August 16, 2023
Two big features since our last published update are Profanity Filters and Keyword Boosting. Complementing our existing autoreplacements feature, these help you use real time captions and translations more safely and accurately.
You can enable profanity filtering when you create a session. Any profanity detected will be made blank. If it filters a word you want (e.g. the name Dick), for now you will need to disable the profanity filter.
If the captioner is not detecting a word you need (e.g. specific terminology you use or a person’s name), you can boost that word to help the captioner detect it. This boosting is based on how a word sounds, so if your word has unconventional spelling, you will need to input it in a phonetic form and use an ARDB entry to automatically get it to the correct spelling.
- Improve responsiveness for certain languages
- Add support for keyword boosting and profanity filters
- Add support for emoji reactions
- Improved transcript extraction from subtitle files to do logical paragraph breaks and take into account speaker labels.
- Fix occasional out of order release of utterances in autopilot.
- Speech recognition engine enhancements
- Trim entries to the Autoreplacements database (ARDB) so that replacements do not accidentally remove spacing.
- Support using a screenshare of chrome tab audio as an input to the captioner.
- Add support for keyword boosting and profanity filter.
- Fixes autosegment to not line break on periods in urls.
- Beta: Added ability to restrict sessions to guests with an access code.
- Improved generation of transcript from session history.
- Trim entries to the Autoreplacements database (ARDB) so that replacements do not accidentally remove spacing.
- Created an options section in session editor for features like profanity filter and keyword boosting.
- Add ability to search by session title.
April 10, 2023
- Projector view
- You can now pick a presentation to show from the projector controller instead of having to link it via the session editor.
- Adjust subtitle alignment to be right, left, centered or justified.
- Adds option to disable dynamic text resizing
- Adds dynamic chunking for multilingual display to ensure subtitles are readable and in sync.
- Adds tunable per language settings for subtitle length and display of characters per second.
- Document editor
- Adds AI-powered ability to summarize text, generate titles, extract keywords and actions items from a transcript, and suggest social media posts
- Quality of life improvements like better mobile support and keyboard shortcuts (press ctrl+s to save!)
- Improved line splitting to preserve lines when switching between rich text and alignment editor.
- Media (AVCS)
- Increased support for other media file types like mov and webm.
- Beta support for screen recording from within spf.io to create screencasts.
March 12, 2023
- AI-powered Multilingual Polls have arrived! Enable your audience to vote in their own language during your meetings.
- Multilingual conversations feature is rolling out in private beta
- Added the ability to edit i18n.json files in the document editor so you can use it to translate supported localization files for apps and websites.
- Added Ukrainian and Russian localization to audience view
January 4, 2023
Happy New Year!
- Create hardsubbed/burned-in caption videos. These are easy to share in WhatsApp, Line, Telegram, WeChat, Signal, etc.
- Translation memory analysis and cleanup tools are here!
- Check out the free cloze / fill-in-the-blank tool we made for teachers!
- Improved performance of the subtitle editor when dealing with large files.
- You can now upload PPTX files as slides (in addition to PDFs). The slide images will still be static, but this is more convenient than having to convert to PDF first.
- Adds support for creating custom captioner traces on windows so you can practice post-editing live captions on your own content.
- Autoreplacement database fixes to properly support autoreplacing multiple phrase matches.
- Bugfix for when a new channel is added to an instance and activated via translation on demand.
November 3, 2022
- Adds native support for ZoomGov with Cloud Loopback
- You can edit the line height of slide translation overlays to improve the appearance of translated slides.
- You can now lock slots in a projector view so that remote slot changes will not affect it. This solves the issue with “projector view hijacking” that sometimes happened when someone using a speaker view changed channels. The speaker view also now has its own channel setting independent of projector views.
- Use the “ctrl+shift+1” (or other numbers) keyboard shortcut in the captioner to quickly switch languages. You can do these with numbers 1 through 9, corresponding to each channel’s position in the language switcher.
- Simplifies setting channel endpoints in a session so that by default only the first active channel is shown. This is the typical behavior users need and the per channel endpoint configuration is now hidden in advanced settings.
September 30, 2022
Happy International Translation Day! Here are some exciting updates from this past month:
- Cloud Loopback now supports YouTube and Vimeo as an audio source
- You can download a PDF of slide translations with one click
- Keyword boosting means you can more easily add the unique words you use for the captioner to detect.
- The subtitle editor now shows gaps to alert you to places that may need additional cues (e.g. “Applause” or “Music playing”).
- Scrub to a place in a video and the subtitle editor will jump to the nearest subtitle for faster navigation. Edits you make can be live previewed in the video.
August 29, 2022
Here are some updates from this summer!
- Adds Burmese and Tigrinya presets
- Adds Google Meet support to Cloud Loopback
- Adds word and character counts to document info
- Improved error logging
- Adds speaker diarization support
- Adds basic support for E3 ASR keyword boosting
- Enable fine positioning of video in overlays view
- Add enhanced subtitle editor that supports flagging subs for review
May 3, 2022
In addition to bugfixes and performance improvements in the database and frontend, this latest update includes enhancements to the following parts of spf.io:
- New “Engine 3” automatic captioning preview for English!
- You can now autosegment lines based on punctuation–this makes autopilot even more usable for translation!
- You can set per language engine preferences.
- Streaming “volatile” captions: caption engines can stream results before they are confident in them. This feature means the audience view and remote captioner views can show previews of these results to start reading text even faster than before.
- Add live captions and translations to your own webpage! Read more about the new “subtitle only embed mode” and see examples of what you can do with it in this article.
- In-person events are making a comeback! Now, your audience can quickly share translation with each other through a QR Code in the audience view.
- Spf.io’s audience scaler powerup enables you to serve tens of thousands of simultaneous viewers. Now this view adds a statistical summary that breaks down the most number of concurrent connections and the average duration of connections, broken out by language. This helps you estimate your audience size per language so you can make informed decisions about how to grow your multilingual audience.
- Before, you could easily delete session history, and now you can also delete sessions themselves! To do so archive the session and then view your list of archived sessions. Now a delete button will appear enabling you to delete archived sessions.
- Added presets for the Slovak language.
- Improved preconfiguration for remote operators and voice writers so they can dive into an event and get started working right away. This includes a new “remote audio shortcut” that with one-click streams the event audio to the captioner.
- You can now play videos you uploaded to spf.io in the projector view!
- You can now flag subtitles that need to be fixed in the subtitle editor for easier collaboration and community-sourcing.
March 14, 2022
Happy Pi Day! Here’s some update announcements for you 🙂
- Added an additional language slot–projector view can now show up to four languages at the same time!
- Setting more than 2 language slots persistently shows translation instead of showing slides
- Small bug fixes when displaying mini-apps
- You can now select a locale (e.g. en-GB or en-US) when transcribing videos
- There is a new subtitle editor that makes it easier to edit individual subtitles. Access it by clicking on this button in the document editor when viewing an srt document:
- We added a button for removing non-speech captions from srt files.
- Persist selected captioning engine
- Some engine/language combinations support automatic punctuation.
- Add autoreplacements from captioner live by selecting text and adding to the ARDB.
- Improved key macros and abbreviations.
- You can delete pages of history when you want to clean it up for use with FastSRT.
- Admin can send password reset email to members.
- Autoreplacements applied server-side now include word boundaries in matching for greater accuracy.
- Updated some stats in analytics view.
October 20, 2021
*NEW* Apply autoreplacements to documents:
- You can now apply your captioner and translation autoreplacements to your documents!
- Use this personal “autoreplacement database” (ARDB) to quickly correct transcripts, subtitles and translations
- As you add to your ARDB and train spf.io to apply corrections, your editing process will become quicker and more automated
- Learn more about how to “Fix your captions and transcripts faster with autoreplacements“
- Fill gaps between subtitles – set a specific threshold so subtitles stay on the screen for longer before the next one appears
- Read the article: What is fastSRT?
Cloud Loopback (for Zoom and StreamYard captions with multiple speakers):
- You can now enter cloud loopback settings immediately in quick start mode (Zoom/StreamYard link, event hours, etc.)
- Add cloud loopback hours in captioner
- In captioner, you now have the option to reset the remote captioner for easy troubleshooting
- You can now recreate subtitles/transcripts from scratch by unlinking documents from the original media
- Remote control of captioner now supports segment mode (new line is inserted when a speaker pauses)
- Usernames can now include periods
- Improved reliability of Autopilot max wait time and Caption Token API integration
September 24, 2021
OBS to Vimeo caption and subtitling – You can now use spf.io to ingest caption or subtitle to OBS and stream it to Vimeo.
September 17, 2021
Learn something new: Subscribe to spf.io’s YouTube channel to get the latest updates, and level up your skills with our newest how-to videos.
Add picture-in-picture to your OBS stream – show slides as a thumbnail in your stream with picture-in-picture for a professional look. Easily control your presentation from your laptop or even your phone!
Bidirectional language mode for interpreters – Speaking more than one language in Zoom? Interpreters can now switch seamlessly between languages when subtitling a call.
- Learn how to set up cloud loopback for Zoom/StreamYard calls with multiple speakers
- Improve the quality of your speech-to-text content
August 18, 2021
Twitch captions integration (OBS) – You can now see spf.io-generated captions in your Twitch Stream Manager preview.
Cloud loopback session powerup – Cloud loopback allows you to caption multiple speakers on Zoom by capturing the sound of your Zoom call (documentation here). Cloud loopback is now a quick and easy feature requiring no extra installs!
- Embed your Audience View in a web page
- Add session notes
- Disable audio features in Audience View
- Stream raw captions in different modes to meet CART/voicewriter needs
- Process files more easily with larger upload capacity
- Analytics: access a variety of usage reports now
- Access support quickly: request office hours through your account
- Need more details? Click here
August 16, 2021
We created this change log to streamline how product updates are communicated with customers. We will include simple lists of updates along with links to articles that go into greater depth.
See change log summary prior to this date
May 11, 2021
- Adds translation coordinator role
- Adds more language presets including Catalan, Galician, Uyghur, Slovenian, Serbian, Albanian
- Admins can now view user activity on an instance
- Improves document usage reporting, including difference between autotranslate and post-edited
- Adds media usage report
- Increases selection of translation engines
- Enable bidirectional bilingual mode for Zoom translation
- Adds ability to stream raw captions from plaintext editor for CART/voicewriters to use
- Plaintext editor now supports right-to-left languages
- Adds OBS captions integration for use with Twitch
- Session notes enable you to share key information with your team and remote operators
- Adds ability to request office hours within spf.io
- Improved file upload ability to support larger PDFs and media files
- Adds the cloud loopback session powerup to support captioning and translating Zoom calls without needing to install additional software
- Give users more control of the audience view experience by enabling disabling of audio features
- Improved autopilot performance with remote captioning
- Adds support for clearing translation of a specific slide
- Improves handling of markdown headers in documents
- You can now embed the audience view in your own webpages
- You can set a delay for the audience view without needing to specify a video
- Improves YouTube live captioning
Jan 14, 2021
- Spanish defaults to es-US so offline transcript works out of the box
- High quality subtitle autotranslation feature. Automatically generate properly timed high quality translations of subtitles.
- Import .srt subtitle files in the document editor
- Onboarding requires customers to setup their channels when they first login.
- Youtube Channel Integration: Seamlessly import captions from YouTube videos in your channel, automatically create translations and push the translated subtitles back into Youtube–all from within spf.io!
- New virtual teleprompter in spf.io overlays makes you look good on camera! Spf.io overlays makes it easy to open caption livestreams too.
- Adds magic link login support.
- Adds audio monitoring support to operator view so audio can pass through the captioner and still be heard.
- Adds smart document chunking/wrapping for long sentences so that you can release good sized chunk captions and translations in real time.
- Adds document printing capability.
Aug 25, 2020
- Adds download speed test
- Zoom captioning support (including translation into 1 language)
- Embed video streams from FB, YT and Zoom in audience view
- Waterline metrics for per channel connections
- Persists document editor configuration like light/dark theme and font size
- Improve autopilot by enforcing a max wait time for automatic release
- Adds plain text editor mode to captioner view for stenographers/CART providers
- Adds ability to generate timed subtitles from an aligned transcript and translation
- Stores raw caption history
- Add differ to compare automatically generated transcripts/translations with post edited versions
- Adds ability to upload media (audio and video), transcribe it automatically and generate captions
- Edit subtitles in the document editor with support for jumping to timestamps in the source media
- Enable live captioning and/or translation of YouTube Live
- Autoreplacements are applied to offline transcriptions
Feb 12, 2020
- Export translation memory strings to tsv
- Enable cleaning up deleted presentations
- Updated logging
- Improved access control
Nov 5, 2019
- Adds support for translation on demand of slides
- Upgrade server runtime
- Simplify navigation to put “documents” and “slides” under “resources”
- Selecting lines of text in the audience view is disabled for sessions that “Do not store history”.
Oct 5, 2019
- Improved scalable audio livestreaming for interpretation and assisted listening.
- Improve readability of auto-translated slides by letting users read them without the background images.
- Download session channel history as an .srt file to help with video subtitling.
- Enables changing subtitle font color and background color.
- Bugfix: Hide announcement text in projector view when a new slide is shown.
Jul 24, 2019
- Speaker view and slide controller are resilient to network problems.
- Display raw captions in multiple languages in the projector view.
- Bugfixes in audio livestreaming
Jul 18, 2019
- Speaker view supports smooth scrolling on iPad and easy navigation between playlist entries.
- Projector view and speaker view will keep device screens awake to prevent accidental falling asleep during events.
- Show slide progress count in slide controller
- Adds audio livestreaming support for Safari on iOS 12 devices
- Speaker view supports translation on demand
Jun 24, 2019
- Released new speaker view for manuscripters.
- Captioner view and projector controller have superseded interpreter view.
- Captioner view is now compatible with certain stenokeyboards.
- Adds support for sessions without history (for transient or sensitive content).
- Improved display of raw captions on projector view.
- Release paragraphs as slides in the captioner view.
- Adds one click swap channels feature to streamer for bilingual events.
Dec 14, 2018
- Added simple apps dedicated to audio streaming and spoken interpretation.
- Audience views can receive raw real time captions for accessibility.
- Enhanced operator chat and audio streaming reliability.
- Quickly switch channels in the beta captioner to support different languages spoken on stage
- Added Amharic, Lao, Farsi, Urdu, Javanese & Sundanese presets.
- User feedback can be automatically translated.
- Adds translation memory support so pretranslated materials can be easily re-used.
- Adds an easter egg game to the projector view.
- Adds initial designs for a simplified projector view controller.
Jul 11, 2018
- Markdown formatting preserved in audience and projector views
- The beta captioner can be controlled via clicker
- Experimental support for document line suggestions in beta captioner
- Stream audio to up to 10 devices through the beta captioner
- Added panel discussion practice traces
- Documents can be downloaded with one click
- Filter results by channel in multichannel viewer
- Enable document translation of SRT (subtitle) files and i18n json
- Added ability to request user feedback from interpreter view