Video editing glossary

9:16 is the vertical aspect ratio used by TikTok, Reels, and Shorts, typically 1080 by 1920 pixels. Example. a phone held upright records 9:16. In Vyra. "export a 9:16 version at 1080x1920 for TikTok."

A-roll is the primary footage, usually the person talking to camera or the main action, that carries the story. Example. the interview is a-roll; the office shots over it are b-roll. In Vyra. the agent identifies talking-to-camera clips as the spine and treats other clips as coverage when you ask for a cut.

An AI video editor is editing software that performs cuts, captions, and effects from natural-language instructions rather than manual timeline work. Example. "remove the pauses and add captions" instead of dragging clips. In Vyra. the editor understands your real footage and edits it; it does not generate synthetic video.

Aspect ratio is the width-to-height proportion of a video frame, such as 16:9 for YouTube or 9:16 for TikTok. Example. a 1920x1080 video is 16:9; 1080x1920 is 9:16. In Vyra. "make a 9:16 version that keeps me centered" reframes with a mask.

Audio sync is aligning a separately recorded audio track, or multiple camera clips, so sound and picture match to the frame. Example. a lav mic recording on a phone lined up with the camera's scratch audio. In Vyra. the agent aligns clips by transcript matching; for music, beat sync uses BPM, beat positions, and onsets.

B-roll is supplementary footage cut over or between primary footage to show what is being talked about or to hide an edit. Example. a shot of the brushes while the speaker says "I use three brushes." In Vyra. "when I mention the brushes, cut to the brush footage for 2-3 seconds, then back to me."

Beat sync is cutting or animating on the beats of a music track so the edit feels locked to the sound. Example. one clip per beat in the chorus, one every two beats in the verse. In Vyra. upload the track and say "cut on the beats, start on the first downbeat."

Bitrate is the amount of data used per second of video, measured in megabits per second, and it sets the ceiling on how much detail an export can keep. Example. the same 1080p clip at 5 Mbps looks soft in motion, while at 20 Mbps the grain and text stay crisp. In Vyra. the export slider runs from 5 to 50 Mbps for MP4 and 5 to 20 Mbps for WebM; the Instagram and TikTok preset uses 20 Mbps.

Cadence is the rhythm of cuts and speech across a video, felt as how often the picture changes and how the pacing rises and falls. Example. an intro that cuts every second, a middle that holds for five, and a fast finish. In Vyra. give a rhythm in numbers ("average shot length three seconds, hold the talking clips") and the agent cuts to it.

A codec is the compression scheme that encodes video or audio into a file and decodes it for playback, separate from the container it lives in. Example. an MP4 container can hold H.264 or HEVC video; the codec decides file size and compatibility, not the .mp4 extension. In Vyra. exports use H.264 in MP4 for the widest compatibility, or VP8 in WebM for web delivery.

A cold open starts a video in the middle of action or a strong moment before any intro or greeting. Example. opening on the canvas being torn instead of "hey everyone." In Vyra. "start on the reveal shot, no intro, then go back to the setup."

A color grade is the deliberate adjustment of color and contrast across a video to create a consistent look. Example. warm shadows and muted highlights for a calm vlog. In Vyra. "grade this warmer and slightly desaturated, consistent across all clips."

Conversational editing is editing video through a back-and-forth chat where each message adjusts the cut. Example. "tighter," then "swap the intro clip," then "bigger captions." In Vyra. one correction per message keeps results accurate.

A CTA (call to action) is the line at the end of a video that tells the viewer what to do next, like follow, comment, or click. Example. "comment 'guide' and I will send it to you." In Vyra. "end on a 2-second title card that says 'comment GUIDE for the link'."

A cutaway is a brief shot away from the main action, used to show what is being discussed or to hide an edit in the primary footage. Example. a shot of the brushes while the speaker says "I only use three brushes." In Vyra. "when I mention the brushes, cut to the brush footage for two seconds, then back to me" is a cutaway request.

Ducking is lowering music volume automatically whenever speech is present so the voice stays clear. Example. music at 25 percent while talking, 70 percent during b-roll. In Vyra. "duck the music to about 20 percent whenever I am speaking."

An editing agent is an AI that operates an editor's tools directly, reading the footage and executing cuts, captions, and graphics from instructions instead of only suggesting them. Example. an agent that searches a transcript, trims pauses, and places captions in one turn. In Vyra. the built-in agent and any MCP client (Claude, ChatGPT, Cursor) get the same editing tools and see transcripts and scene descriptions.

Emphasis captions are caption phrases given a larger or accented treatment because they carry the hook, the turn, or the payoff. Example. "never again" set bigger and yellow while the surrounding captions stay plain white. In Vyra. AI caption templates classify each phrase for emphasis automatically, and you can steer them with guidance like "emphasize the hook in the first three seconds."

An export preset is a bundled resolution, bitrate, and format choice matched to a destination so you do not set them by hand each time. Example. 1080p at 20 Mbps MP4 for a Reel; 720p at 5 Mbps for a quick review file. In Vyra. presets are Instagram / TikTok, YouTube Shorts, High Quality, Balanced, Small File, and Custom; each shows an estimated file size.

A faceless video is content built from voiceover, screen recordings, b-roll, or stock footage without the creator appearing on camera. Example. a narrated explainer over stock clips and text callouts. In Vyra. assemble from your footage plus Pexels stock and Freesound effects, then add captions; the voice over panel is experimental.

A fine cut is the refined version of a rough cut with pacing, transitions, and trims adjusted, close to final. Example. pauses tightened, b-roll placed, hook moved to the top. In Vyra. iterate on the rough cut one note per message until it is a fine cut.

Frame rate is the number of still images shown per second, with 24 feeling cinematic, 30 standard for social, and 60 smooth for motion and slow-motion source. Example. a 60 fps clip slowed to half speed still plays at a clean 30 fps. In Vyra. the timeline keeps your source frame rate; ask for slow motion on a 60 fps clip and the agent applies time remapping.

H.264 is the most widely supported video codec, playable on every phone, browser, and platform, which is why most social exports use it. Example. TikTok, Instagram, and YouTube all ingest H.264 MP4 without re-encoding problems. In Vyra. the MP4 export option is H.264 at your chosen resolution, from 480p up to 4K upscale.

A highlight caption is a render mode where the full phrase stays on screen and the currently spoken word changes color as it is said. Example. a five-word line in white with each word turning yellow in turn. In Vyra. highlight is one of five caption render modes, alongside default, word-by-word, slide-up, and slide-left.

A hook is the first one to three seconds of a video, designed to stop a viewer from scrolling by promising a payoff, posing a question, or showing something surprising. Example. "I quit my job with $400 in the bank" is a hook; "hey guys, welcome back" is not. In Vyra. say "open on the line where I say X" or "start on the reveal shot" and the agent places it first.

Hook rate is the share of viewers who keep watching past the first few seconds, typically measured at three seconds. Example. 70 percent hook rate means seven of ten people stayed past second three. In Vyra. "open on the line where I say X" or "start on the reveal shot" moves the hook to the first frame.

A hyperlapse is a timelapse where the camera moves through space between frames, producing a fast, traveling shot. Example. walking through a city with the shot compressed to a few seconds. In Vyra. "speed the walking clip up 8x and keep it stabilized-looking by trimming the shaky start."

An insert shot is a close-up detail of something already in the scene, cut in to show texture, text, or a hand action clearly. Example. a close-up of a knife slicing after the wide shot of the cutting board. In Vyra. search your footage by what is in it ("the close-up of the palette") and ask the agent to insert it at the matching line.

A J-cut is an edit where the audio of the next shot starts before its picture appears. Example. you hear the street before you see the street. In Vyra. "let the next clip's sound start half a second early" or describe the effect in plain words.

A jump cut is an edit between two shots of the same subject from the same angle, which makes the subject appear to jump. Example. cutting out a pause in a talking head produces a jump cut. In Vyra. "punch in 10 percent on the second clip instead of a jump cut" smooths the transition.

A keyframe is a point on the timeline where a property such as position, scale, opacity, or volume is fixed, with the editor interpolating values between keyframes. Example. two volume keyframes take music from 100 percent to 20 percent over one second when the speaker starts. In Vyra. volume, position, scale, and effect parameters can be keyframed; say "fade the music down over one second when I start talking."

Kinetic typography is text that moves, scales, or appears in time with speech or music to add emphasis. Example. the word "never" popping up larger as it is spoken. In Vyra. "animate the three tool names on screen one at a time as I say them, then clear them together."

An L-cut is an edit where the audio of the current shot continues after its picture has changed. Example. the speaker keeps talking while the video shows the product. In Vyra. "keep my voice going while you show the product shot" produces an L-cut.

A layout is a multi-cell arrangement of clips on one canvas, such as a split screen, a grid, or a picture-in-picture inset. Example. a reaction video shows the source clip on top and the creator's face below in a vertical split. In Vyra. layout presets include split horizontal and vertical, three-up, 2x2 grid, PiP in any corner, and feature left or right.

A lower third is a text graphic in the bottom third of the frame that names a speaker, place, or topic. Example. "Sulan Zhang, @sulansart" appearing for the first four seconds. In Vyra. "add a lower third with my name and handle for the first 4 seconds, slide in from the left."

A LUT (lookup table) is a preset color transformation applied to footage to achieve a specific look quickly. Example. a "teal and orange" LUT applied to every clip. In Vyra. describe the look you want; the agent applies matching color effects without requiring a LUT file.

A mask is a shape that hides or reveals part of a layer, used for cropping, reframing, spotlighting a subject, and depth effects. Example. a rectangle mask crops a 16:9 clip to the vertical center for a Reel. In Vyra. masks are how cropping and reframing work; there is no separate crop tool. Rectangle, ellipse, polygon, feathered, and tracked smart masks are available.

A match cut joins two shots that share a shape, motion, or framing so the edit feels continuous even though the subject changes. Example. a coffee cup set down in a cafe cuts to a paint jar set down in the studio at the same spot in frame. In Vyra. describe both moments and the shared element; the agent can align the cut so the motion carries across.

MCP (Model Context Protocol) is an open standard that lets AI assistants connect to external tools and data; Vyra uses it so Claude, ChatGPT, and Cursor can control the editor. Example. Claude Desktop connected to Vyra can list your footage and cut a video. In Vyra. connect your assistant once, then edit from its chat.

A montage is a sequence of short clips cut together, usually to music, to compress time or build energy. Example. a 30-second recap of a trip cut on the beat. In Vyra. "build a 30-second montage from the strongest 10 clips, fast to slow to fast, synced to the track."

Motion graphics are animated graphic elements such as titles, callouts, charts, and shapes, layered over or between footage. Example. an animated arrow pointing at a button in a screen recording. In Vyra. the agent builds motion graphics from a description or a reference; they are not templates.

Multicam is editing footage from two or more cameras that recorded the same moment, cutting between angles on one synced timeline. Example. a podcast with a wide two-shot and a close-up per host, switching to whoever is talking. In Vyra. multi-camera sync aligns angles by matching their transcripts; then ask the agent to cut to the speaker.

Noise reduction lowers constant background sound such as hum, hiss, or room tone so speech sits cleaner in the mix. Example. a fridge hum under a kitchen vlog drops away while the voice stays intact. In Vyra. cleanup is part of the audio guides; describe the problem ("there is a hum under my voice") and keep expectations honest, since heavy noise is best fixed at recording.

An open loop is a question or promise raised early in a video and resolved later, so viewers stay to see the answer. Example. "the third mistake cost me a client, more on that in a second." In Vyra. "keep the line where I tease the third mistake near the top and the payoff at the end."

A pattern interrupt is a sudden change in visuals, sound, or pacing meant to reset attention and keep viewers watching. Example. a cut to black, a zoom, a sound effect, or a change of location. In Vyra. "add a pattern interrupt around the 8-second mark: cut to the wide shot with a sound hit."

Picture lock is the point where all visual edits are final and only sound, captions, and color remain. Example. no more clip changes after this; captions and music are added on top. In Vyra. lock the cut before asking for captions and music so they do not need redoing.

A POV video is shot or framed so the viewer feels they are seeing from the creator's point of view, often with an on-screen line such as "POV, you are the new intern." Example. a phone camera at eye level walking into a room with the text on screen. In Vyra. "add the POV text at the top for the first 3 seconds, then remove it."

A preset is a saved, reusable configuration for a text treatment, motion graphic, caption style, or export setting. Example. a lower third with your handle saved once and reused across every video. In Vyra. select a motion graphic on the timeline and click "Save as Preset"; the Presets panel also ships title cards, lower thirds, info cards, panels, overlays, and collages.

In video editing, a prompt is a plain-language instruction to an editing agent that names the target, what to keep, what to cut, and how text and sound should behave. Example. "Cut this to 45 seconds for Reels, keep the best take of each point, 3 to 5 word captions, no music." In Vyra. see the prompt library for minimal, standard, and full versions of each task; specifics beat adjectives.

A proxy file is a smaller, lower-resolution copy of a source clip used for smooth playback while editing, with the original swapped back in for export. Example. a 4K phone clip plays as a 720p stand-in on the timeline so scrubbing does not stutter. In Vyra. every upload gets a 720p proxy automatically, and export always uses the original file, never the proxy.

Proxy mode is an editor setting that plays lightweight proxy copies instead of full-resolution originals so the timeline stays responsive. Example. scrubbing a 4K project feels instant because the editor is reading 720p stand-ins. In Vyra. proxy mode is on by default; exports always render from the originals, so quality is never affected.

A punch-in is a cut to a slightly tighter framing of the same shot, used for emphasis or to hide a jump cut. Example. zooming from a medium shot to a close-up on the same take. In Vyra. "punch in 10 percent on the key line" or "use punch-ins instead of jump cuts."

A reference frame is a single still from a reference video used to show an agent the exact look, framing, or text treatment to match. Example. a frame from a favorite creator's video with big yellow captions under the chin. In Vyra. scrub the reference video and click "Add to chat" at any timestamp to attach that frame to your request.

A reference video is an example video whose style, pacing, or layout you want your own edit to match. Example. attaching a trend video and asking for the same cut timing. In Vyra. attach it, then say what to copy: "match its cut rhythm and caption placement."

Retention is the percentage of viewers still watching at each point in a video; platforms use it to decide how widely to distribute the video. Example. a 30-second video where half the audience leaves by second 5 has a retention problem at the hook, not the ending. In Vyra. ask for tighter pauses, an earlier hook, or a mid-video re-hook to address specific drop points.

Room tone is the ambient sound of a recording space with no one speaking, used to fill gaps so cuts do not go dead silent. Example. the faint hum of a studio behind an interview. In Vyra. "keep the original ambience under the cut" preserves room tone through edits.

A rough cut is the first full assembly of a video with the right structure and takes, before captions, music, and polish. Example. a 60-second cut with the best takes in order and nothing else. In Vyra. "make a rough cut, no captions or music yet, straight cuts only."

The safe zone is the part of the frame that stays visible and unobstructed by platform interface elements like buttons and captions. Example. on TikTok, the bottom 20 percent and right edge are covered by UI. In Vyra. "keep captions above the bottom 20 percent and away from the right edge."

Scene detection is automatic identification of where one shot ends and another begins, plus what each shot contains. Example. an hour of footage split into 200 labeled clips. In Vyra. every upload is scene-analyzed, so "find every shot of the canvas" works.

Shares are viewers sending a video to someone else, saves are viewers bookmarking it, and platforms weight both as strong intent signals beyond likes. Example. a tutorial gets saves; a relatable skit gets shares. In Vyra. choose the format for the signal you want, then prompt for it; the format pages list which formats earn which.

A sound effect is a short non-musical audio cue, such as a whoosh, click, or impact, placed to underline a cut, a reveal, or a text pop. Example. a soft whoosh on a fast pan; a single click when a caption lands. In Vyra. the Audio panel searches Freesound with preview; the default direction is selective cues, not one per word.

A speed ramp is a gradual change in playback speed inside one clip, usually fast to slow or slow to fast, timed to a beat or a gesture. Example. a walk-in starts at 2x, slows to 0.5x as the person turns to camera, then returns to normal. In Vyra. ask for a ramp with timestamps and the agent sets time remapping keyframes; 60 fps source keeps the slow part smooth.

Stock footage is licensed video or photo clips from a library, used to cover gaps when you have no footage of your own for a moment. Example. a drone shot of a city to open a travel recap you filmed on the ground. In Vyra. the Videos and Images panels search Pexels; click to add, or ask the agent to find "a wide city skyline at night."

A storytelling video is a short built on a narrative arc with a setup, a turn, and a payoff, rather than a list or a demonstration. Example. "I quit my job with $400" as the setup, the first failed launch as the turn, the first sale as the payoff. In Vyra. quote the exact lines that mark each beat and the agent structures the cut around them.

In Vyra, a style is the visual identity applied to a project, covering the caption face, title face, accent face, and color palette. Example. Vyra Default uses Inter Bold for captions and Montserrat Black for titles; Cinematic Film shifts type and color toward a filmic look. In Vyra. choose a style when creating a project, or ask the agent to switch it mid-edit; templates set structure, styles set look.

Style matching is replicating the pacing, text treatment, color, and music behavior of a reference video on different footage. Example. your studio footage cut like a creator's vlog you admire. In Vyra. "use the reference as the style guide; do not match its content or length."

A talking head is a video format where a person speaks directly to camera, usually framed from the chest up. Example. a founder explaining a feature from their desk. In Vyra. "clean up this talking head: remove pauses over 0.4 seconds, filler words, and retakes."

In Vyra, a template is structural guidance for the agent about how a kind of video is assembled, such as Short Talking Head, Long Podcast, or Vlog. Example. the Vlog template favors music-driven pacing and location changes; Long Podcast favors speaker switching and chapter marks. In Vyra. pick a template at project creation or let the agent set one in conversation; the template can be adapted with your own preferences.

Time remapping is changing a clip's playback speed over time, covering slow motion, fast forward, freeze frames, and speed ramps. Example. a two-minute pour becomes a five-second timelapse with a hold on the final frame. In Vyra. time remapping handles timelapses, slow motion, and freezes; say "speed this section up 8x and hold the last frame for one second."

A timelapse is video captured at a low frame rate and played back at normal speed so slow processes appear fast. Example. a painting completed in 20 seconds of screen time. In Vyra. speed up a clip with time remapping: "play the painting clip at 20x."

A title card is a full-frame or near-full-frame text screen that names a section, a chapter, or the video itself. Example. a black card reading "Day 3" between two locations in a travel vlog. In Vyra. ask for a title card at a timestamp and the agent builds it from the title face in your style, or pick one from the presets panel.

Transcript editing is cutting video by editing its text. Delete the words and the matching footage is removed. Example. removing a false start from the transcript trims the retake from the timeline. In Vyra. every upload is transcribed with word timestamps; say "remove every sentence I restart" or quote the exact line to cut.

Transcription is converting spoken audio into timestamped text. Example. a 40-minute recording becomes searchable text with word timings. In Vyra. speech is transcribed on upload, which powers search, cutting by phrase, and captions.

A transition is the treatment between two shots, from a straight cut to a dissolve, wipe, or whip pan, and the straight cut is the default for a reason. Example. a slow dissolve between a sunset and a night skyline signals time passing. In Vyra. transitions are applied on request; the default creative direction favors clean cuts, so say "add a quick dissolve between the two locations" when you want one.

UGC (user-generated content) is video that looks like it was made by a regular customer rather than a brand, often used in ads. Example. a phone-shot "I tried this for a week" review. In Vyra. cut a UGC-style ad with "open on the hook line, show the product in use, end on the result."

Upscaling is exporting at a higher resolution than the source, which enlarges the frame but cannot add detail the camera did not capture. Example. a 1080p phone clip exported at 4K is smoother to scale on a big screen but no sharper than the original. In Vyra. the export dialog offers 4K and 1440p as upscale options alongside Original, 1080p, 720p, and 480p.

Voiceover is narration recorded separately and laid over the picture, common in faceless, tutorial, and recap formats. Example. a calm narration over a montage of studio shots. In Vyra. upload your own recording and duck the music under it, or try the experimental voice over panel that generates speech from a script.

Watch time is the total minutes viewers spend on a video, and average view duration is the per-viewer version platforms use to rank it. Example. a 60-second Reel with a 45-second average view duration is retaining 75 percent. In Vyra. improve it with tighter pauses, an earlier hook, and a mid-video re-hook; see the retention guides for prompts.

WebM is an open web video container that pairs with VP8 or VP9 video, useful for browser embeds and lightweight web delivery. Example. a product page background loop is often served as WebM to keep the file small. In Vyra. WebM export uses VP8 and is limited to 1080p and 20 Mbps; pick MP4 for social platforms.

Word-level captions display each word on screen at the moment it is spoken, rather than a whole sentence at once. Example. TikTok-style captions that pop one word at a time. In Vyra. "caption the whole video word by word, bold white, slightly above center."