.

Step by step

1. Open video recording and tap the text icon at the top right.
2. Write your script, import a text file or tap “Sample text” to try out the feature.
3. Choose the language you’ll speak and turn on “Use prompter”. Back in the camera, the text scrolls along as you speak.

Tips

• The prompter only appears in the preview, never in the recorded video.
• You change size and position directly on the prompter; drag it close to the camera so your gaze looks natural.
• Tap a line to jump there. Pause stops the text from moving on.
• Under “Follow”, choose “Word follow” (speech recognition, word by word) or “Voice-paced”: the text then moves at the set pace while you speak and holds in pauses. Two sliders set the pace and the sensitivity.

If something doesn’t work

↑ Back to top

Interviews & USB MicrophonesiPhone and iPad only

Two cameras, two people and, if needed, two separate microphone channels. CaptionAgent automatically cuts to whoever is speaking.

Step by step

1. On a supported iPhone or iPad: Record → Interview. Hold the device between the two of you: rear camera toward the guest, front camera toward you.
2. Use the zoom levels to frame the guest and start recording. Both cameras record at the same time.
3. After recording, CaptionAgent creates an automatic edit based on lip movement and sound. In the Cut section, you can correct any camera switch with a tap.
4. Under Style, you can give each person their own caption color or even a completely separate style.

Tips

• Connect a USB receiver before opening interview recording and set it to stereo or separate L/R output. Default: Guest = left, You = right; use the double arrows to swap the assignment.
• Test the microphones one at a time before you start: only the matching level should rise clearly.
• When cutting, the sound follows the visible camera – even after manual changes and when exporting.

If something doesn’t work

↑ Back to top

Checking the Text

Speech recognition is good, but not flawless. A quick look over the text is worth it – especially for names and technical terms.

Step by step

1. Open a project. Below the video, you’ll see the transcript in blocks.
2. Tap into a block to correct the text. The captions update immediately.
3. Uncertain spots are highlighted. “Next Uncertain Spot” jumps straight there – on the Mac with ⌘G.

Tips

• Select a block’s time to jump to that spot and listen to it.
• With “Emphasize Word” you can highlight individual words in the caption.
• Use search to quickly find names and technical terms and replace them everywhere.

If something doesn’t work

↑ Back to top

Cutting

Trim the start and end, or remove individual words, pauses and filler words – right in the text. Your original always stays unchanged.

Step by step

1. In the editor, open the “Cut” tool.
2. Drag the handles on the filmstrip to set the start and end.
3. Remove individual words or whole blocks with a tap. Video, sound and captions are shortened together.

Tips

• “Find filler words” marks “uh”, “um” and the like; “Shorten pauses” tightens long pauses. You decide what goes.
• On devices with Apple Intelligence, “AI suggestions” propose further cuts.
• Every cut can be reversed – with “Undo” or in the ⋯ menu of the Cut section with “Reset all cuts”.

If something doesn’t work

↑ Back to top

Caption Style

The style determines the font, colors, background and animation of your captions. More than 60 designs are ready to use – and each one can be customized.

Step by step

1. In the editor, open “Style” and choose a design from the catalog. The preview shows it right away with your text.
2. Adjust size, position, colors, outline and background. You can also move the captions directly in the preview image.
3. If you especially like a style, save it under “My styles” or set it as the default for new projects.

Tips

• With brand colors, you use your brand’s colors for captions and scenes.
• For interviews, choose “Guest” or “You” at the top: the default is a style with its own color for each person, and if you like, each person can get a completely separate design.
• Turn off motion if you want the captions to look calmer.

If something doesn’t work

↑ Back to top

Scenes & Cutout

Scenes place you in front of a designed background: full screen, split screen or picture-in-picture – ideal for vertical clips.

Step by step

1. In the editor, open “Scene” and choose a template and the aspect ratio, for example vertical 9:16.
2. Under Cutout, choose “Cut out person” or – in front of a green or blue wall – “Remove green background” or “Remove blue background”.
3. Add a headline if you like, check the preview and tap “Apply”.

Tips

• With a green screen, even curls and fine hair are cut out cleanly; without a backdrop the edge is good, but the finest flyaway hairs may be lost.
• “More precise cutout” steadies the edge in calm shots, but the export then takes considerably longer.
• Effects like shadows or a cardboard cutout give the cut-out person more depth.

If something doesn’t work

↑ Back to top

Enhance Audio

“Enhance Audio” reduces noise and echo in the original sound – right on your device.

Step by step

1. In the editor, open the “Sound” tool.
2. Set the strength and start processing.
3. Compare “Original” and “Enhanced”. The version you choose is used for the preview and the export.

Tips

• Set too strong, a voice quickly sounds artificial. A medium strength is often enough.
• With long videos, processing takes a while; you can cancel it at any time.

↑ Back to top

Export & Share

Export your video with captions in the picture, or just the captions as a file for YouTube and editing software.

Step by step

1. In the editor, tap “Export” and choose “Video with captions” or “Subtitle file”.
2. Choose the quality and frame rate. Full HD is enough for Reels and Shorts.
3. After exporting, save the video to Photos, as a file, or share it directly.

Tips

• 4K is only worth it if your original was recorded in 4K.
• You can use SRT and VTT files on YouTube, for example, or in Final Cut and DaVinci Resolve.
• From the ⋯ menu on the home screen, you can export subtitle files from several projects at once (Pro).

If something doesn’t work

↑ Back to top