Talking Head Videos: How to Film and Edit One That Holds Attention

What is a talking head video?

A talking head video shows a person speaking on camera, usually framed from the chest or shoulders up, with their voice carrying the content. The speaker may look straight into the lens or answer someone off camera. Most explainers, opinions, tutorials, announcements and interview clips on Shorts, Reels, TikTok, YouTube and LinkedIn are talking head videos.

The format is simple on purpose. One person and one clear voice can explain an idea quickly, because nothing on screen competes with what is being said. That simplicity is also the risk. With no action to watch, every pause, mumble and dead second is visible. The edit decides whether the video feels direct or feels long.

Five talking head video formats

Pick the format by what the viewer needs to see. If the value lives in your reasoning, experience or opinion, keep your face on screen. If it lives on a screen or in an object, show that, and use your face as the anchor between demonstrations.

You can mix formats inside one video. A direct-to-camera opening that hands over to a screen recording, then comes back to your face for the takeaway, suits most tutorials. The course study above saw the same pattern: cutting between a speaker and slides held students longer than slides alone.

Where talking head videos work

The same recording can serve several places, but each platform sets its own frame and length. Check the current specs before you film, because they change.

• YouTube Shorts. Shorts are square or vertical videos up to 3 minutes long, according to YouTube’s upload page. YouTube’s Shorts tips say you need no special equipment: you can record, edit and upload from your phone.
• Instagram Reels. Instagram’s creator FAQ advises making the first 3 seconds engaging so people don’t move on and uploading at the highest resolution possible. It recommends reels of 3 minutes or less to people who don’t follow you.
• TikTok. TikTok’s creative best practices recommend vertical 9:16 video, at least 720p, sound, and keeping content inside the interface safe zone. They are written for ads, so treat them as layout advice rather than an organic ranking rule.
• YouTube long-form. A talking head often carries a longer video, with screen recordings and B-roll around it. YouTube’s retention report shows what share of viewers are still watching after the first 30 seconds, and suggests reworking that opening if many leave.
• LinkedIn. LinkedIn’s video specifications for Pages accept shapes from 1:2.4 to 2.4:1, so the vertical file you made for Reels fits.
• Courses and onboarding. Short lessons that cut between your face and the material are the setup the edX study looked at.

For the three short-form apps, export a 9:16 vertical file. 1080 × 1920 pixels is the usual working size and the resolution TikTok’s video ad specifications recommend. Meta’s Reels ad specs currently list 1440 × 2560, the same shape at a higher resolution.

Plan one idea and a strong first line

Write the point as one sentence

Before you film, write what the video says in one sentence. “Move your light beside the camera, not behind you” is a video. “Lighting, sound and framing tips” is three videos. One idea makes the video easier to follow, easier to title and much easier to cut.

Open with the point, not a greeting

Viewers decide quickly whether to stay, which is why Instagram’s creator guidance singles out the first 3 seconds. Skip “Hey everyone, today I want to talk about…” and start with the claim or the problem: “Your microphone is too far away, and no filter fixes that.” Then deliver on it. The Shorts retention guide explains how to check whether an opening worked.

Script, bullet points, or both

A full script keeps you on time and on point, but many people sound flat reading one. Bullet points sound natural, but they invite rambling and more retakes. A middle path works for most creators: write the first and last lines word for word, and keep the middle as three or four bullet points you talk through.

Then check the length. If you speak at about 150 words a minute, a 60-second video holds about 150 words. Paces vary, so time yourself once. The words to minutes calculator turns a word count into speaking time, and the script timer times each paragraph and shows which words fit in your first three seconds.

Film it: framing, eye line, light and sound

Frame for a vertical screen

For a 9:16 video, frame from the chest up with your eyes roughly a third of the way down the frame. Leave a little space above your head, not a lot. Then think about what will cover the frame later: app buttons on the right, your post caption and account name along the bottom, and your own captions somewhere in the middle.

Meta’s Reels ad guidance suggests keeping the top 14%, the bottom 35% and 6% on each side free of text, logos and other key elements. Each app’s layout differs and changes by device, so test a frame in the Reels safe zone checker before a long shoot.

Put the lens at eye level

Raise the camera until the lens sits level with your eyes. A lens below you looks up your nose; a lens far above makes you look small. Sit or stand at least an arm’s length away, because a wide phone lens close to a face stretches the nose and forehead. Look into the lens, not at your own image on the screen. In the edX study, the instructor who often looked into the lens drew more viewing time, though one pair of courses cannot prove that eye contact caused it.

Light your face, not the room

Face a window, or put your main light in front of you and slightly to one side. Avoid a bright window behind you, which turns your face into a silhouette, and a single overhead bulb, which drops shadows into your eyes. Switch off lights that are a different color from your main source so your skin doesn’t pick up two tints. If you film at different times of day, note where the light was so the next video matches.

Get the sound right first

Treat sound as the first thing to fix. In two experiments, Eryn Newman and Norbert Schwarz played identical conference talks and radio interviews in high or low audio quality. Listeners rated the researchers and their work less favorably when the audio was poor, even though the words were the same. Read the 2018 study.

Put the microphone close to your mouth: a clip-on lavalier or a small wireless mic beats a built-in mic across the room. Record in a room with soft furnishings, switch off fans and air conditioning, and silence notifications. Record ten seconds, then listen on headphones before you film the whole thing.

A phone is enough

YouTube’s own Shorts tips say you can record, edit and upload from your phone. Wipe the lens, lock focus and exposure so they don’t shift mid-sentence, record at the highest resolution your phone offers, and turn on Do Not Disturb. A dedicated camera helps in dim rooms and can soften the background, but it doesn’t fix muddy sound or a weak first line.

Use a teleprompter without looking like it

If you read a script, put the text as close to the lens as you can and keep the lines narrow so your eyes barely move. Set the speed so the next line is always just ahead of you, and read in phrases rather than word by word. The online teleprompter runs in your browser, scrolls at a steady speed and can mirror the text for a glass rig.

Record retakes an editor can use

Record your first line several times, because it carries the most weight. When you stumble, stop, pause for a beat, and restart the whole sentence rather than the last word. A clean pause before a full restart gives you, or your editor, an obvious place to cut.

Edit it: cuts, captions, B-roll and sound

Edit in passes, in this order, so later work doesn’t get undone by earlier changes. The complete talking head editing workflow walks through each pass with examples.

1. Choose takes and cut the dead air. Keep the best take of each line, then remove false starts, filler words and pauses that interrupt a thought. Keep the pauses that let a point land. The guide to removing pauses and filler words shows how to judge each cut by listening.
2. Smooth the jump cuts. Cutting between sentences of the same shot is normal in this format. To make a jump read as a choice, punch in slightly on every other section. Recording at a higher resolution than you export leaves room to do that without a soft image.
3. Add captions. Captions let viewers follow without sound and make the video accessible to more people. Proofread names, numbers and negations, break lines where the sentence makes sense, and keep the text clear of your face and the app’s buttons. See video captions people can actually read.
4. Add B-roll where words fall short. Show the setting, object or screen the sentence describes, then return to your face. A stock shot that only matches a keyword adds motion without meaning. See choosing B-roll for talking head videos.
5. Even out the sound. Reduce hum or background noise if you can, then normalize loudness so each video plays at a consistent level next to your others. The audio normalizer measures loudness in LUFS and lets you pick a target.
6. Watch the export on a phone. Watch once with sound, once without, and once with the app’s interface visible. Fix anything covered, unreadable or out of sync before you post.

Doing it with EditorOP

EditorOP is an AI video editor built for this format. You upload your raw talking head clips, every take included. It chooses between retakes, cuts the pauses and stumbles, and adds captions, B-roll and motion graphics to make a vertical reel up to 90 seconds long. You review the cut, change any caption, shot or graphic on an editable timeline, and export when it reads right. The filming advice above still matters: an editor can remove a stumble, but it cannot relight a face filmed against a window or fix a voice recorded across the room. See how the AI reel editor works.

Talking head video mistakes checklist

Run through this list before you post. Each item makes a talking head video feel longer or harder to follow than it needs to be.

• Opening with a greeting, your name or a logo instead of the point.
• Covering two or three ideas in one video.
• Camera below eye level, or eyes on the screen instead of the lens.
• A bright window behind you.
• A microphone across the room, or a fan running in the background.
• Reading a script in a flat voice, or improvising with no plan at all.
• Leaving every pause, false start and repeated sentence in the cut.
• Captions that cover your face, sit under the app’s buttons or misspell a name.
• Stock B-roll that matches a keyword but not the sentence.
• Posting without watching the export on a phone.

The downloadable checklist below turns the planning and filming steps into a list you can tick off on set, with blank rows for your own checks.