# Glossary of caption, dubbing and video terms

Page: https://muvomi.com/glossary
Updated 2026-11-10. Each word has an anchor of its own: https://muvomi.com/glossary#srt is SRT.

Short, plain definitions of the words you meet when you caption, dub, voice and edit a video. Under each one is what it means in Muvomi.

## Captions and subtitles

### Captions

Captions are the words of a video shown on screen in time with the speech. They can be in the language that is spoken or in another one, and they may also describe sounds, such as music or laughter, for viewers who cannot hear them.

In Muvomi: [Auto captions](https://muvomi.com/tools/auto-captions) write them from the speech in a video, with a time for every word, in 51 languages.

### Subtitles

Subtitles are on-screen text of what is said in a video. Strictly, subtitles are for viewers who can hear the sound but do not know the language, so they are often a translation, while captions also carry sounds for viewers who cannot hear. In everyday use the two words mean the same thing.

In Muvomi: Captions can be written in the language that is spoken or translated into 51 languages, and downloaded as SRT or VTT subtitle files.

### Closed captions

Also: CC.

Closed captions are kept apart from the picture, as a separate track or file, so a viewer can switch them on or off in the player. Most players mark them with the letters CC.

In Muvomi: For captions a viewer can switch off, download the SRT or VTT file and give it to the platform's own caption track.

### Open captions

Also: burned-in captions, hardcoded subtitles.

Open captions are drawn into the picture itself, so they are always on and show wherever the video plays, even where a player has no caption setting. They cannot be switched off or restyled afterwards.

In Muvomi: An MP4 exported from the [video editor](https://muvomi.com/tools/online-video-editor) has its captions drawn into the picture, in the look you chose.

### SRT

Also: SubRip, .srt file.

SRT (SubRip) is a plain text subtitle file. Each caption in it has a number, a start and an end time, and its text, and most video players and platforms accept it.

In Muvomi: The Captions panel downloads an SRT file of the captions as they are written. It carries the text and the times; the look and the word highlighting exist only in the exported video.

### VTT

Also: WebVTT, .vtt file.

VTT, in full WebVTT (Web Video Text Tracks), is the subtitle format of web browsers, read by the caption track of a web page's video player. It is close to SRT, with a full stop before the milliseconds where SRT has a comma, and it can also say where on the screen a caption sits.

In Muvomi: The Captions panel downloads a VTT file beside the SRT, and a dub comes with its subtitles in both.

### Word-by-word captions

Also: karaoke captions, word-level captions.

Word-by-word captions give every word its own start and end time, so the word being spoken can be lit, underlined or revealed as it is said. They are common on short videos, where the style is often called karaoke captions.

In Muvomi: Every word carries its own time, and the looks in the group Follow the voice use it to mark the word being said: [caption templates](https://muvomi.com/caption-templates).

### Auto captions

Also: automatic captions, AI captions, speech to text.

Auto captions are captions written by software that listens to the speech in a recording and turns it into timed text. How accurate they are depends on the recording, so they are read through and corrected before a video is published.

In Muvomi: Press Auto-caption in the [video editor](https://muvomi.com/tools/online-video-editor) and the speech becomes captions on the timeline. A captions job uses 10 credits.

### Transcription

Also: transcript, video to text.

Transcription is writing down what is said in a recording as text. A transcript is the whole text, with or without times; captions are the same words cut into short lines that appear as they are said.

In Muvomi: A recording is transcribed once and its transcript is kept, so its captions can be written again another way without listening again.

### Caption template

Also: caption style, caption look.

A caption template is a ready-made look for captions: the typeface, the colours, the box or outline behind the words, and the way each word appears.

In Muvomi: There are 87 caption templates in 9 groups, each playing on the [caption templates page](https://muvomi.com/caption-templates).

## Languages and scripts

### Translation

Translation carries the meaning of what is said into another language. Translated captions let people who do not speak a video's language follow it.

In Muvomi: Captions can be translated into 51 languages, phrase by phrase, each phrase shown while it is said: [caption languages](https://muvomi.com/languages).

### Transliteration

Transliteration writes the words of a language in the letters of another script, without changing the language. Hindi written in Latin letters is still Hindi; Hindi put into English is a translation.

In Muvomi: In the Captions tool, Write in: Latin letters writes a language the way people type it, and the English words in it stay as they were said.

### Romanised captions

Also: romanisation, captions in Latin letters, captions in English letters.

Romanised captions are captions for a language that has a script of its own, written in Latin (English) letters instead. They suit viewers who speak the language but usually type and read it in Latin letters.

In Muvomi: Auto captions write a language in Latin letters under the name people know it by: Hinglish, Banglish, Telugish, Tenglish, Tanglish, Gujlish, Roman Urdu, Kanglish, Manglish, Punglish, Pinyin, Romaji, Finglish, Greeklish.

### Hinglish

Hinglish is Hindi written in Latin (English) letters, the way many people type it in messages and on social media, usually with English words mixed in. The word also names that everyday mix of Hindi and English in speech.

In Muvomi: [Hindi captions](https://muvomi.com/languages/hindi) can be written in Hinglish, in Devanagari or as a translation.

### Banglish

Banglish is Bengali written in Latin (English) letters, the way many people type it in messages and on social media, usually with English words mixed in.

In Muvomi: [Bengali captions](https://muvomi.com/languages/bengali) can be written in Banglish, in Bengali script or as a translation.

### Telugish

Also: Tenglish.

Telugish, also called Tenglish, is Telugu written in Latin (English) letters, the way many people type it in messages and on social media, usually with English words mixed in.

In Muvomi: [Telugu captions](https://muvomi.com/languages/telugu) can be written in Telugish or Tenglish, in Telugu script or as a translation.

### Tanglish

Tanglish is Tamil written in Latin (English) letters, the way many people type it in messages and on social media, usually with English words mixed in.

In Muvomi: [Tamil captions](https://muvomi.com/languages/tamil) can be written in Tanglish, in Tamil script or as a translation.

### Gujlish

Gujlish is Gujarati written in Latin (English) letters, the way many people type it in messages and on social media, usually with English words mixed in.

In Muvomi: [Gujarati captions](https://muvomi.com/languages/gujarati) can be written in Gujlish, in Gujarati script or as a translation.

### Roman Urdu

Roman Urdu is Urdu written in Latin letters instead of its own right-to-left script, the way many people type it on phones and online.

In Muvomi: [Urdu captions](https://muvomi.com/languages/urdu) can be written in Roman Urdu, in Urdu script or as a translation.

### Kanglish

Kanglish is Kannada written in Latin (English) letters, the way many people type it in messages and on social media, usually with English words mixed in.

In Muvomi: [Kannada captions](https://muvomi.com/languages/kannada) can be written in Kanglish, in Kannada script or as a translation.

### Manglish

Manglish is Malayalam written in Latin (English) letters, the way many people type it in messages and on social media, usually with English words mixed in. In Malaysia the same word names a local form of English.

In Muvomi: [Malayalam captions](https://muvomi.com/languages/malayalam) can be written in Manglish, in Malayalam script or as a translation.

### Punglish

Punglish is Punjabi written in Latin (English) letters, the way many people type it in messages and on social media, usually with English words mixed in.

In Muvomi: [Punjabi captions](https://muvomi.com/languages/punjabi) can be written in Punglish, in Gurmukhi or as a translation.

### Pinyin

Pinyin is the standard way of writing Mandarin Chinese in Latin letters, with marks over the vowels for its four tones.

In Muvomi: [Chinese (Mandarin) captions](https://muvomi.com/languages/chinese) can be written in Pinyin, in Chinese characters or as a translation.

### Romaji

Romaji is Japanese written in Latin letters. The Hepburn spelling is the most widely used.

In Muvomi: [Japanese captions](https://muvomi.com/languages/japanese) can be written in Romaji, in kanji and kana or as a translation.

### Finglish

Finglish, also spelled Fingilish, is Persian written in Latin letters, the way many people type it in messages and online.

In Muvomi: [Persian captions](https://muvomi.com/languages/persian) can be written in Finglish, in Persian script or as a translation.

### Greeklish

Greeklish is Greek written in Latin letters, the way many people type it in messages and online.

In Muvomi: [Greek captions](https://muvomi.com/languages/greek) can be written in Greeklish, in the Greek alphabet or as a translation.

## Voice and dubbing

### AI dubbing

Also: AI video dubbing, video translation.

AI dubbing replaces the speech in a video with the same lines spoken in another language. The words are translated and then voiced by synthetic voices, and the picture stays as it was.

In Muvomi: [AI Dubbing](https://muvomi.com/tools/ai-video-dubbing) says every line again in one of 92 languages, each speaker in a voice close to their own, over the video's own music and sound, with subtitles in the new language.

### Voice cloning

Voice cloning makes a synthetic voice that sounds like one particular person, built from recordings of that person speaking.

In Muvomi: AI Dubbing gives each speaker a new voice close to their own, not their own voice, and the Voiceover tool takes no voice sample.

### Voiceover

Also: voice-over, narration.

A voiceover is a voice heard over a video while the speaker is not seen: the narration of an explainer, an advert or a lesson.

In Muvomi: [Voiceover](https://muvomi.com/tools/ai-voiceover) reads a script of up to 1,200 characters exactly as written, in one of 12 voices and 70 languages, and Motion Design writes and records a whole film's narration.

### Text to speech

Also: TTS.

Text to speech turns written words into spoken audio in a synthetic voice.

In Muvomi: The 70 languages a script can be read in are listed on [voiceover languages](https://muvomi.com/voiceover).

### Lip-sync

Also: lip sync.

Lip-sync is the matching of a speaker's mouth movements to the words that are heard. In dubbing it means changing the picture so that the lips fit the new language.

In Muvomi: AI Dubbing does not change anyone's lip movements: each new line is placed where the original line was said.

## Sound

### Audio ducking

Also: ducking.

Audio ducking lowers one sound, usually the music, while another is heard, usually a voice, and brings it back afterwards.

In Muvomi: In Motion Design films, voiced templates and shorts the music drops while a line is said, and in the video editor it can be ducked under the speech as volume keyframes you can change: [music and sound](https://muvomi.com/tools/background-music).

### Noise reduction

Also: noise removal, denoising.

Noise reduction lowers steady unwanted sound in a recording, such as hiss, hum, a fan or room tone, so that the voice is clearer.

In Muvomi: [Audio Cleanup](https://muvomi.com/tools/audio-cleanup) takes hiss, fans, room tone and mains hum out of a recording, and plays each result against the original before you keep it.

## Video and editing

### Aspect ratio

Also: frame shape, 16:9, 9:16, 1:1, 4:5.

Aspect ratio is the shape of a video frame, written as its width to its height. 16:9 is the wide frame of YouTube and television, 9:16 the upright frame of Reels, Shorts and TikTok, 1:1 a square, and 4:5 a tall frame for social feeds.

In Muvomi: The video editor and Motion Design work in 16:9 at 1920 × 1080, 9:16 at 1080 × 1920, 1:1 at 1080 × 1080 or 4:5 at 1080 × 1350.

### Resolution

Also: 720p, 1080p, 4K.

Resolution is the number of pixels in a video frame. For a wide 16:9 frame, 720p is 1280 × 720 pixels, 1080p is 1920 × 1080 and 4K is 3840 × 2160.

In Muvomi: Exports go up to 720p with the free credits, and up to 1080p at no cost, or 4K for 5 credits, while you have credits you bought: [pricing](https://muvomi.com/pricing).

### Motion graphics

Also: motion design.

Motion graphics are graphic design that moves: animated type, shapes, charts, logos and product pictures, as opposed to footage filmed with a camera.

In Muvomi: [Motion Design](https://muvomi.com/tools/ai-motion-design) designs every scene of a film as animated motion graphics from an idea, a script or a website, with voice-over, music and sound.

### Motion template

Also: video template.

A motion template is a finished animated film with places for your own words, colours, pictures and logo, so a new video starts from a design that already works.

In Muvomi: There are 52 motion templates, each with a preview film, in [the template library](https://muvomi.com/templates).

### B-roll

B-roll is extra footage cut in over the main footage: establishing shots, cutaways and details that show what is being talked about.

In Muvomi: [AI Video](https://muvomi.com/tools/ai-video-generator) makes a short clip from a description, or from a picture used as its first frame, to fill a gap in a cut with a shot nobody filmed.

### Timeline

Also: multitrack timeline, multitrack editing.

A timeline is the editing view that lays clips out in the order and at the moments they play. A multitrack timeline stacks layers, so that video, pictures, titles and sound can overlap.

In Muvomi: The [video editor](https://muvomi.com/tools/online-video-editor) has up to 16 video and 16 audio tracks, and an edit runs for up to 60 minutes.

### Keyframe

A keyframe fixes a value, such as a position, a size or a volume, at one moment. The software fills in the change between two keyframes, which is how a zoom, a pan or a fade is made.

In Muvomi: In the video editor, framing points make a pan and zoom of your own, and music ducking is written as volume keyframes you can change.

### Brand kit

A brand kit keeps a brand's logo, colours and typeface in one place, so that every video and graphic uses the same ones.

In Muvomi: A [brand kit](https://muvomi.com/tools/brand-kit) holds your logo, an accent, a background and a text colour, and a typeface, and is applied to a video edit in one press.

### Text to video

Text to video is the making of a video clip from a written description: an AI model generates footage that was never filmed.

In Muvomi: Describe a scene and [AI Video](https://muvomi.com/tools/ai-video-generator) makes it as a clip, with a camera direction and a look of your choosing. For a whole film from a brief, there is [Motion Design](https://muvomi.com/tools/ai-motion-design).

### Image to video

Also: first frame to video.

Image to video turns a still picture into a moving clip: an AI model takes the picture as its starting frame and generates what happens next.

In Muvomi: In AI Video, a picture from your library is the first frame of the clip and your prompt says what happens next.

## Credits and exports

### Credits

Credits are units of use that a service sells: each piece of work costs a set number of them, so you pay for what you make.

In Muvomi: A new account gets 100 credits free, once. More are bought once, from $5 for 500 credits, and stay in your account until you use them; there is no subscription. A template film uses 40, captions 10 and an AI edit 20: [pricing](https://muvomi.com/pricing).

### Watermark

A watermark is a mark, usually a logo or a line of text, laid over a picture or a video to show where it was made.

In Muvomi: Work made with the free credits carries a Muvomi watermark. While you have credits you bought, exports have none.

## Questions

**What is the difference between captions and subtitles?**

Subtitles are for viewers who can hear a video but do not know its language, so they are often a translation. Captions are for viewers who cannot hear it, so they also carry sounds such as music or laughter. In everyday use, and on most platforms, the two words mean the same thing.

**What is the difference between open and closed captions?**

Open captions are drawn into the picture and are always on. Closed captions are a separate track or file, such as an SRT, that a viewer can switch on or off in the player.

**What is the difference between SRT and VTT?**

Both are plain text subtitle files with a start and an end time for each caption. SRT is the older of the two and is accepted by most players and platforms. VTT (WebVTT) is the web's own format: it writes the milliseconds after a full stop where SRT has a comma, and can also say where on the screen a caption sits.

**What is the difference between translation and transliteration?**

Translation changes the language and keeps the meaning. Transliteration keeps the language and changes the letters: Hindi written in Latin letters, as Hinglish, is still Hindi.

**What is the difference between Hinglish and Hindi captions?**

Both are Hindi. Hindi captions are written in Devanagari, the script of Hindi; Hinglish captions are the same words in Latin (English) letters, the way many people type them.

**What is the difference between dubbing and a voiceover?**

Dubbing replaces the speech that is already in a video with the same lines in another language. A voiceover adds a new voice over a video, reading a script written for it.

---

From Muvomi (https://muvomi.com/). Muvomi is an AI creative platform that runs in the browser. In one studio you can turn an idea, a script or a website into a finished motion design film, make AI video clips, character shorts and images, and edit on a real timeline. The same place adds captions in 51 languages, AI voiceover in 70 and dubbing into 92, and keeps your designs on brand.
The site as a map for language models: https://muvomi.com/llms.txt
