LiveLink
LiveLink Lite User Guide
A practical guide to creating projects, composing live scenes, and running real-time captions for presentations.
What LiveLink Lite Does
LiveLink Lite helps you run a presentation with live video, shared screens, and real-time captions in one place. You can show a presenter camera and slides or another screen source, then add multilingual captions, translated slide text, names or titles, backgrounds, drawing, recording, and output windows as needed.
Assigned microphones feed speech-to-text, and the app turns speech into captions, translates them when needed, and places them on the live scene. The same project can record video, archive transcripts, show a caption-only window, and provide a separate output window so operators can monitor what the audience will see.
A standout feature is the built-in virtual camera: LiveLink Lite can present its finished scene — video and live captions together — as a webcam that Zoom, Microsoft Teams, OBS, and other apps can pick from their camera list. Using LiveLink Lite as the camera in a Zoom or Teams meeting is one of the main ways the app is used.
Contents
-
Editor Panel
-
Session Library
Important Concepts
LiveLink Lite has many controls because it combines presentation layout, live audio, captions, translation, recording, and output routing. These concepts make the rest of the app easier to understand.
Quick Start
01
02
03
04
05
06
Choose a project. Open LiveLink Lite and select an existing project from the Welcome page, or start a new project with the setup wizard.
Select inputs. Pick the Actor camera and Screen source, then confirm that the preview shows the right content.
Assign audio. Choose the microphones or audio devices for each speaker group and check the audio meter while someone speaks.
Choose caption routing. Select a caption preset or wizard template that matches the event: same-language captions, translated captions, multiple speakers, or bilingual room.
Prepare scenes. Review scene buttons 1-4 and confirm the layouts you plan to use during the presentation.
Go live. Switch to Present mode, enable speech-to-text, open any needed output windows, and start recording if the session should be archived.
The wizard only sets a good starting point. Almost every setting can be changed later in Edit mode or from the toolbar.
Before captions can work, confirm the required services are ready: a valid license, Google sign-in for Google speech-to-text or translation, and any API keys/accounts for services you plan to use, such as ElevenLabs or DeepL.
Real-Time Captions
Real-time captions are the main feature of LiveLink Lite. The app can show up to four caption outputs on screen at the same time, and each output can use a different language. This makes it possible to run events such as same-language captioning, live translation for one audience, bilingual rooms, or multiple caption streams for different viewers.
Captions begin with microphone audio. Speech-to-text listens to the assigned speaker groups, produces live text, the caption routing system decides which outputs should receive that text, and translation is applied when an output language differs from the speaker language. The final captions are shown in the live scene and can also appear in the separate caption output window.
How spoken audio becomes on-screen captions. Translation only runs when an output language differs from the speaker.
Caption building block
Speaker group
What it means
The person or audio source being captioned. A speaker group has one or more microphones or audio devices and one or more spoken languages.
Speech-to-text session
The live transcription process for a speaker group and language. The app can run up to four active speech-to-text sessions at once.
Caption output
One visible caption stream on the screen. LiveLink Lite supports up to four caption outputs, each with its own language, color, and style.
Caption preset
The saved routing setup that connects speaker groups to caption outputs. Presets decide which languages appear and whether translation is used.
Common Caption Setups
-
One speaker, same-language captions: one microphone produces one caption output in the speaker language.
-
One speaker, translated captions: one microphone produces captions in a different audience language.
-
Original plus translation: one speaker appears in both the spoken language and a translated language at the same time.
-
Two speakers or bilingual room: each speaker group can have its own microphone, spoken language, and caption routing.
-
Up to four caption outputs: use multiple outputs when different audiences need different languages or when original and translated captions should remain visible together.
Caption Presets and Routing
A caption preset is the most important caption setup item. It tells the app which speaker groups exist, which microphones belong to them, what language each speaker uses, and which caption outputs should receive the text. The wizard provides common templates so most projects do not need manual routing.
When you do need manual control, two routing tools are available during setup or live operation. The routing graph lets you drag connections between speakers and outputs, including holding Ctrl to connect several outputs at once. A compact connection matrix offers the same control with keyboard toggles and no dragging. Use either when the templates are not enough — for unusual events, shared microphones, multiple outputs, or cases where one speaker should feed several caption languages. For a labeled screenshot and a step-by-step walkthrough, see the Caption Routing Editor in the Interface Reference.
Speech-to-Text and Caption Stability
Speech-to-text provider and model choices are available from the title bar. They affect supported languages, latency, interim caption behavior, and recognition quality. Google models and ElevenLabs Scribe are intended for live transcription.
Some providers show interim text while the speaker is still talking. Interim captions are useful because they feel immediate, but they can also change as the provider revises the sentence. Caption stabilization helps reduce distracting rewrites before text is committed visually.
Word Hints help with names, technical terms, product names, and event-specific vocabulary that recognizers often miss. Open Word Hints from the Speech to Text menu and add terms manually, or let the app extract likely terms for you. Supplying hints before an event noticeably improves accuracy on specialized content.
Caption Translation
Translation is used when the caption output language is different from the speaker language. The translation engine choice affects speed, language coverage, and wording quality. Faster engines are usually better for rapid conversation. Higher-quality engines may be better for prepared presentations, slower speech, or events where phrasing matters more than minimum latency.
The Region setting (Options → Region) is a process-wide hint for cloud services. Google speech-to-text and Google Translate honor it directly; Gemini maps to its nearest available region; DeepL and ElevenLabs are single-host services and ignore it; and Google Translate's high-quality LLM model always runs from one fixed region regardless of the setting.
Caption Appearance and Placement
Caption routing decides what text appears; caption appearance decides how it looks. You can adjust the caption region, font size, text color, outline, background box, spacing, layout, and timeout behavior. For live use, choose strong contrast and keep each caption region large enough for comfortable reading.
An optional scrolling (teleprompter) mode anchors captions to the bottom of their region and slides older lines upward as new lines arrive, with an adjustable per-line scroll speed. Leave it off for a fixed caption block, or turn it on when you want a continuous, rolling-transcript feel.
When using multiple caption outputs, make sure the layout leaves enough room for each language. A readable two-language layout is usually better than four cramped caption regions.
Caption Setup Checklist
-
Choose or create a caption preset that matches the event.
-
Assign the correct microphone or audio device to each speaker group.
-
Confirm each speaker language and each caption output language.
-
Pick the speech-to-text provider/model and translation engine before the event starts.
-
Speak into each microphone and verify that the audio meter moves.
-
Enable speech-to-text and confirm each expected caption output appears in the preview.
-
Check caption contrast, size, and placement in all scenes you plan to use.
If captions are enabled but nothing appears, check audio first: microphone assignment, meter activity, speech-to-text enabled
state, sign-in/API credentials, and license requirements.
Microphones
1–4 speaker groups
Speech-to-Text
Google / ElevenLabs
Caption routing
preset decides outputs
Translation
when languages differ
On-screen
up to 4 caption
Welcome Page
The Welcome page is the first screen you use to choose or create a project. It also includes account and license controls needed for cloud speech-to-text, translation, and licensed features.
Area or Button
What it does
Google Sign In / Sign Out
License
Connects the app to Google services used for speech-to-text, translation, and related cloud features.
Audio
Opens the license dialog. Some features, such as OCR and speech-to-text controls, require a valid license.
Projects
Browse...
Shows the currently configured speaker groups and microphones. Use this area to confirm that each speaker has the correct input device before starting.
Project actions menu
Caption Presets
New Project
Start Presenting
Start Editing
Lists recent projects. Click a project to select it, or double-click to start presenting with it.
Loads a project file that is not already in the recent project list.
Use the menu on a project card to edit the project note, open its folder, remove it from the recent list, or delete the file.
Selects the routing preset that controls which microphone languages feed which caption outputs.
Starts the setup wizard for a new project.
Loads the selected project in Present mode, with editing panels hidden and live controls available.
Loads the selected project in Edit mode so you can adjust visual layout and settings before going live.
New Project Wizard
The wizard is the easiest way to build a new project. It walks through five steps in order, shows a live preview beside each step, and fills in a safe default for every choice, so you can move quickly and refine anything later in Edit mode.
Step
1. Sources
Main choices
Pick the Camera / Actor source and the Slides / Screen source. On the camera you can also remove the background and flip (mirror) the image; on the screen you can turn on Use OCR with translation to translate text found in the capture (a gear button opens its settings; requires a license).
2. Scene Layout
Set up each of the four scenes on its own tab. For a scene, pick a ready-made layout from the template grid (picture-in-picture, presenter, side-by-side split, sidebar camera, full screen, full camera, and more), or drag and resize the Screen and Actor directly in the preview. Each scene can also set its fill, border, rounded or circle / oval crop, size, and position.
3. Captions & Languages
Start from a caption preset — Single Translation, Bilingual, Four Languages, or Two Speakers — then set each speaker's device and language and each caption output's language. Choose Custom to design your own routing in the full node editor.
4. Captions & Background
Style the captions and choose the background side by side. Captions: pick a ready-made color style, then adjust font, size, color, outline, and background box, and drag the caption region in the preview. Background: a solid color, an image, or a generated animated effect.
5. Review & Start
Move through the steps with Next → and ← Back. Because every step has sensible defaults, Skip to Review jumps straight to the last step at any time. On the first step the Back button reads Cancel and closes the wizard.
The project name is set on the final Review & Start step, and title / lower-third text overlays are no longer part of the wizard — add them anytime from the Text tab in Edit mode.
StartName the project, optionally add notes and choose a save folder, check the summary of your sources, scenes, captions, and background, then click Start Presenting.
Present Mode vs Edit Mode
LiveLink Lite separates live operation from detailed editing so the screen stays clean during a presentation.
Mode
Use it when
Present mode
You are running the session live. Editing panels are hidden, scene buttons are ready, output windows can be shown, speech-to-text can run, and telestrator drawing is available.
Edit mode
You need to adjust layout, background, screen and actor properties, caption placement, caption styling, or text overlays. This is the mode for preparing the project before the event.
More Features
Beyond live captioning, LiveLink Lite bundles the production tools you need to build and run a complete scene.
Audio and Microphones
Audio setup decides what speech-to-text hears. Assign microphones or audio devices to speaker groups, then confirm that the audio meter moves when each person speaks. A speaker group can use one device or multiple devices mixed together.
The app can keep track of offline devices stored in a project, so a missing microphone can reappear when Windows detects it again. Use gain, mute, and noise reduction carefully: they can improve caption quality, but an incorrectly muted or silent device will prevent captions from appearing. Noise reduction can use the built-in DeepFilterNet AI filter or, on supported NVIDIA RTX systems, NVIDIA Broadcast.
Scene Composition
The scene is everything the audience sees: background, Screen layer, Actor layer, captions, text overlays, and drawing. Each project has four scene buttons, and each scene stores its own layout and visibility choices. This lets you prepare common looks before the event, such as full-screen slides, actor-only, split view, picture-in-picture, or a caption-focused layout.
Front
Telestrator
Live freehand drawing
Text overlays
Titles & lower-thirds
Captions
Up to 4 language output
Actor
Camera or presenter
Screen
Slides, desktop or window
Background
Color, image or animation
Back
Scene layers stack from back to front. Each scene button stores its own layout and which layers are visible.
-
Background: choose a solid color, an image, or a generated animated effect.
-
Screen: show slides, a desktop, a captured window, or another visual source, with optional borders and rounded or circular cropping.
-
Actor: show a camera or presenter feed, with optional background removal, a horizontal flip (mirror), and the same border and rounded/circle framing.
-
Captions: place one or more caption outputs where they remain readable over the scene.
-
Text overlays: add titles, speaker names, lower-thirds, or other fixed text.
OCR Slide Text Translation
OCR reads text visible in the Screen source and can translate it separately from spoken captions. This is useful for slide decks, captured documents, browser windows, or shared screens that contain text the audience may not understand.
OCR is best treated as a slide and screen aid, not as a replacement for speech captions. It may take extra processing time and works best when the source text is stable, readable, and not too small.
Recording and Transcript Sessions
Recording and transcript sessions serve different purposes. Media recording captures the session output according to recording settings. Transcript recording archives the caption/session text so it can be reviewed later.
The Session Library lets you browse previous sessions by date, open transcript artifacts in tabs, search across recorded content, and review AI summaries when summary generation is enabled. A live session can also appear while recording is active, then become an archived session when recording finalizes.
Output Windows
Output windows make it easier to monitor or share what LiveLink Lite is producing. The caption output window shows captions independently from the main scene. The extra render output window shows the composed scene in a separate window.
These windows are useful for confidence monitoring, second-screen display, or situations where another tool captures a separate application window.
Virtual Camera (Use in Zoom, Teams, and OBS)
LiveLink Lite installs a virtual camera so its finished scene — the composed video with live captions included — can be used as an ordinary webcam in other applications. This is the primary way to bring LiveLink Lite into a video call: run your presentation here, and let Zoom, Teams, or your streaming tool show it as the camera.
In the other application's camera or video settings, choose “LiveLinkLite Output 1080p” from the camera list. It works in Zoom, Microsoft Teams, OBS Studio, Google Meet, and most other apps that accept a webcam, and several apps can use it at the same time.
-
Automatic: the camera is registered when LiveLink Lite is installed, and it feeds live video whenever LiveLink Lite is running and an app is viewing it — there is no separate switch to turn on.
-
Output format: always 1920×1080 (Full HD), at 30 or 60 fps following the render frame rate you choose in the Options menu.
-
If the picture is missing or frozen: make sure LiveLink Lite is open and showing a scene, then reselect the camera in the other app; some apps must be restarted to detect a newly installed camera.
Telestrator Drawing
The telestrator lets the presenter draw temporary annotations over the live output. It is useful for pointing at slide details, highlighting an area, or marking motion during a presentation. Strokes fade out automatically by default, keeping the scene clean without manual erasing. Hold Ctrl while drawing (or just after) to freeze a stroke on screen for as long as you need, then release to let it fade.
Hover over the toolbar telestrator button for quick favorite colors, or click it for full drawing settings such as enable/disable, color, width, fade duration, smoothing, and clear.
Main Toolbar
The toolbar is used during both editing and presenting. It gives quick access to the live controls you are most likely to change during a session.
Control
Present / Edit
Use it for
Toggles between Present mode and Edit mode. Present mode is for live operation; Edit mode opens the editing controls.
Recording
Starts or stops session recording and opens recording settings.
Scene buttons 1-4
Switches between saved scene layouts. Use these for quick changes such as screen-only, actor-only, split view, or picture-in-picture.
Actor
Selects the camera or video source for the Actor layer. The popup also includes actor background removal.
Screen
Selects the slide, desktop, window, or file source for the Screen layer.
OCR
Turns slide text detection/translation on or off and opens OCR options. OCR requires a valid license.
Captions
Opens caption preset selection so you can change language routing during setup or live operation.
Speech-to-text button
Enables or disables speech-to-text. The button also shows whether caption sessions are active. Speech-to-text requires a valid license.
Audio devices
Opens audio capture device configuration. Use this when assigning microphones to speakers.
Audio meter
Shows mixed audio level and provides quick gain adjustment.
Telestrator
Controls live drawing. Hover for favorite colors; click for full drawing settings.
Output windows
Opens or closes caption and render output windows used for monitoring or external display.
For a labeled screenshot and a control-by-control breakdown including indicator colors and license requirements, see Toolbar in the Interface Reference.
Title Bar Menus
The title bar contains application-wide menus and the User Guide button. These are less frequently changed during a live presentation, but they are important for setup, maintenance, and troubleshooting.
The title bar also carries quick-access icon buttons (web-streaming toggle, license, Word Hints, Welcome page, Session Library, theme, and language). For a labeled screenshot and every menu item with its options and trade-offs, see Title Bar in the Interface Reference.
Troubleshooting Basics
1 / No captions appear
Confirm speech-to-text is enabled, the correct microphone is assigned, the audio meter moves, the caption preset has a valid route, and required license/provider/API settings are ready.
2 / Captions use the wrong language
Check the speaker language in the caption preset, the output language, and the selected speech-to-text provider/model.
3 / Translation feels slow
Try a faster translation engine, or simplify the routing if many outputs are translating at once.
4 / Screen or actor source is missing
Open the Actor or Screen toolbar popup, reselect the source, and confirm the source window or device is still available in Windows.
5 / OCR does not translate slide text
Confirm OCR is enabled, the Screen source is visible, the text is readable, and the license requirements are satisfied.
Before Going Live
-
Confirm the correct project and caption preset are selected.
-
Check that the license is active (it covers every cloud service).
-
Verify Actor and Screen sources in the preview.
-
Speak into each microphone and watch the audio meter.
-
Enable speech-to-text and confirm captions appear in the preview.
-
Switch through scene buttons 1-4 to confirm each layout looks correct.
-
Open any required output windows before the event begins.
-
If using Zoom, Teams, OBS, Google Meet, or another video app, select “LiveLinkLite Output 1080p” as the camera there and confirm it shows the finished scene with captions.
Interface Reference
This section is a detailed, control-by-control companion to the high-level overviews above. It walks the title bar, the main toolbar, the editor panel, and the Session Library in the order the controls appear on screen. Use it as a lookup reference when you need to know exactly what a specific button or menu item does.
Title Bar
The title bar runs across the very top of the window. The left side shows the application icon and the current project title — drag it to move the window, or double-click it to maximize and restore. While a transcript session is recording, a blinking red REC indicator appears next to the menus. The right side holds the menus, quick-access icon buttons, the language toggle, and the standard window controls.
Control
What it does
1
Web Streaming toggle (satellite icon)
Leftmost icon button. Starts or stops live caption web streaming with one click — the same as Web Streaming → Start/Stop. While streaming it brightens and shows a blinking green “on-air” dot with a green underline. Requires a valid license.
2
Web Streaming
Menu: Start/Stop Live Stream Captions; Copy Captions URL (silent — copies the audience link to the clipboard); Show QR Code (the join QR-code window); Render QR Code (the in-scene QR overlay). See the menu notes below the table.
3
Transcript Sessions
Menu: Open Session Library; Enable Recording (archive caption text to disk); Open Data Folder; Set Data Folder…; Enable AI Summary; Summary Language.
4
Project
Menu: Save As…; Save; Load…; App Data Folder; Clear Settings… (resets stored settings after a confirmation).
5
Speech to Text
Menu: Word Hints…; LLM Autocorrect; and the recognizer choice — Google latest_long, Google chirp_3, or ElevenLabs Scribe 2 (see notes below the table).
6
Translation
Menu (pick one engine): Gemini Flash 3.1 LLM, Google Translate (LLM), Google Translate (NMT), DeepL (Fast), DeepL (Accurate).
7
? (User Guide)
Opens this guide in your default browser.
8
Options
Menu: Caption Output Window; Extra Render Window; Render 60 fps / Render 30 fps (disabled while recording); Show Logs.
9
License (key icon)
Opens the license activation dialog. A valid license is required for Speech-to-Text and OCR.
10
Word Hints (clipboard icon)
Quick shortcut to the Word Hints editor (same as Speech to Text → Word Hints).
11
Session Library (archive icon)
Temporarily shows the Welcome page so you can switch project or caption preset without restarting.
12
Welcome page (disk icon)
Opens or closes the Session Library page (same as Transcript Sessions → Open Session Library).
13
Theme (sun/moon icon)
Toggles the application between light and dark mode.
14
14Language (ENG / 한국어)
Menu: Caption Output Window; Extra Render Window; Render 60 fps / Render 30 fps (disabled while recording); Show Logs.
15
Window controls
Minimize, maximize/restore, and close the window.
Web Streaming menu
Web Streaming publishes your live captions to a cloud relay so audience members can read them on their own phones or laptops — useful for accessibility, overflow rooms, or remote viewers. The leftmost satellite icon button ( ) starts and stops streaming with a single click and mirrors the menu's start/stop item.
-
Live Stream Captions — starts or stops streaming. When it starts, a session is created on the relay and a join link (and QR code) become available. Only captions spoken after you start are sent; nothing said earlier is shared. Requires a valid license.
-
Copy Captions URL — copies the audience link to the clipboard. This action is silent — there is no on-screen confirmation. Paste the link into chat, email, or a slide so viewers can open it.
-
Show QR Code — shows or hides a separate window with a large scan-to-join QR code for the audience. This choice is remembered between runs, and the window can be shown even before streaming has started so you can present it while people get ready.
-
Render QR Code — overlays the join QR code directly inside the broadcast scene (and any output windows), so the code appears in your composed video without a separate window.
1
The audience opens a lightweight web page that follows the transcript live and lets each viewer pick which output language to read, so one stream can serve several languages at once.
Transcript Sessions menu
-
Open Session Library — opens (or closes) the Session Library page; the same as the archive icon ( ).
-
Enable Recording — a checkbox that turns transcript archiving on or off. While it is on, each caption session is written to disk in the data folder and the blinking red REC indicator appears in the title bar. This records the caption text, not video — media capture is the separate Recording control on the toolbar ( ).
-
Open Data Folder — opens the current transcript data folder in Windows Explorer. Disabled until a data folder has been set.
-
Set Data Folder… — chooses the folder where transcript sessions (and their AI summaries) are stored.
-
Enable AI Summary — a checkbox that allows AI summaries to be generated for recorded sessions. When it is off, the Generate Summary action in the Session Library is unavailable.
-
Summary Language: … — opens a searchable language picker that sets the language AI summaries are written in. The current choice is shown in the menu item itself.
Project menu
-
Save As… — saves the current project to a new file you choose.
-
Save — saves to the project's existing file. If the project has never been saved, this behaves like Save As…
-
Load… — opens an existing project file.
-
App Data Folder — opens the application's data folder (settings, logs, and supporting files) in Windows Explorer.
-
Clear Settings… — resets stored application settings to their defaults after a confirmation prompt. This affects app-wide settings, not your saved project files.
Recognizer choices (Speech to Text)
-
Google latest_long — emits interim (partial) text as the speaker talks for low-latency captions; finals follow when confirmed.
-
Google chirp_3 — finals only, no interim text; better accuracy for prepared speech.
-
ElevenLabs Scribe 2 — interim text plus finals, streamed through the DenShow cloud. Falls back to Google for languages it does not support.
Translation engines
-
Gemini Flash 3.1 LLM — highest quality and most stable phrasing, highest latency. With an interim STT engine it is effectively used for finalized text.
-
Google Translate (LLM) — higher quality than NMT, moderate latency.
-
Google Translate (NMT) — lowest latency and quality.
-
DeepL (Fast) — classic NMT, lowest DeepL latency, strong on European languages. Unsupported pairs fall back to Google.
-
DeepL (Accurate) — next-generation model, higher quality and markedly higher latency. Same fallback to Google for unsupported pairs.
Options menu
-
Caption Output Window — a checkbox that shows or hides the separate caption-only window (also toggled from the toolbar's output controls, 27).
-
Extra Render Window — a checkbox that shows or hides the second window displaying the full composed scene (also the toolbar button 28).
-
Render 60 fps / Render 30 fps — sets the render and output frame rate. The two act as a pair, so one is always active. Both are disabled while a recording is in progress so the recorder's timing is never changed mid-session.
-
Show Logs — opens the in-app log viewer, useful when diagnosing caption, audio, license, or cloud-service problems.
Word Hints dialog
Open Word Hints from Speech to Text → Word Hints… or the clipboard icon (10). Word hints bias speech recognition toward words it commonly mishears — names, places, product names, and technical jargon. Adding them before an event noticeably improves accuracy on specialized vocabulary.
-
Language tabs — hints are kept separately for each speaker language. Every configured speaker language gets a tab, and any language that already has saved hints is listed too.
-
Hint list — enter one single-word hint per line, up to 20 per language. A live counter shows the number of hints and an estimated token cost, and flags any entries that need fixing.
-
Extract hints from text — opens an assistant where you paste prepared notes or load a document (.txt, .md, .srt, .pdf, .docx, and more). It detects the language automatically and proposes single-word hints for names, places, and jargon, which are merged into the matching language tab for you to review.
-
Clear Text — empties the current language. When several languages are present this removes that language entirely; when only one remains it just clears the text.
-
Apply / Cancel — Apply validates and saves all changes (it shows a "Fix Word Hint Issues" list if anything is invalid); Cancel closes without saving.
12
17
Hints take effect on the next recognition session, not the current one. Set or edit them before the speaker starts, or briefly turn speech-to-text off and on to apply them.
License Activation dialog
Open it from the key icon ( ) or the License button on the Welcome page. A valid license is required for Speech-to-Text and OCR; without one, those features stay disabled.
-
Status — shows whether the product is activated. When active, it also lists the licensed name, email, and expiry date (or "Never" for a perpetual license).
-
Activate — enter the account name and password from your purchase welcome email and click Activate. The machine receives its own activation credential; your password is never stored. An internet connection is required.
-
Account snapshot — once active, the dialog shows your license expiry, seats in use, activated machines, and the remaining credit balance, refreshed live while the app runs.
-
Deactivate — enter the same account credentials to release this machine's activation slot so the license can move to another computer.
9
ToolBar
The main toolbar sits just below the title bar and is available in both Edit and Present modes. Controls are listed left to right.
The main toolbar, callouts 16–28. The numbers match the table below.
Control
What it does & indicators
16
Present / Edit
Toggles between Present mode (clean, live) and Edit mode (editor panel visible).
17
Recording
Record button, elapsed-time counter, and a settings gear. Starting prompts for a file name first; the control shows a busy state while a recording is being finalized.
18
Scene buttons 1–4
Switch the active scene layout. Each button is a live mini-preview of that scene's Screen/Actor/Caption rectangles and visibility; the active scene is highlighted.
19
Actor
Picks the Actor (camera/video) source. The popup has a Background Removal toggle on top; hovering a device shows a detail popup listing its available capture formats.
20
Screen
Picks the Screen source — a monitor, an application window, or a media file.
21
OCR
Enables on-screen text detection/translation. Requires a valid license. The button pulses green while OCR is recognizing or translating; hovering opens OCR options, and the gear opens the OCR appearance settings.
22
Captions
Opens the Caption Routing Editor to change speaker→caption language routing during setup or live operation.
23
Speech-to-Text
Enables or disables live transcription. Requires a valid license. Pulses green while a session is active and shows a blue border when enabled; a red diagonal line warns that you are not signed in to Google, so transcription will not work.
24
Audio devices
Opens audio capture configuration, where microphones are assigned to speaker groups.
25
Audio meter
Shows the mixed audio level and provides a quick gain adjustment.
26
Telestrator
The dot previews the current stroke color and width. Hover for the quick favorite-color popup; click for full drawing settings (enable, clear, color, width, fade, smoothing).
27
Output windows
windowsQuick toggles for the optional output windows (caption output and extra render), mirroring the matching items in the Options menu.
28
Extra Output Window
Square button at the right edge that shows or hides the separate render-output window.
Connect a speaker to a caption
-
Find the speaker on the left and press the mouse on its output port — the dot on its right edge (6).
-
Drag toward the caption output on the right; a curved preview line follows the cursor.
-
Release over the caption's input port (8). A link is created.
-
If the speaker and caption languages are different, that caption is translated automatically — no extra step is needed.
Faster ways to edit links
-
Connect several at once (Ctrl+drag): hold Ctrl while dragging from a speaker port; each caption links as the cursor reaches it, without releasing the mouse. A small "+" appears by the cursor while Ctrl is held.
-
Move a link: drag from an already-connected speaker port onto a different caption to move the route there.
-
Remove one link: click it to select, then press Delete (or use its inline delete button); or drag a connected port to empty space and release.Remove every link on a caption: drag from the caption's input port away from it and release past a short distance to clear all speakers feeding that output.
-
Keyboard: Tab / Shift+Tab and the arrow keys cycle the selected link, Delete removes it, and Esc clears the selection or cancels a drag in progress.
The four-session limit (2) counts speaker languages, not connections. One speaker language can feed many caption outputs at no extra cost, so to show the same talk in several languages, connect one speaker to several outputs rather than duplicating the speaker.
Session Library
Open the Session Library from Transcript Sessions → Open Session Library (or the archive icon in the title bar). It indexes the finalized transcript sessions stored in your data folder and presents them in a searchable, split-panel workspace.
Top bar
-
Folder field — shows the active library folder.
-
Browse — choose a different library folder.
-
Open — open the library folder in Windows Explorer.
-
Find (Ctrl+F) — open the search popup.
-
Close — return to the main window.
Filtering
A date calendar offers quick presets and a custom range. Videos only / All sessions limits the list to sessions that have a recorded video file, and Live today jumps to the current live session.
Find popup
Search is explicit: type, then press Enter or click Find — it does not search as you type. The scope toggles are Everywhere (notes, video filenames, and transcript file contents — the slowest, since it opens files on disk), Notes, and Video filenames (both in-memory). A counter shows match n / N; the ▲/▼ buttons (or Shift+Enter / Enter) step through matches, opening one session at a time; Esc closes the popup.
Workspace
The split view shows the session tree on the left (grouped by date; a live session is pinned at the top while recording) and a tabbed document workspace on the right for viewing and editing transcript artifacts.
Per-session actions (right-click)
-
Generate Summary — creates an AI summary for the session (a progress dialog appears, up to about a minute) and opens the new summary as a tab.
-
Open Folder — opens the session's folder in Windows Explorer.
-
Delete — moves the session folder to the Windows Recycle Bin after a confirmation.
Live sessions & loading
While recording, a live session appears at the top of the tree and updates in real time. When it finalizes (idle timeout or a preset change), a "Finalizing Session" dialog flushes pending speech and saves to disk, after which the session becomes an archived entry. When the library is large or you switch folders, a progress overlay indexes the sessions and can be cancelled.
