Our machine transcription capabilities deliver fast, high-quality results that enable you to easily review research data and extract insights. This article will guide users on how to view, download, edit, and search transcripts.
- Discuss' automatically transcribes, translates (when applicable) every Live Research recording, research clip, video & text response for Async Research and Off-platform research.
- All Transcripts are delivered within 30 seconds for our supported languages
- Project owners, editors, and collaborators can view, edit, download, search transcripts and create custom vocabulary list.
- Viewers can see transcripts.
- Transcripts are generated in a .doc format
View Transcript
To save you time and minimize post-session editing, Discuss provides transcripts with fewer filler words, better paragraphing, and enhanced diarization (speaker separation) compared to other vendors. This results in cleaner, more readable transcripts that accurately reflect the conversation flow and speaker contributions.
Instructions to access the transcript
- Login to your Discuss account and select your desired project
- Click on the "Insights" tab and then "Recordings"
- Select your desired recording from the left-hand panel
- Click on the Transcript tab located on the right side of the video
-
For interviews with simultaneous translation, select the language in the dropdown
Download Transcript
- Login to your Discuss account and select your desired project
- Click on the "Insights" tab and then "Recordings"
- Select your desired recording from the left-hand panel
- Click on Tools and Downloads located on the top right of the tab
- In the downloads section, select Original Transcript or Translated Transcript
Bulk Downloads for Transcripts and Recordings
Bulk download lets you export all transcripts or all recordings from a project in one go. From the project Overview page, select Download All Transcripts or Download All Recordings to download a single .zip file (availability depends on download permissions).
What’s included
Transcripts from live and self-paced sessions in .doc format, with filenames that match session names
Recordings in .mp4 format, with filenames that match session names
Edit the Transcript
Using machine transcripts can be helpful when you are creating your post-session report. Sometimes machine-generated transcripts don't get it right or you might want to remove any PII information. That's why we have made it easier than ever to edit those bits that aren't quite right.
- Login to your Discuss account and select your desired project
- Click on the "Insights" tab and then "Recordings"
- Select your desired recording from the left-hand panel
- Click on the Transcript tab located on the right side of the video
- For interviews with simultaneous translation, select the language in the dropdown
Edit a Paragraph in the Transcript
The transcripts are broken up by speaker identification. To edit a specific paragraph, you can click on edit on the right side of the section. Click on save once completed. It’s strongly recommended to break paragraphs before and after the section that needs to be edited, to keep the timestamp accuracy of the transcript
Edit a Word in the Transcript
Easily edit PII, spelling mistakes, repetitions, and filler words from the platform by clicking on the specific words and choosing fix spelling from the menu
Search for Keywords
- Login to your Discuss account and select your desired project
- Click on the "Insights" tab and then "Recordings"
- In the search bar type the keyword you would like to find in the transcript and press enter.
- The targeted word will appear highlighted in the transcript
- The search feature does the targeted keyword research across all available transcripts of all other projects you have access to, the results are displayed on the left-hand side of your screen
Using Find and Replace
The Find & Replace feature allows researchers to quickly correct recurring transcript errors without manually editing each instance. Whether reviewing live sessions, uploaded recordings, or self-paced responses, you can search for specific terms and replace them one at a time or all at once.
This tool also supports transcript reprocessing after Custom Vocabulary updates, helping ensure consistent terminology across your work with less manual effort.
In Live Sessions & Uploaded Recordings
- Open the desired transcript from a live session or uploaded video.
- Click on the Find & Replace option in the transcript toolbar.
- Type the word or phrase you'd like to locate. Matches will be highlighted.
- Enter the replacement term you'd like to use.
- Choose an action:
- Replace: Update one instance at a time.
- Replace All: Update all instances with one click.
- Use Previous and Next buttons to review each match.
Note: Changes apply only to the specific transcript, not across the project.
Find and Replace in Self-Paced Responses and Clips
The Find & Replace feature allows researchers to quickly correct recurring transcript errors without manually editing each instance. You can search for specific terms and replace them one at a time or all at once.
This tool also supports transcript reprocessing after Custom Vocabulary updates, helping ensure consistent terminology across your work with less manual effort.
- Open the transcript from a clip, the main recording, or a self-paced response.
- Click on the word you'd like to change and choose Fix Spelling.
- Choose your edit type:
- Edit Instance: Change a single occurrence.
- Replace All: Update every instance in the transcript.
Note: Changes apply only to the specific transcript, not across the project.
Custom Vocabulary Lists
Using Custom Vocabulary Lists, researchers, project owners, editor,s and collaborators can enter keywords for their projects, that enable the AI to better transcribe audio with specialized terminology, such as brand or product names.
You can create different lists based on language, with up to 100 words per list.
- Affects Live, Self-Captures, and uploaded recordings.
- Lists are automatically paired to session language, meaning the researcher does not have to map lists to sessions manually.
- Researchers can easily edit words within lists, copy lists, and duplicate lists across projects within their organization.
- Proper nouns, product names, and industry-specific terminology are great candidates for keyword-boosting.
- Custom vocabulary lists apply automatically to all future sessions within the project once created.
With Custom Vocabulary Reprocessing, researchers can also retroactively apply these lists to past sessions, updating existing transcripts with the new vocabulary.
To Access Custom Vocabulary
- Go to the Documents section of the Project Overview page.
- Click on the Custom Vocabulary Lists line to review existing lists or create new.
To create a new custom vocabulary list
- Click the “Add Language List” button.
- Select the language the list will be in from the dropdown.
- Type or copy words into the Add Terms section on the left-hand side of the modal.
- Edit if needed or close the modal to save your selection
- When done, click “Add” to review them in the Preview section (the right-hand side of the modal).
To edit, duplicate or delete a custom vocabulary list
- Head to the document section of your project overview or click on "custom vocabulary" in the interview materials section
- Use the 3 dots menu to manage (edit) lists, duplicate lists, or delete lists.
Custom Vocabulary Library Reprocessing
Custom Vocabulary Reprocessing allows researchers to retroactively apply updates from their Custom Vocabulary Library to existing transcripts. This helps ensure consistency and reduces the need for manual corrections, especially when branded terms, acronyms, or commonly misheard phrases are added after a session is recorded.
- Update the Custom Vocabulary Library: Add new terms that should be recognized across transcripts.
- System Prompt: If relevant transcripts already exist, the system will list them and offer reprocessing.
- Select Sessions: Use checkboxes to choose which transcripts to reprocess.
- Confirm and Launch: Click confirm to begin the reprocessing job. Visual feedback will show progress.
- Review Updated Transcripts: Once done, click into updated sessions to confirm changes.
Benefits of Custom Vocabulary:
- Improved Recognition of Specialized Terms: Enhances the transcription accuracy of proper nouns, product names, and industry-specific terminology that the model may not have been trained on.
- Customization Without Full Model Training: Allows for targeted improvements in transcription accuracy without the need for extensive model retraining.
- Control Over Transcription Output: Enables users to specify the desired spelling and capitalization of terms in transcripts by providing them as keywords.
- Efficient Adaptation to Unique Vocabulary: Facilitates the model's adaptation to uncommon or new vocabulary pertinent to specific applications or industries.
Medical Transcription Settings
Using the Medical Transcription project-level setting, researchers can indicate that the sessions within their project are medical/healthcare-related. This triggers the AI to listen for medical and healthcare-related words, codes, and acronyms, leading to less time spent editing transcripts post-processing.
Note: This feature can be used in conjunction with Custom Vocabulary Lists.
How to use Medical Transcription
- Go to Project Settings.
- Click the checkbox to enable Medical Transcription.
Notes about Custom Vocabulary
- Affects Live, Self-Captures, and uploaded recordings.
- The setting has to be enabled before sessions are conducted and will only affect future sessions. Past transcripts will not be reprocessed automatically.
- The setting can be enabled and disabled at any time.
Benefits of Medical Transcription
- Improved Recognition of Specialized Terms: Enhances the transcription accuracy of medical and healthcare terminology.
- Customization Without Full Model Training: Allows for targeted improvements in transcription accuracy without requiring extensive model retraining.
Auto-Scrolling the Transcript
This brief guide will provide instructions on how to use auto-scrolling of the transcript. This feature is beneficial in helping you follow along with the conversation that took place and also allows a user to be able to edit the transcript based on what was said in the recording, to help the insights team achieve a more accurate machine transcription. Project owners, editors, collaborators, and viewers can use this feature.
The auto-scrolling of the transcript setting can be enabled to help you read along with the full-length video recording on the recordings page. Auto-scrolling the transcript will archive a set on an archive-by-archive basis, and you will need to set it according to your needs when reviewing each recording and transcript in the project.
- Login to your Discuss account and select your desired project
- Click on the "Insights" tab and then "Recordings"
- Select your desired recording from the left-hand panel
- Click on the Transcript tab located on the right side of the video
- Navigate to the "Auto-Scroll" slider, on the bottom left of the transcript
***Note: Editing the machine transcript at any point will disable auto-scrolling. You will need to manually re-enable the toggle to ensure it continues to follow along with the recording.
Machine Transcription & Translations
Discuss' automatically transcribes (and translates, if applicable) session recordings, video self-capture responses, or uploaded recordings conducted in a language supported by the platform. Our machine transcription and translation features deliver fast, high-quality results that enable users to easily review research data and extract insights and are extended to clips, highlight reels, and shared assets. Users can download all generated transcripts (including machine-generated transcripts) by accessing the Tools & Downloads dropdown within the Recordings page and the Self Captures menu.
- Translations are available for recording, even if no translator is present during the session.
- Users can review a translated transcript by selecting the appropriate option from the dropdown menu within the recordings page or the Self Captures menu.
Translations are available for recording, even if no translator is present during the session. Users can review a translated transcript by selecting the appropriate option from the dropdown menu within the recordings page or the Self Captures menu.
- You will still need a simultaneous translator if the translation is needed for the backroom when there are observers watching the discussion live.
- You will still need a simultaneous translator if you require dubbing on your recordings, clips, etc. since subtitles are not available through the platform yet.
- If you wish to turn off the automatic session translation feature, you can disable machine translations by contacting our support team through this support form
- Languages supported for this machine translation by the Discuss platform can be found here. If a session or self-capture response is recorded in an unsupported language, an error message will indicate that a transcript could not be generated after processing.
Tier One Language Transcription Stats:
- Speed: Transcribe one hour of video in just 30 seconds.
- Accuracy: Achieve highly accurate results with a low English word error rate of 8.4%, making Discuss twice as accurate as other vendors.
- Non-English Transcription: Over 30% higher accuracy in non-English languages compared to other vendors.
View the full list of supported languages here
Benefits of AI-Powered Translation:
- Context-Aware Translations: Understands context better than traditional translation tools, producing more natural and coherent translations.
- Handling Complex Sentences: Effectively manages nuanced and idiomatic expressions, especially in widely spoken languages.
- Multilingual Support: Covers a broad range of languages and dialects.
- Domain-Specific Terminology: Performs well in specialized fields such as technology, healthcare, and law when context is provided.
- Conversational Tone: Ideal for informal and formal texts, preserving tone and intent.
Setting Discussion Guide Questions
Discussion guides are meant to be helpful tools in the 'Meeting Room' to help navigate your conversations. The moderator works best when they are fully concentrated in the conversation that is taking place, so sometimes checking the discussion guide questions in the conversation just isn't feasible or gets added at the wrong time.
The ability to drag and drop the discussion guide questions after the fact helps you create a more defined machine transcript and helps play as the true marker of where the questions capture the start of the response from a respondent.
Note: Project viewers, password-protected single share viewers, and anyone with a copy of the downloaded discussion guide for a recording will be able to view the placement of the discussion guide questions, but not edit them.
- Login to your Discuss account and select your desired project
- Click on the "Insights" tab and then "Recordings"
- Select your desired recording from the left-hand panel
- Click on the Transcript tab located on the right side of the video
- For questions that were not checked off in the meeting room or 'Not Answered' at the bottom of the transcript, drag those questions in front of the moment in the transcript that the question should be.
- For questions that were checked off in the meeting room and landed at the incorrect moment that the question should have been checked off at, drag those questions in front of the moment in the transcript that the question should be.
Human Transcription
Human Transcription can be requested for your project. If you order human transcript(s), they will replace the machine transcript(s) in the dropdown.
Please Note: Human transcripts take 24-72 hours to produce.
Off-Platform Research Transcript
You can generate transcripts for your Off-Platform Research. Start by uploading your research, select the video language, and a transcript will be generated.
Self-Captures Submissions
For Self Captures video submissions and AI-led conversations, the platform generates a transcript in the original spoken language and an AI-generated English translation of that transcript, so you can review and analyze responses in either language.
Speaker Identification
Automatic Speaker Identification is available for Live sessions and for Off-Platform Research.
Speaker identification enhances the clarity and readability of market research reports. Including the names or identifiers of speakers alongside their responses makes it easier for readers to follow the conversation flow and understand who is expressing certain opinions or providing feedback. This facilitates comprehension and ensures that the findings are presented coherently.
Note: The transcript will capture the insights of all moderators, respondents, and translators but participants connected via phone bridge (dial-in), during a live session won't be identified in the transcript.
Live Meeting Room Transcript
The Live Meeting Room Transcript provides a real-time text feed of your session directly within the Meeting Room. This allows moderators and observers to follow the conversation as it happens in the original spoken language.
How it Works
Real-Time Display: The transcript appears in the right-side panel (under the Chat tab) and updates instantly as moderators and respondents speak.
Session Language: The transcript is generated in the language set for the session. For example, if the session is conducted in Spanish, the transcript will appear in Spanish.
Individual Control: Each participant can start or stop their own live transcript without affecting others.
Visibility: Only moderators and observers can view the live transcript; it is never visible to respondents.
Helpful Tips for Use
Real-Time Viewing: The live transcript is a temporary tool designed for real-time engagement and is not saved. Once the session ends, the live text is purged.
Toggling Transcription: If you stop the transcript during a session, the current history is cleared for your view. Restarting it will begin the transcript again from that specific moment.
Custom Vocabulary: The live transcript pulls from any Custom Vocabulary Lists created for the project. If your session involves technical subjects, specific jargon, or unique acronyms, adding them to a list beforehand will significantly improve the accuracy of the live transcript.
AI Simultaneous Translation in the Meeting Room
AI Simultaneous Translation enhances the Meeting Room experience by adding a real-time English translation layer to non-English sessions. This allows observers to follow sessions live in English without requiring a human translator in the backroom.
How it Works
The Toggle: Moderators and observers can switch between the original session language transcript and the AI-generated English translation.
Viewing Layer: The translation is a visual tool for live viewing only; it does not change the audio or the final video recording.
User-Specific: Each moderator or observer chooses their own setting. Enabling translation for yourself does not change the view for other participants.
Access Roles: This feature is available to Moderators, Observers, and Support agents. It is not available to respondents.
Post-Session Behavior
Finalized Transcripts: Regardless of whether you use the live feature, the platform automatically generates a finalized transcript in the session language and an English translation for post-session access on the Recordings page.
Things to Note
Non-English Sessions Only: If the session language is set to English, the translation toggle will not be visible.
Compatibility with Human Translators: AI Simultaneous Translation cannot be used at the same time as a human simultaneous translator in the same Meeting Room. If your meeting is configured to use a human simultaneous translator, the AI Simultaneous Translation feature is not supported and will not be available.
Session Persistence: Like the standard transcript, the translation feed is not saved. If you stop the transcription mid-session or when the meeting ends, the live translation history is cleared.
Enablement: If you do not see this option in your Meeting Room, please reach out to your Account Manager for assistance.
Custom Vocabulary: The live transcript pulls from any Custom Vocabulary Lists created for the project. If your session involves technical subjects, specific jargon, or unique acronyms, adding them to a list beforehand will significantly improve the accuracy of the translation.
Supported Languages for Live Meeting Room Transcription and Translation
View the full list of supported languages here
Note: Tagalog is currently not supported for the Live Meeting Room Transcription and Translation