> For the complete documentation index, see [llms.txt](https://evidano.gitbook.io/ailyze-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://evidano.gitbook.io/ailyze-docs/reference/settings/transcription.md). # Transcription Settings Every button, field, and option in the Transcribe Audio/Video workflow, screen by screen, from upload sources to the transcript editor and sharing. This page lists every control in the **Transcribe Audio/Video** workflow and says what each one changes: the upload sources, the settings for each file, the charge dialog, the transcript list, sharing and downloads, and the transcript editor. It is for anyone who has run the walkthrough in [Transcribe Audio & Video](/ailyze-docs/prepare-research-data/transcribe-audio-and-video.md) and wants to know what a particular button or checkbox does before clicking it. Controls are listed in the order you meet them. The numbers in each table match the red badges in the figure above it. Prices and allowances are on [Plans, Limits & Credits](/ailyze-docs/reference/plans-limits-and-credits.md); accepted formats are on [Supported File Types](/ailyze-docs/reference/supported-file-types.md). Last verified against the staging app on 14 October 2026. ## Screen 1: Upload sources From your account page, click **Transcribe Audio/Video** (or **View previous** for the list) to open **Audio/Video Transcript Creation**. The tabs at the top choose where your recordings come from. Files from every tab land in the same list below, so one project can mix sources.
Audio/Video Transcript Creation page with seven numbered source tabs above the Local Files drop zone

The drop zone accepts audio and video files up to the size shown under Browse files.

| # | Tab | What it does | Notes | | --- | ------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------ | -------------------------------------------------------------------- | | 1 | **Local Files** | Drag recordings onto the drop zone, or click **Browse files**. | Default tab. Up to three files upload at the same time. | | 2–4 | **Google Drive**, **Dropbox**, **OneDrive** | Opens that service's file picker after you connect your account. | A **Logout** link in the picker disconnects the service. | | 5 | **Zoom** | Lists the cloud recordings on your Zoom account; tick the ones to import. | Only cloud recordings appear, not recordings saved on your computer. | | 6 | **Teams** | Lists meeting recordings from your OneDrive or SharePoint, each badged **My recordings**, **OneDrive**, or **Shared with me**. | Narrow the list with **Search recordings...** and the date picker. | | 7 | **From URL** | Paste direct links to audio or video files and click **Upload**. | See the next figure. |
From URL tab with the Download Audio/Video URL switch, a link box and the Upload button

The link must point at the media file itself, not at a page that plays it.

| # | Control | What it does | | - | ---------------------------- | -------------------------------------------------------------------------- | | 1 | **Download Audio/Video URL** | The import mode. It is the only mode offered. | | 2 | Link box | One link per line. Press Enter to upload, or Shift+Enter for a new line. | | 3 | **Upload** | Fetches each link. Finished links can be cleared with **Clear completed**. |
What the uploader tells you while files arrive * **Retry**, **Preview**, and **Remove** icons sit on every file tile. **Clear All** empties the list and **Add more** reopens the picker. * A counter above the tiles reads "Uploading 2 files", then "2 files uploaded". A failed upload reads "1 file failed. Click retry to upload again." * If a file looks like one you already added, the uploader says it "looks like a file you already added, so it was not uploaded again" and offers **Upload anyway**. Uploading it again does not merge the two. * "Processing uploaded file(s)…" means Evidano is reading each file's length. You can't submit until that finishes, because the length sets the price. * A file you dropped on the home page, or kept from the free preview there, arrives with a notice saying so; choose its language below once it is ready.
## Screen 2: The batch toolbar and cost banner Once a file is in the list, a toolbar, a cost banner, and the settings appear below the uploader.
Toolbar with Project Name, Use previous settings, the two settings modes and view icons above a blue cost banner

Two sample tones are uploaded here, so the settings panel reads "Settings for all 2 files".

| # | Control | What it does | Notes | | - | ---------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | 1 | **Project Name (optional)** | Opens the **Project name** dialog, where you name the folder that will hold these transcripts. | Up to a hundred characters. Leave it blank for a default name; the list shows an unnamed folder as "New Folder (ID)". The name saves as you type; **Close** closes the dialog. | | 2 | **Use previous settings** | Opens **Reuse previous settings**: pick an earlier transcription and copy its setup onto these files. | Copies the language, dictionary, vocabulary, spellings, speakers, AI analysis, redaction, and translation options. Trim ranges and file names are never copied. | | 3 | **Same for all files** | One settings panel for the whole batch. | Default when you upload two or more files. Trimming stays per file. | | 4 | **Custom per file** | One settings panel per file. | Applying previous settings to only some files switches to this mode. | | 5 | **Detailed view** | Full settings panels. | | | 6 | **Compact list** | One row per file: name, language, duration, trim start and end, and speakers. | The sliders icon in a row switches back to the detailed view. | | 7 | Cost banner | States how long the batch is, how much free audio it uses, and what the rest costs. With a translation, a second sentence gives the translation charge. | These are the amounts the charge dialog asks you to confirm. | | 8 | **Settings for all N files** | The shared settings panel (in **Same for all files** mode). | In **Custom per file** mode each file gets a panel headed by its name, with play, preview, and remove icons. |
Project name dialog with an optional name field and a Close button

Name the folder after the project, for example "Mobile clinic interviews", so it is easy to find later.

Reuse previous settings dialog listing one earlier transcription, Apply to All files or Selected files, and Apply settings

A badge under each earlier transcription summarises its setup; a yellow note flags anything you must set again.

In **Reuse previous settings**, **Apply to** offers **All files** or **Selected files** (tick at least one). **Apply settings** copies the setup; **Cancel** closes without changes. With no earlier transcriptions, the dialog says you don't have any previous settings yet.
Compact list with two files, a Select language field per row, durations, trim start and end, and row actions

In Same for all files mode a note above the table reminds you that settings apply to every file.

## Screen 3: Language and translation Each panel starts with a required language and an optional translation.
Language of Audio list open with a search box, the Best Accuracy group and the auto-detect entry first

Type part of a name to filter, for example "English" lists every English variant.

| # | Control | What it does | Notes | | - | ---------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | 1 | Search box | Filters the list as you type. | "No languages found." means no entry matches. | | 2 | Accuracy group | Languages are grouped into **Best Accuracy**, **Better Accuracy**, and **Good Accuracy**; the number on the right counts each group. | Settings a language cannot support stay visible but locked, with a note such as "Not available for Arabic". A Good Accuracy pick shows a hint naming a better entry when one exists. | | 3 | "2 or more languages in 1 audio (auto-detect)" | For recordings that switch language. | It clears any translation and locks the translation options. |
Translate To list showing options such as Translate US English to Spanish

The recording's own language is never offered as a target.

**Translate To (optional)** also produces a translated transcript in the language you pick. Each option reads "Translate (audio language) → (target)". **Remove translation**, or the clear icon in the field, removes it. Translation is charged by audio length; see [Plans, Limits & Credits](/ailyze-docs/reference/plans-limits-and-credits.md). The play icon on a file panel plays the recording so you can confirm which one it is; the eye icon (**Preview**) opens it in a player; the bin removes it from the batch. ## Screen 4: Optional Settings Click **Optional Settings** to open six groups. In **Same for all files** mode the section opens by itself. In **Custom per file** mode, **Apply these optional settings to all files** copies this section to every file, leaving each file's language, translation, and trim alone. ### Speaker Labels
Speaker Labels with the Number of Speakers menu and the Generic labels, Role and Name cards

Generic labels is the default: speakers come back as Speaker A, Speaker B, and so on.

| # | Control | What it does | Notes | | - | ---------------------- | ---------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------- | | 1 | **Number of Speakers** | Tells the transcriber how many voices to separate. | **Auto-detected** by default. Best and Better Accuracy languages offer one to ten; Good Accuracy languages two to ten. | | 2 | **Generic labels** | Keeps the labels Speaker A, Speaker B, Speaker C. | | | 3 | **Role** | Names each speaker by role, such as Interviewer and Participant. | Needs a Best or Better Accuracy language. Always uses known values. | | 4 | **Name** | Uses each speaker's real name. | Needs a Best or Better Accuracy language. Adds the **Auto-detect** / **Known values** switch. |
Role selected with two speaker cards filled in as Interviewer and Participant with descriptions

Role starts with two empty speaker cards; fill in one per role you expect.

| # | Control | What it does | Notes | | - | -------------------------- | ------------------------------------------------------------------------------------ | ---------------------------------------------------------------------------------------- | | 1 | **Role** field | The label for this speaker, for example Interviewer. | Up to thirty-five characters. With **Name** and **Known values**, the field is **Name**. | | 2 | **Description (optional)** | A short note that helps tell similar voices apart, for example "asks the questions". | Up to two hundred characters. | | 3 | **Add speaker** | Adds another speaker card. **Remove** on a card deletes it. | |
Name selected with the Auto-detect and Known values switch and the No speaker inputs needed note

Auto-detect works the names out from the conversation, for example when people introduce themselves.

| # | Control | What it does | | - | ---------------- | ----------------------------------------------------------- | | 1 | **Auto-detect** | Evidano infers names from what is said. No input is needed. | | 2 | **Known values** | You list each person's name, with an optional description. | ### Improve Transcription Accuracy
Improve Transcription Accuracy with Custom Vocabulary locked, a spelling rule, two dictionary terms and the filler-words box

The sample entries are the guide's mobile-clinic place names; type your own names and jargon.

| # | Control | What it does | Notes | | - | -------------------------------------- | --------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | 1 | **Custom Vocabulary** | Words and phrases the transcriber should listen for, in their spoken form. | Only some languages support it; for the others, US English among them, it is locked with "Not available for …". Type a term and press Enter, or paste a list. Spell numbers out ("seven", not "7"), use letters, spaces, and apostrophes only, and remove spaces in acronyms. | | 2 | **Custom Spelling/Formatting** | Rewrites a spoken form as the written form you want, such as "sequel" as "SQL". | Each row takes one or more **Original** words (press space after each) and one **Replacement**. The plus and minus icons add and remove rows. | | 3 | **Custom Dictionary** | Brand, company, person, and place names, so they are spelled and capitalized correctly. | Press Enter or comma to add; Backspace removes the last term. Up to a thousand terms of at most six words each. Use this one for names. | | 4 | **Include Filler Words in Transcript** | Keeps "um", "uh", and "hmm" instead of removing them. | Off by default. Turn it on for a verbatim transcript. | ### AI Analysis These add panels to the transcript editor; none of them changes the transcript text. All six start unticked, and until you choose a language they are locked with "Not available for this language".
AI Analysis group with six numbered checkboxes and their descriptions

Entity Detection and Topic Detection have an info icon listing every type or topic they look for.

| # | Checkbox | What you get | Default | | - | ---------------------- | ------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------- | | 1 | **Auto Chapters** | The recording split into chapters, each with a summary, in the editor's **Chapters** panel. | Off | | 2 | **Multichannel Audio** | Treats each audio channel as its own speaker. | Off. Only for recordings captured with one speaker per channel; it locks the speaker settings. | | 3 | **Auto Highlights** | Key moments and phrases, in the **Key Phrases** panel. | Off | | 4 | **Entity Detection** | People, organizations, places, dates, and similar items, in the **Entities** panel. | Off | | 5 | **Sentiment Analysis** | Positive, negative, or neutral sentiment for each sentence, in the **Sentiment** panel. | Off | | 6 | **Topic Detection** | Topics matched against a predefined list, in the **Categories** panel. | Off | ### Redact Sensitive Information
Redact Personally Identifiable Information ticked, showing a table of information types with Redact checkboxes

Tick only the types you need removed; at least one is required.

**Redact Personally Identifiable Information** removes personal details from the transcript. Ticking it shows a table of information types (**Policy**, **Description**, **Redact?**): names, email addresses, phone numbers, financial and medical details, and more. When redaction ran, the editor's download dialog lets you choose the redacted or the original version. ### Translation This group appears only when a **Translate To** language is set.
Translation options with glossary rows of original and translated words and the Formality choice

Use the glossary to keep product and place names consistent in the translation.

* Glossary rows of **Original Word** and **Translated Word** force how a term is translated. **Add another term** adds a row; **Remove term** deletes one. * **Formality** sets the tone: **Default**, **More Formal**, **Less Formal**, **Prefer More**, or **Prefer Less**. ### Trim Audio
Trim Audio section with a start and stop slider for one sample file and its total duration

Only the selected range is transcribed and charged.

**Trim Audio** (headed **Trim Audio (per file)** in **Same for all files** mode) holds an **Audio Start/Stop Time** or **Video Start/Stop Time** slider with two handles for each file. The range must be longer than ten seconds; a shorter one shows "Selected range must be at least 10 seconds." For audio, a **Preview Progress** bar plays the selection; for video, a player shows it. ## Screen 5: Submitting and paying Click **Process All Transcriptions**. While anything is missing, the button is greyed out and a red note above it names the problem.
Red Please fix: Language is required note above a greyed Process All Transcriptions button

Fix the named setting and the button turns blue.

Other notes you may see above the button: "Calculating file durations… you'll be able to submit in a moment.", a file whose length could not be read, or "Please complete all required fields before submitting." While submitting, the button reads **Preparing files…** and then **Processing…**.
Cost banner with a second sentence giving the translation charge

With a translation, the banner adds the translation charge to the transcription cost.

{% hint style="warning" %} **Costs money:** When the batch goes beyond your free audio, a **Create project** dialog shows your **Current balance**, the **Charge**, and the **Final balance**. Nothing is charged until you click **Create project** there. If your balance is short, the dialog offers **Top up your wallet** presets and, when free time remains, **Transcribe … for free instead**. Rates are on [Plans, Limits & Credits](/ailyze-docs/reference/plans-limits-and-credits.md). {% endhint %}
The free-trim option, step by step * **Transcribe … for free instead** appears under "Or continue without paying" only when your balance is short and some free audio remains. * It opens **Transcribe for free**: each file shows its "Selected range" and what it is "Trimmed to"; files beyond the free time are marked "Beyond free quota — won't be submitted". Untick files to choose which ones use the free time. * **Total to transcribe (free)** sums the result. **Transcribe N file(s)** submits; **Cancel** goes back. * The free option drops any translation you chose.
After submitting, "Analysis Starting" says you will receive an email once completed. If a file contains no audio, the title is "Project Submitted" and the message lists the skipped files. You are then taken to the transcript list. ## Screen 6: The transcript list The **Transcribe Audio/Video** list shows one folder per project, with its files underneath.
Transcript list with search, Folders and Transcripts, three action buttons and folder rows with owner, sharing and status

Folders open by default so you can see each file's status without clicking.

| # | Control | What it does | | - | ----------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | 1 | Search box | In **Folders** mode, matches a folder name or ID. | | 2 | **Folders** | Search folder names and IDs. | | 3 | **Transcripts** | Search the words inside every finished transcript. A **Match by** menu chooses **Hybrid** (your words plus anything that means the same), **Semantic** (meaning only), or **Exact**. | | 4 | **Upload audio/ video files** | Starts a new project. | | 5 | **Analyze transcripts** | Opens document analysis with your transcripts offered as files. | | 6 | **Refresh** | Reloads the statuses. The list also refreshes itself while anything is processing. | | 7 | **Owner / Role** | Who owns the folder, and whether you are the **Owner** or a **Collaborator**. | | 8 | **Shared with** | How many people the folder is shared with. | | 9 | Status pill | **Validating**, **Processing**, **Initial transcript**, **Enhancing with AI**, **Transcript ready**, **AI reviewed**, **Completed**, **Partially completed**, or **Failed**. A failed folder adds "You will not be charged."; a failed file says "Transcription failed for this file." | The **Language**, **Files / Duration**, and **Date Created** columns describe each folder and file. A folder made in the last hour carries a **New** badge, and an open folder shows chips summarising its setup: the language, any translation target, the speaker setup, and the AI features used. The list shows ten folders per page. ### Folder and file actions
An open folder row and its file row with eleven numbered icons

Click a file row to open its transcript; click a folder row to show or hide its files.

| # | Icon | What it does | | -- | --------------------------------------- | --------------------------------------------------------------------------------------------------------- | | 1 | Pencil (**Rename folder**) | Renames the folder in place. Enter saves, Escape cancels. Hidden while the folder is processing. | | 2 | Eye (**Show files** / **Hide files**) | Opens or closes the folder. | | 3 | **Download** | Opens the download dialog for every ready file in the folder ("Nothing ready to download yet" otherwise). | | 4 | **More actions** | The folder menu (next figure). | | 5 | **Analyze transcript** | Opens document analysis with this transcript loaded. | | 6 | **Translate transcript** | Opens [Translate Files](/ailyze-docs/prepare-research-data/translate-files.md) with this transcript. | | 7 | **Share transcript** | Opens the share dialog for this folder. Only the owner can share. | | 8 | **Download audio** / **Download video** | Downloads the original recording. | | 9 | **Download transcript** | Opens the download dialog for this file. | | 10 | **Delete file** | Deletes this file. Owner only, once the folder is finished. | | 11 | **Open in editor** | Opens the transcript editor. | A file with a translation also shows **View translated transcript** and **Download translated transcript** (a Word file).
Folder More actions menu with Analyze, Translate, Share, Download audio/video and Delete folder

Share and Delete folder are greyed out for collaborators.

The folder menu offers **Analyze**, **Translate**, **Share**, **Download audio/video**, **Download translation** (when one exists, as a Word file or a ZIP), and **Delete folder**. Deleting a folder that is still processing stops the transcription and reverses any charge for it. **Delete folder** and **Delete file** remove the transcripts for good, after a confirmation.
Selection bar reading 1 transcript selected with Clear, Download transcripts, Share and Download audio/video

Tick a folder to select all its files, or tick single files.

Ticking files shows a selection bar: **Clear**, **Download transcripts**, **Share**, **Download audio/video**, and **Delete** (only when whole folders are ticked). Several transcripts download as one ZIP with a folder per project; transcripts that are not ready yet are skipped with a notice.
Transcripts search for a phrase, with Match by Hybrid and matching passages listed under their files

Click a passage to open the editor at that moment of the recording.

## Screen 7: Sharing **Share** on a folder, a file, a selection, or in the editor opens one dialog, titled **Share Folder**, **Share N Folders**, or **Share transcript**. Recipients get an email from Evidano.
Share Folder dialog with an email box, Add message, Add as collaborator, Share through link and a Collaborators list

The two send buttons give different access: edit for paid users, view-only for anyone.

| # | Control | What it does | Notes | | - | ----------------------- | ------------------------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------ | | 1 | Email box | One or more addresses, separated by commas or spaces. | A likely typo shows a **Use …** suggestion. | | 2 | **Add message** | Adds a personal message to the invitation (dialog "Custom message (optional)"). | Sent only with collaborator invites. | | 3 | **Add as collaborator** | Gives a paid Evidano user edit access to the whole project, from their own account. | They see the folder in their list with the role **Collaborator**. | | 4 | **Share through link** | Sends anyone, including free and guest users, a secure view-only link to the shared files. | Pressing Enter in the email box does this too. | | 5 | **Collaborators** | Everyone the folder is shared with, when, and how (**Collaborator** or **Link**). | **Resend** sends the email again; the bin removes access. Hidden when sharing several folders at once. | {% hint style="warning" %} **Public link:** Anyone holding a **Share through link** email can open the shared transcripts without an account. They can read, search, play, download, and clip, but not edit; **Analyze transcript**, **Translate**, **Share**, and chat are greyed out with "Available to the project owner only". {% endhint %} ## Screen 8: Downloads
Download transcript dialog with Include timestamps and five format cards

Clicking a format card downloads it straight away.

| # | Control | What it does | Notes | | - | ---------------------- | ----------------------------------------------------------------------------------------- | ------------------------------------------ | | 1 | **Include timestamps** | Adds timestamps to DOCX and TXT downloads. | Your choice is remembered in this browser. | | 2 | Format cards | **DOCX** (Word), **XLSX** (Excel), **SRT** and **VTT** (subtitles), **TXT** (plain text). | Several files download as a ZIP. | In the editor, a redacted transcript adds **Transcript version:** with **Redacted** (PII removed) or **Unredacted** (original, includes PII); it always starts on **Redacted**. **Download translation** in the editor opens the same dialog for the translated transcript. ## Screen 9: The transcript editor Open a file to reach the editor. The transcript is shown as one paragraph per speaker turn, each with the speaker's name and a timestamp. The player sits at the bottom, or beside the video for video files.
Editor toolbar with sixteen numbered buttons from Back to More actions

On a narrower window, buttons that don't fit move into More actions at the right end.

| # | Button | What it does | Notes | | -- | --------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------ | --------------------------------------------------------------------------------------------------------------------------------------- | | 1 | **Back** | Returns to the list. | | | 2 | **Refresh** | Reloads the transcript. | | | 3 | **Search** | Opens **Find & replace** and **Search by meaning**. | Ctrl+F (Cmd+F on a Mac) opens it too. | | 4 | **AI Edits** | Lists the corrections the AI review made, with **Undo** on each. | Opens by itself the first time. | | 5 | **Analyze transcript** | Opens document analysis with this transcript loaded. | Available once the AI review finishes. | | 6 | **Translate** | Opens translation with this transcript. With a translation, **View translation** / **View original** switches between the two. | Hidden until the transcript is ready. | | 7 | **Accuracy** | Average confidence and **Words to check**. | Shown when confidence scores exist. | | 8 | **Analytics** | Talk time and turns per speaker. | Other panels appear when you switched on their AI analysis: **Chapters**, **Entities**, **Sentiment**, **Key Phrases**, **Categories**. | | 9 | **Edit text** | Turns on paragraph editing (the button reads **Editing**). | Available once the transcript is ready. | | 10 | **Undo edit** | Steps back through your edits. | Hover for a preview of what it will undo. | | 11 | **Redo edit** | Steps forward again. | | | 12 | **Export clip** | Cuts and downloads a short clip. | Runs in your browser, free. | | 13 | **Download audio** / **Download video** | Downloads the original recording. | | | 14 | **Download transcript** | Opens the download dialog. | | | 15 | **Share** | Opens the share dialog. | | | 16 | **More actions** | Buttons that don't fit, such as **Fix speaker identification** and **Text size**. | |
Editor with a speaker turn, its timestamp, the player controls and the round chat button

Press Space to play or pause when you are not typing.

| # | Control | What it does | | - | ------------------------------ | ----------------------------------------------------------------------------------- | | 1 | Speaker name | Opens **Rename speaker**. | | 2 | Timestamp | Plays from the start of this turn. | | 3 | **Back 10s** / **Forward 10s** | Skips ten seconds. | | 4 | **Play** | Plays or pauses (Space). | | 5 | Speed | **0.5×**, **1×**, **1.5×**, or **2×**. | | 6 | **Mute** and volume | Mutes or sets the volume. | | 7 | Chat button | Opens **Chat with this transcript**, a chat about this one file. Needs a paid plan. | ### AI Edits
AI Edits panel reading Changes have been applied automatically and No edits applied

This sample transcript has no corrections left to review.

Each correction appears as a card with **Before** and **After** and an **Undo** button. While the review is still running, the panel says "Final AI review in progress"; "AI changes are not yet accepted" means the changes will appear after a refresh. ### Search
Find & replace with a search term, Match case, Match whole word, the replace toggle and Replace all

Enter and Shift+Enter jump to the next and previous match.

| # | Control | What it does | Notes | | - | --------------------------------- | ------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------- | | 1 | **Search** field | Highlights every match and shows "N of M". | | | 2 | **Match case** | Only matches the same capitals. | | | 3 | **Match whole word** | Ignores matches inside longer words. | | | 4 | Replace toggle | Shows the **Replace with** row. | | | 5 | **Replace** / **Replace all (N)** | Replaces the current match, or every match. | Saved straight away, keeping word timings. **Replace all** asks you to confirm, and you can take it back with **Undo edit**. |
Search by meaning tab with a described phrase, Match by chips and matching passages

Describe what was said in your own words; matches don't need the same wording.

### Accuracy and Words to check
Accuracy panel with 96% average confidence, the Words to check list and highlighted words in the transcript

Words below 75% confidence are highlighted; hover one to see its score.

| # | Control | What it does | | - | -------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------- | | 1 | Average transcription confidence | How sure the transcriber was overall. | | 2 | **Words to check (N)** | Each uncertain word with a few words either side, its time, and its confidence. Click one to hear it. The same words are highlighted in the transcript. | ### Menus and dialogs
More actions menu with Fix speaker identification and Text size

Which items appear here depends on the window width.

Menu on a clicked word with Play from here, Start a new speaker here, Edit text from here and Clip audio from here

Click any word to start playing, editing, clipping, or a new speaker from that point.

* **Play from here** plays from the word. * **Start a new speaker here** splits the turn at this word; pick an existing speaker or a new one. * **Edit text from here** opens the paragraph for editing. * **Clip audio from here** (or **Clip video from here**) opens **Export clip** starting at this word.
Rename speaker dialog with a New name field, a Speaker B chip, This turn only and Rename all

Press Enter to rename every turn.

| # | Control | What it does | | - | ------------------ | ---------------------------------------------------------------------------- | | 1 | **New name** | The new speaker name. | | 2 | Speaker chips | Pick a speaker already in the transcript, to merge this turn into theirs. | | 3 | **This turn only** | Renames this paragraph only. | | 4 | **Rename all** | Renames every turn by this speaker (shows the count when there are several). |
Fix speaker identification dialog with Role and Name cards, the speaker details switch and Re-identify speakers

The wording stays the same; only the speaker labels change.

**Fix speaker identification** relabels the whole transcript by **Role** or **Name**, with the same **Auto-detect** / **Known values** choice as at upload. It replaces the current speaker names. Click **Re-identify speakers**; you get an email when the updated transcript is ready.
Text size popover with a percentage and Compact, Cozy and Relaxed line density

Reset to default restores both settings.

Export clip dialog with a waveform, Set start, Set end, the clip length and Download clip

Press I and O to set the start and end while the preview plays.

| # | Control | What it does | Notes | | - | ----------------- | -------------------------------- | ---------------------------------------------------------------- | | 1 | **Set start** | Marks the start at the playhead. | Shortcut I. The start and end tags take ±0.5 second nudges. | | 2 | **Set end** | Marks the end. | Shortcut O. | | 3 | **Clip length** | The length of the selection. | Preview speed and loop affect listening only, not the clip. | | 4 | **Download clip** | Downloads the clip. | For video, choose **No captions** or **Burn in captions** first. |
Analytics panel with talk time per speaker, total talk time and turn counts

Useful for checking that an interviewer didn't dominate the conversation.

Who speaks in this segment? dialog with one line per speaker and Split segment

Use it when one paragraph holds two people's words.

Hover a paragraph for **Clip this paragraph**, **Copy paragraph text**, and **Break into speaker segments**. The last opens **Who speaks in this segment?**: give each line a speaker, use **Join to the line above** to merge lines, **Start over** to reset, and **Split segment** to apply.
Edit text mode with the first paragraph open for typing and the Editing button highlighted

Your edit also saves when you click outside the paragraph.

In edit mode, click a paragraph to type. **Save** (or Enter) keeps the change, Escape discards it, and **Done editing** closes the paragraph. Each paragraph gains **Play from here**, an editable speaker name (this turn only), and **Rename this speaker everywhere**, which opens **Rename speaker**.
Keyboard shortcuts | Key | Where | What it does | | ------------------------------------------------- | ------------------------- | ----------------------------------- | | Space | Editor (not while typing) | Play or pause | | Escape | Editor | Close the open side panel | | Ctrl+F / Cmd+F | Editor | Open **Find & replace** | | Enter / Shift+Enter / Escape | Find bar | Next match / previous match / close | | Enter / Escape | Editing a paragraph | Save / discard | | Enter | **Rename speaker** | **Rename all** | | Space, I, O | **Export clip** | Play or pause, set start, set end | | Left / Right (Shift for five seconds), Home / End | **Export clip** | Move the playhead | | Enter / Escape | Renaming a folder | Save / cancel |
## Next steps * [Transcribe Audio & Video](/ailyze-docs/prepare-research-data/transcribe-audio-and-video.md) — the walkthrough these controls belong to. * [Transcription Issues](/ailyze-docs/troubleshooting/transcription-issues.md) — accuracy, speaker labels, locked editing, sharing, and charges. * [Document Analysis Settings](/ailyze-docs/reference/settings/document-analysis.md) — what happens after you click **Analyze transcript**.