How to Audio Describe Different Types of Educational Videos: A Format-by-Format Guide

Not all educational videos are the same, and the way you audio describe them should not be the same either. A screencast with code scrolling across the screen presents completely different challenges than a chemistry lab demonstration or a talking head over a slide deck. For example, typical reccomendations of describing all on-screen text don't work well when the entire screen is covered in code and rapidly shifts between opened files in an IDE. The visual information that matters, the pacing, and the level of detail required all change depending on the format.

This guide breaks down the most common educational video formats and explains what to prioritize, what you can safely omit, and whether standard or extended audio descriptions are the better fit. Whether you are producing descriptions manually or reviewing AI-generated output, these format-specific techniques from Recap Innovations will help you create descriptions that are accurate, useful, and well-timed.

Why format matters for audio description

Audio description has a fundamental constraint: time. Descriptions must fit into pauses in the existing audio (standard mode) or the video must be paused to allow for longer narration (extended mode). Either way, you cannot describe everything. You have to make choices about what visual information is most important for a viewer who cannot see the screen.

Those choices depend entirely on the video format. In a lecture capture, the slides carry most of the information. In a lab demonstration, the physical actions and their visible results are what matter. In a screencast, the on-screen interface and cursor movements tell the story. Knowing what to focus on for each format is the difference between a description that helps a viewer follow along and one that overwhelms them with irrelevant detail or leaves out what they actually need.

For a deeper look at how standard and extended modes work, see our audio descriptions product page.

Lecture capture

Lecture capture is the most common format in higher education. A camera records the instructor, the room, and often a screen or projector displaying slides. The instructor's spoken words carry the primary content, but slides, gestures, and board work provide essential context that is invisible to a listener relying on audio alone.

What to describe

• Slide content: Read or summarize text, bullet points, and headings when the instructor does not read them aloud. If the instructor reads the slide verbatim, a description is not needed.
• Diagrams and figures on slides: Describe charts, graphs, diagrams, and images that appear on screen. Prioritize the data or relationship being illustrated rather than every visual detail.
• Transitions between slides: Note when a new slide appears, especially if the instructor does not verbally signal the change.
• Instructor actions: Describe when the instructor points to something specific on the screen, picks up a prop, or moves to a different part of the room in a way that changes context.
• On-screen annotations: If the instructor draws on slides, circles items, or highlights content, describe what is being marked and why it appears significant.

What you can skip

• The instructor's physical appearance, clothing, or general position in the room (unless it changes in a meaningful way).
• Minor movements like pacing or shifting weight.
• Repeated slide layouts or formatting that does not change the meaning.

Standard vs. extended

Standard mode works well for most lectures. Instructors naturally pause between slides and between ideas, giving enough space for brief descriptions. Extended mode is appropriate when slides contain dense diagrams, data tables, or complex figures that require more time to describe than the instructor's pauses allow.

Screencast

Screencasts record what is happening on a computer screen, often with a voiceover narrating the steps. They are common in software tutorials, coding lessons, IT training, and demonstrations of digital tools. The visual information is almost entirely on-screen interface elements, text, and cursor movements.

What to describe

• Interface elements: Identify the application, window, or webpage visible on screen when it first appears or changes. Mention menus, toolbars, panels, modals, and dialog boxes as they become relevant.
• Cursor and click actions: Describe where the cursor moves and what is being clicked, especially when the narrator says something like "click here" or "select this" without specifying the element by name.
• Text entry and code: When the instructor types code or text, describe what is being typed if the narration does not cover it. For code, focus on the structure and purpose rather than reading every character.
• Output and results: Describe what appears on screen after an action: error messages, successful output, visual changes in the interface, or new content loading.
• Navigation: Note when the instructor switches tabs, opens a new file, scrolls to a different section, or navigates to a different page.

What you can skip

• Mouse movements that do not result in an action.
• Repetitive interface elements that have already been described and have not changed.
• Exact pixel positions or layout details that do not affect understanding.

Standard vs. extended

Extended mode is often the better choice for screencasts, especially those involving code or multi-step procedures. Screencasts tend to have continuous narration with few natural pauses, and the on-screen content changes rapidly. Pausing the video to describe interface changes or code output gives the listener time to absorb both the narration and the description. Standard mode can work for slower-paced tutorials where the instructor pauses between steps.

Talking head with slides

This format combines a webcam view of the speaker with a presentation displayed alongside or behind them, often in a picture-in-picture layout. The speaker's face and body language are visible, but the slides carry the substantive content.

What to describe

• Slide content: Same priorities as lecture capture. Describe text, diagrams, and images that the speaker does not verbalize.
• Layout changes: Note when the layout shifts, for example from full-screen slides to a split view, or when the speaker's video becomes full screen.
• Speaker gestures with context: If the speaker points at a specific part of a slide or holds up an object, describe the target of the gesture.
• Embedded media: If the slides include embedded video clips, animations, or interactive elements, describe what plays.

What you can skip

• The speaker's facial expressions in most cases (unless they are reacting to something that changes the meaning, such as expressing skepticism about a claim on screen).
• The speaker's general appearance after the initial introduction.
• Decorative slide backgrounds or transitions that do not convey information.

Standard vs. extended

Standard mode is usually sufficient. This format tends to have a natural speaking rhythm with pauses between slides. Extended mode may be needed for slides with complex data visualizations or detailed diagrams.

Whiteboard and lightboard

Whiteboard videos show an instructor writing, drawing, or building diagrams on a physical or digital whiteboard. Lightboard videos are similar but use a transparent board that the instructor writes on while facing the camera. In both cases, the content is built incrementally and the visual progression is the core of the lesson.

What to describe

• What is being written or drawn: Describe text, equations, diagrams, and labels as they appear. For equations, use a readable spoken format (for example, "x squared plus two x equals zero" rather than trying to replicate the visual layout).
• Spatial relationships: When the instructor draws a diagram, describe the structure: "a box labeled 'input' connects with an arrow to a box labeled 'process.'" The relationships between elements matter more than exact positions.
• Color coding: If the instructor uses different colors to distinguish categories, variables, or steps, note the color and what it represents.
• Erasing and revising: When the instructor erases content or crosses something out, describe what was removed and what replaces it.
• Gestures to the board: When the instructor circles, underlines, or taps on a specific element, describe which element they are emphasizing.

What you can skip

• The instructor's hand or arm movements as they write (the result of the writing is what matters, not the physical act).
• Minor adjustments to handwriting or positioning.
• Content that the instructor reads aloud as they write.

Standard vs. extended

Extended mode is frequently the right choice. Whiteboard content builds continuously, and the instructor often talks while writing, leaving no natural pauses. The visual content is also cumulative: early elements remain relevant as new ones are added. Extended descriptions allow the listener to keep a mental model of the entire board, not just the latest addition.

Lab demonstration

Lab demonstrations show physical experiments, medical procedures, technical assembly, or other hands-on activities. The visual content is dense, sequential, and often safety-critical. These videos are common in STEM courses, healthcare education, and vocational training.

What to describe

• Setup and equipment: At the start, describe the workspace layout, the equipment and materials present, and any safety gear being used.
• Physical actions: Describe what the instructor does with their hands, the tools, and the materials. Be specific: "pours the clear liquid from the beaker into the flask" is more useful than "adds the liquid."
• Observable results: Describe color changes, phase changes, movement, sounds, readings on instruments, and other visible outcomes. These are often the entire point of the demonstration.
• Measurements and readings: When the instructor checks a gauge, thermometer, scale, or other instrument, describe the reading if it is visible on screen.
• Safety-relevant actions: Describe the use of safety equipment, protective steps, and any warnings demonstrated through action rather than speech.

What you can skip

• Background details of the lab environment that do not change.
• Repeated actions that have already been described (for example, if the instructor stirs a solution multiple times, describe the first instance and note that it continues).
• The instructor's facial expressions unless they are reacting to a result.

Standard vs. extended

Extended mode is strongly recommended. Lab demonstrations are visually dense, and the observable results are often the entire learning objective. There is rarely enough silence in the narration to fit adequate descriptions of physical processes. Pausing the video to describe a chemical reaction, a tissue sample under a microscope, or the reading on an oscilloscope ensures the viewer receives the same information as sighted students.

Animated explainer

Animated explainers use motion graphics, illustrated characters, kinetic typography, or other animated elements to explain a concept. There is typically a voiceover narrating the concept while the animation reinforces or extends the spoken content. These are common in introductory courses, MOOCs, and corporate training.

What to describe

• Animated sequences: Describe what the animation shows, especially when it adds information beyond the voiceover. If the narrator says "cells divide" and the animation shows the specific stages of mitosis with labeled phases, describe the visual details the narration omits.
• On-screen text and labels: Read text callouts, labels, and titles that appear as overlays on the animation.
• Diagrams and data visualizations: Describe charts, graphs, flowcharts, and diagrams that appear, focusing on the data or relationships they illustrate.
• Transitions and scene changes: Note when the visual context shifts significantly, such as moving from a macro view to a microscopic view, or from one location to another.

What you can skip

• Decorative animation that does not convey new information (for example, background particles or abstract motion).
• Visual transitions like fades or wipes that do not change the meaning.
• Character movements that match exactly what the narrator is saying.

Standard vs. extended

Standard mode works well for most animated explainers. These videos are typically well-paced with natural gaps between concepts. Extended mode may be needed for animations that pack multiple visual elements on screen simultaneously, such as a detailed process diagram that builds over several seconds while the narrator discusses only part of it.

Interview and conversation

Interviews and panel discussions feature two or more people talking. The visual content is primarily the speakers themselves, their expressions, and any materials they reference. These videos appear in courses as guest lectures, expert discussions, and case study interviews.

What to describe

• Speaker identification: Identify who is speaking when the camera cuts between participants, especially at the start or when a new speaker joins. Include names and roles if displayed on screen.
• On-screen graphics: Describe lower thirds (name/title overlays), B-roll footage shown during the conversation, and any images, clips, or documents referenced.
• Significant non-verbal communication: Note meaningful gestures, emotional reactions, or body language that convey information beyond the words. For example, if one panelist visibly disagrees with a claim through a head shake or crossed arms while another is speaking, that is relevant.
• Physical demonstrations: If a speaker holds up an object, shows a document, or demonstrates something with their hands, describe what is shown.

What you can skip

• Routine camera switches between speakers during normal back-and-forth conversation.
• Nodding, smiling, and other conversational body language that reinforces but does not add to the spoken content.
• The physical appearance of speakers beyond the initial identification.

Standard vs. extended

Standard mode is almost always sufficient. Conversations have natural pauses between speakers, and the visual content is relatively sparse. Extended mode would only be needed if the conversation references complex visual materials shown on screen.

Direct to camera

Direct-to-camera videos feature a single speaker addressing the viewer, often in a simple studio setup or office. There is usually no screen share, no slides, and minimal visual complexity. These are common in course introductions, announcements, assignment explanations, and motivational or reflective content.

What to describe

• Setting and context: Briefly describe the setting at the start if it is relevant (for example, "the instructor is seated at a desk in an office with bookshelves behind them").
• On-screen text: Describe any titles, captions, or graphic overlays that appear.
• Props or materials: If the speaker holds up a book, shows a document, or references a physical object, describe what is shown.
• Significant changes: Note if the setting, lighting, or framing changes mid-video, or if another person enters the frame.

What you can skip

• The speaker's appearance beyond a brief initial description.
• Minor movements, gestures, and expressions during routine speech.
• Background details that do not change or contribute to the content.

Standard vs. extended

Standard mode is sufficient for nearly all direct-to-camera content. These videos tend to have the least visual complexity of any format, and the speaker usually provides enough verbal context on their own. Descriptions are typically brief and infrequent.

Documentary style

Documentary-style videos combine narration with footage of real locations, people, events, and archival materials. They are used in history, social science, environmental science, journalism, and cultural studies courses. The visual content is rich, varied, and central to the meaning.

What to describe

• Locations and settings: Describe where the footage was captured, including landscapes, buildings, interiors, and environmental conditions.
• People and activities: Describe who is on screen and what they are doing, especially when the narration discusses a concept and the footage illustrates it with specific examples.
• Archival materials: Describe photographs, historical footage, documents, maps, and artifacts shown on screen. Include any visible text, dates, or labels.
• Data and graphics: Describe maps with highlighted regions, timelines, statistical overlays, and other informational graphics.
• B-roll footage: When the narration plays over supporting footage, describe what the footage shows. This is often where the visual content diverges most from the audio.

What you can skip

• Establishing shots that are purely atmospheric when the narrator already describes the location.
• Repeated footage used as filler between segments.
• Decorative transitions.

Standard vs. extended

Standard mode works for most documentary content if the narration has regular pauses. Extended mode is valuable for segments with dense visual information, such as a sequence showing multiple archival photographs in quick succession, or a scene where the visual content tells a different story than the narration.

Split screen and picture-in-picture

Split screen and picture-in-picture layouts show multiple video feeds simultaneously. Common examples include a speaker alongside their slides, two participants in a remote conversation shown side by side, or a software demo with a webcam overlay. The challenge is that multiple visual streams compete for the listener's attention.

What to describe

• Layout orientation: At the start, describe the layout: "the screen is split vertically, with the instructor on the left and the slide deck on the right," or "the instructor's webcam appears in a small window in the lower right corner over the screen recording."
• Content priority: Focus on whichever feed carries the most information at any given moment. If the slides are the focus, describe the slides. If the speaker is demonstrating something physically, describe that instead.
• Layout changes: Note when the layout changes, such as one feed going full screen or a new feed being added.
• Feed-specific content: Apply the description techniques for the underlying format of each feed. If one feed is a screencast and the other is a webcam, use the screencast guidelines for that portion and the direct-to-camera guidelines for the other.

What you can skip

• Trying to describe both feeds simultaneously. Prioritize and alternate based on where the meaningful content is.
• The static presence of a webcam overlay when the speaker is simply talking and the slides carry the content.

Standard vs. extended

Depends entirely on the content in each feed. If one feed is a dense screencast or data visualization, extended mode may be needed. If both feeds are relatively simple (a speaker and basic slides), standard mode works fine.

Khan Academy style (digital drawing with voiceover)

This format features a digital canvas where an instructor draws, writes, and annotates in real time while narrating. The visual content builds progressively, similar to a whiteboard, but with digital tools that allow for color, layering, and precise graphics. It is common in math, science, and economics courses.

What to describe

• Drawn elements: Describe figures, graphs, equations, and diagrams as they are drawn. Focus on structure and relationships: "a right triangle with the hypotenuse labeled c and the two legs labeled a and b."
• Color and highlighting: Note color changes used to distinguish elements, such as "the supply curve is drawn in blue and the demand curve in red."
• Progressive construction: Describe how the visual builds over time. When a graph is extended, an equation is modified, or a new element is added to an existing diagram, connect the new content to what is already on screen.
• Annotations and emphasis: Describe underlines, circles, arrows, and other marks used to draw attention to specific elements.

What you can skip

• The drawing tool itself (pen, highlighter, eraser) unless the tool change signals something meaningful.
• The physical act of drawing when the result is what matters.
• Elements the instructor describes verbally as they draw.

Standard vs. extended

Extended mode is often necessary. The instructor typically narrates continuously while drawing, leaving few pauses. The visual content is also cumulative and mathematical, requiring careful description that takes more time than brief pauses allow. For simple conceptual sketches, standard mode may suffice.

Green screen and studio production

Green screen videos place the instructor in front of a digitally composited background, which might be a still image, a video, an animation, or an interactive environment. Studio productions may include multiple camera angles, professional lighting, and post-production effects. These formats are common in high-production-value course content and broadcast-style educational media.

What to describe

• Virtual environment: Describe the composited background and any changes to it. If the background shifts from a cityscape to a laboratory, note the change.
• Integrated graphics: Describe charts, diagrams, or images that are composited into the scene alongside the instructor.
• Camera angles and framing: If the production uses multiple camera angles, note significant changes, such as cutting from a wide shot to a close-up of a detail.
• On-screen effects: Describe text overlays, animated lower thirds, transition effects, or other post-production elements that convey information.

What you can skip

• The fact that a green screen is being used (the viewer does not need to know the production technique, only what appears on screen).
• Camera angle changes that do not alter what information is visible.
• Atmospheric effects that are decorative rather than informational.

Standard vs. extended

Standard mode is usually sufficient unless the composited environment contains detailed visual information that needs time to describe. If the background is a complex data visualization or an annotated 3D model that the instructor references, extended mode may be appropriate.

Hybrid lecture (in-person and remote)

Hybrid lectures combine an in-person classroom recording with remote participants shown on a screen or integrated into the video feed. Some students are physically present, while others attend via video conference. The recording captures both the room and the remote feeds.

What to describe

• Room layout: Briefly orient the listener at the start: "the instructor is at the front of a lecture hall with approximately 30 students seated in rows; a large screen behind the instructor shows a grid of remote participants."
• Remote participant interactions: When a remote participant speaks, identify them if possible and note if their video feed becomes prominent on screen.
• Shared content: Describe any content projected in the room, following the guidelines for whichever format the projected content takes (slides, demonstration, etc.).
• Physical interactions: If the instructor moves to interact with in-room students, equipment, or the whiteboard, describe the action.

What you can skip

• The faces of remote participants who are not actively contributing.
• Routine technical interactions like muting, unmuting, or adjusting the camera.
• The classroom environment after the initial description unless it changes meaningfully.

Standard vs. extended

Standard mode is typically sufficient. Hybrid lectures have natural pacing similar to in-person lectures. Extended mode may be needed when the instructor switches rapidly between in-room activities and projected content.

Live webinar recordings

Recorded webinars feature a speaker or panel presenting to a remote audience, often with slides, screen sharing, polls, and Q&A interactions. The visual content combines the presenter's feed with shared materials and audience interaction elements.

What to describe

• Shared content: Apply the appropriate format guidelines for whatever is being screen-shared (slides, software demo, document, etc.).
• Presenter video: Describe the presenter's video feed briefly when it first appears and when they demonstrate something physical.
• Interactive elements: Describe poll results, chat highlights, or Q&A panels that appear on screen if they contribute to the content.
• Transitions between segments: Note when the presentation shifts from slides to a live demo, or from a presentation to an audience Q&A.

What you can skip

• The webinar platform interface elements (toolbar, participant list, chat sidebar) unless they are the focus of the content.
• The presenter's video when it is a small overlay and the shared content is the focus.
• Technical moments like "let me share my screen" if the presenter narrates them.

Standard vs. extended

Standard mode works for most webinar recordings. The pacing is typically conversational with natural pauses. Extended mode may be needed during dense screen-sharing segments.

Choosing standard vs. extended: a quick reference

Video format	Recommended mode	Reason
Lecture capture	Standard (usually)	Natural pauses between slides and ideas
Screencast	Extended (usually)	Continuous narration; fast-changing visuals
Talking head with slides	Standard	Conversational pacing with slide pauses
Whiteboard / lightboard	Extended	Continuous drawing and talking simultaneously
Lab demonstration	Extended	Dense visual information; safety-critical details
Animated explainer	Standard (usually)	Well-paced with conceptual gaps
Interview / conversation	Standard	Natural turn-taking pauses
Direct to camera	Standard	Minimal visual content
Documentary style	Standard or extended	Depends on density of visual material
Split screen / PiP	Varies	Depends on content in each feed
Khan Academy style	Extended (usually)	Cumulative visual content with continuous narration
Green screen / studio	Standard (usually)	Unless composited visuals are dense
Hybrid lecture	Standard	Natural lecture pacing
Live webinar recording	Standard (usually)	Conversational pacing

For a detailed comparison of standard and extended audio descriptions, including when to use each and how they work technically, see our guide on AI vs. human audio description workflows.

General principles across all formats

Regardless of the video format, a few principles apply to every audio description:

1. Describe what matters, not everything visible. The goal is equivalent access to the information, not a comprehensive visual inventory. Prioritize content that a sighted viewer would use to understand the lesson.

2. Follow the instructor's focus. If the instructor is directing attention to a specific element, that element should be described. If they do not reference something visually, it is usually lower priority.

3. Use precise, neutral language. Say "a bar chart showing enrollment increasing from 500 to 1,200 between 2020 and 2025" rather than "a nice graph." Describe what is shown, not your interpretation of it.

4. Do not duplicate the audio. If the instructor reads a slide aloud, you do not need to describe it again. Focus description time on visual content that is not covered in the spoken narration.

5. Orient the viewer. When the visual context changes significantly, briefly establish the new setting or layout before describing specific elements.

6. Be consistent. Use the same terminology throughout a video or series. If you call something a "flow diagram" in minute two, do not call it a "process chart" in minute eight.

Try it with your own content

We are actively building a library of audio description examples across every format listed in this guide. If you have a video that fits any of these categories that you'd like to share with us, submit it through our intake form and we will generate a free AI audio description for it. Videos must be five minutes or shorter, and we may feature the described version on our site as an example.

You can also explore our AI audio description generator to learn more about how Recap processes videos at scale, or read our technical guide on how the AI technology works.