Changelog
v5.1
Release Date: 2025-09-04
Timeline Optimizations
This release introduces significant timeline optimizations, delivering improved speed and performance. The updated technology foundation also enables future enhancements and advanced features.
v5.0
Release Date: 2025-08-20
Voice Quality & Expansion
This release expands the variety and quality of Revoiceit's voice capabilities. With an upgraded expressive model, support for a wide range of new source and target languages, and nearly twice as many available voices, users now have more flexibility and can achieve higher-quality results and better localization coverage than ever before.
Updated Expressive Voice Model
The expressive voice model has been improved with new prompts, resulting in more natural and higher-quality speech output.
New Source and Target Languages
This release significantly expands language coverage. Users can now generate AI voices in Bulgarian, Danish, Dutch, Finnish, Greek, Indonesian, Norwegian, and Swedish. Several new source languages are also supported for transcription and as input for dubbing workflows. These include all of the aforementioned languages, as well as Bengali, Czech, Estonian, Hungarian, Malay, Romanian, Russian, and Thai.
New Voices
With a total of 93 voices, the platform now offers nearly twice the selection, providing greater variety in terms of gender, age range, and voice timbre.
v4.2
Release Date: 2025-08-12
Batch and Project API
This release introduces significant enhancements to the Revoiceit API, providing greater flexibility for integrating workflows. The new Projects & Batches API enables developers to automate the setup and management of dubbing workflows by creating and modifying projects and batches. The Media API enables secure file uploads via pre-signed URLs and the deletion of specific files or folders when they are no longer needed. Together, these updates simplify automation, improve scalability, and reduce the manual effort required by teams managing multilingual media workflows.
v4.1
Release Date: 2025-07-22
Faster Results with Custom Audio Input and Rewriting
This release introduces two powerful features that provide users with greater creative control over voice and text. With Dillo, users can guide speech synthesis with their own recordings to achieve natural pronunciation and prosody. The new LLM-based Rewrite feature offers AI-generated text alternatives for easier style adaptation and timing adjustments. The update also includes improvements to script templates, password management, and a new machine translation connector.
Dillo: Enhanced AI Speech Editing with Custom Audio Input
The new Dillo feature gives users greater control over the AI speech editing process. Users can upload an audio file or record themselves, and the system will generate a synthesized voice line that mirrors the pronunciation, prosody, and emphasis of the original input.
LLM Adaptation: Smart Text Rewriting with AI
The text editor now has a new "Rewrite" action that allows users to send a line to an LLM for rewriting. The system returns five alternative versions, and users can select the one that best fits their needs. A standard prompt guides the rewriting process, but users can also add a short custom prompt to further refine the output. This feature is ideal for adapting tone, improving style, and fitting text within video time constraints.
Manuals for Script Templates
A manual has been added to the Line Dubbing and Audio Production script templates. Located on the second sheet of the new template file, the manual explains how to use the templates and lists all supported options and language codes. You can download it from the Documentation section under "Download Reference Data."
Account Settings and Password Management
Users can now change their password on the new Account Settings page, which is accessible by clicking the account icon. When creating new users, managers can also specify whether the account is personal or shared. Password changes are disabled for shared accounts to maintain controlled access.
Globalese Integration
Globalese has been added as a new machine translation provider. To start using it, navigate to the Machine Translation page and create a new profile using your Globalese credentials.
v4.0.2
Release Date: 2025-06-13
Faster Audio Creation and AI Speech Editing
This release focuses on accelerating and refining the AI speech editing process. With faster inference speeds and precise control through the new Chop & Merge feature, users gain more power and flexibility in crafting high-quality voice output.
Chop & Merge Takes
A new Chop & Merge feature has been added to the line history, allowing users to define regions within generated takes and combine the best parts into a single, refined version. This gives users more control and flexibility when crafting the perfect take for the timeline.
Faster Audio Generation
This release brings performance improvements to the inference pipeline, resulting in significantly faster audio generation – particularly when processing multiple lines simultaneously during autoinference.
Machine Translation for Audio Production
Machine Translation is now available in Audio Production projects. Users can upload a script in one language and instantly generate translations into other languages, streamlining multilingual content creation and localization.
API History & Statistics Pages
The API section has been enhanced with new History and Statistics pages. The History page logs all API calls, helping developers monitor and troubleshoot integrations. The Statistics page provides insights into successful and failed requests over time, with filtering options by date range and API key.
v4.0.1
Release Date: 2025-05-21
Audio Production Script Upload and Custom Voice Management
This release brings greater flexibility to Audio Production with support for additional file formats and dynamic language handling. It also introduces a dedicated AI Twin interface for managing custom voices and optimizes Full Conditioning by processing audio faster.
Audio Production: Support for More File Types and Flexible Language Upload
Audio Production now supports the upload of SRT, TXT, PDF, and DOCX files in addition to the existing script template. Users can upload multiple languages at different stages, add new languages to a batch at any time, and re-upload scripts to correct or update content.
New AI Twin Interface
A new section on the Voices page now displays Cloned Voices and AI Twins, providing a dedicated overview for users with custom-created voice models.
Optimization: Audio Compression for Full Conditioning
To improve performance and reduce processing time, Full Conditioning now uses compressed MP3 files instead of uncompressed WAV files.
v4.0
Release Date: 2025-04-29
Text-to-Speech API
The highlight of this month's release is the Revoiceit TTS API, which gives developers access to integrate voice generation into their own tools and workflows. As part of this rollout, we've moved from a minute-based system to a more flexible credit-based model, making it easier to integrate more services into our platform. Other updates include powerful new emotion control features, streamlined line dubbing tools, and enhanced export capabilities.
Revoiceit TTS API
The new TTS API enables seamless integration with external applications. Key features include:
- Authentication & text-to-speech generation with timecodes, source audio, voice selection, styles, and emotions
- List library voices with audio samples
- Glossary management: save, retrieve, and update glossaries to customize pronunciations
- Receive final audio files
New API-related interfaces:
- API Key Management: Generate and manage multiple keys for different apps
- API Credit Usage Dashboard: View credit consumption by period and purchase new credits
- API Statistics: Track request history over time
- API Call History: Debug and monitor integrations by reviewing past requests
- Platform vs API Usage Comparison: Compare used minutes across UI and API
Line Dubbing - Simplified Audio Editing
The timeline has been removed for Line Dubbing projects and replaced with audio tools that allow users to manipulate individual lines more efficiently.
Full Emotion Conditioning
A new Full conditioning option is now available for Line Dubbing and Media Dubbing projects. When selected for a line, Revoiceit analyzes the original source audio and tries to deliver the target audio as close to the original as possible.
Upload Emotion & Intensity in Script Templates
Users can now define emotion and intensity levels directly within script templates during upload into Line Dubbing and Audio Production projects. This gives greater control over how lines are delivered and bypasses the need for automatic emotion detection.
New Excel Script Export
An Excel script export option has been added for Line Dubbing and Audio Production projects, enabling users to review source & target language scripts more easily outside the platform.
v3.3
Release Date: 2025-03-06
Emotion Intensity Control
Three Emotion Intensities per Emotion
A new feature now allows users to adjust emotion intensity when selecting an emotion for a line. Each emotion can be set to low, normal, or high intensity, providing finer control over how a line is delivered. For narrative conditioning, these levels correspond to Epic, Normal, and Announcements, providing greater customization for different use cases.
v3.2.1
Release Date: 2025-02-28
Usability & Performance Enhancements
This release includes a number of usability and performance enhancements throughout the system to improve the overall user experience.
Timeline Improvements
Timeline performance and responsiveness have been improved. In addition, the toolbar has been reorganized to keep relevant actions within easy reach while maximizing video and timeline space.
New Features for Line Dubbing Batches
Several enhancements have been made to improve the line dubbing workflow.
Line Dubbing Review Page: A new source review page ensures that users review and select emotional prompts to be propagated to all target languages.
Export Options: Line dubbing stems can now be exported with multiple naming formats, including source file names, source file names with language codes, or sequential numbers.
Multiple Backgrounds & Improved Asset Management UI
The Asset Management screen has been redesigned for improved usability, making it easier to quickly identify different types of assets. In addition, users can now upload multiple background tracks for use in the timeline and mixing process.
Admin Messages for Remaining Minutes
To prevent unexpected interruptions, administrators now receive notifications when their platform has 30 minutes and 10 minutes of audio remaining. This ensures that they have enough time to top up credits and maintain continuous operation.
Expressivity & Diversity Control for Audio Generation
New Expressivity and Diversity controls give users more flexibility when creating audio. These controls allow users to fine-tune the balance between emotional depth and stability in their generated audio.
Diversity: Lower diversity results in more consistent pronunciation and prosody, while higher diversity increases variation but may produce less natural-sounding takes.
v3.2.0
Release Date: 2025-01-30
Audio Pauses & New Voice Models
This release introduces automatic detection of pauses in generated audio, along with the ability to manually insert pauses using inline tags for greater control. Another key highlight is the addition of our new Arabic language model, complemented by 11 new voices now available for audio generation.
Audio Pauses & Auto-Regioning
This feature gives users more control over their audio output. You can now insert short, medium, or long pauses using pause tags. In addition, a new AI-driven system detects natural pauses in generated audio and automatically segments them into subchunks, allowing for quick and easy pause adjustments without manual editing.
Arabic Language Support
Revoiceit now supports Arabic speech generation. Users can upload or machine translate Arabic text, which the platform automatically diacritizes before generating natural-sounding Arabic audio.
11 New Voices
We've added 11 new voices to provide more flexibility in the casting process. The new options cover a range of timbres (high, medium, and low) as well as different age groups, from young to middle-aged, allowing for more nuanced and diverse voice choices.
DeepL Integration
Users can now create machine translation profiles directly linked to their DeepL account and use its formality and glossary features to get automatic translations for scripts.
Mixing Enhancements
Users now have more flexibility when mixing the final audio output. Mixing is now possible even if not all lines are bounced. A new mode allows the original voice track to be mixed at a lower volume beneath the translated voice track for better audio layering. Additionally, videos can now be downloaded in both MP4 and MOV formats.
Updated "Resynch" Action for Audio Production Projects
The Resynch action has been improved to streamline audio alignment. Instead of returning audio segments to their original positions, it now places all segments in sequence. Users can also specify a pause duration between segments for efficient timing control.
v3.1.0
Release Date: 2024-12-20
Emotion Detection and Mixing
In addition to several UX improvements in the Editor workspace, this release introduces two powerful new features: Emotion Detection and Mixing Options.
Emotion Detection & Transfer
When users upload media, the system now analyzes the audio to determine the emotion with which a line is spoken and automatically assigns it to the line. The detected emotions are then transferred to lines in all languages, reducing the workload for reviewers.
Mixing Options
Users can now mix and export video or audio files directly from within the tool. This feature is especially valuable for users without access to audio post-production tools, allowing them to easily deliver a fully mixed output.
Split/Merge Lines in the Preview Page
Users can now split and merge lines individually for each language. This feature allows users to fix incorrect segmentation and adjust lines to work for their specific use case.
Enhanced Tempo Control
The new tempo control interface makes adjusting tempo more intuitive. In addition, a new tempo control mode has been added to the Timeline. Users can enter this mode, click on an audio chunk, and stretch or shrink it to adjust the tempo directly and ensure that there are no overlaps.
v3.0.1
Release Date: 2024-12-12
Machine Translation
Language Weaver Integration
Revoiceit now integrates with machine translation systems, starting with Language Weaver by RWS. Users can create machine translation profiles directly linked to their Language Weaver account and use its formality and dictionary features to get automatic translations for scripts.
v3.0.0
Release Date: 2024-11-19
Emotional Voice Model and Streamlined User Flow
This version of Revoiceit combines a new Xpressive voice model, which allows users to create emotion-rich voice lines, with a redesigned interface and new functionalities that simplify project creation and management for a more intuitive user experience.
Xpressive Model with 8 Emotions
The Xpressive model comes with 8 emotions: Anger, Curiosity, Fear, Joy, Pain, Pleasure, Sadness and Surprise. To use these emotions, select the "Xpressive" model during project creation. In the Editor, you can then select the desired emotion to generate audio for each line.
New Optimized User Flow and Asset Management for All Project Types
Asset management has been implemented across all project types, allowing users to easily upload and download scripts, voice tracks, background audio, and other media. In addition, all project interfaces have been redesigned for improved usability and a more intuitive experience.
Automatic Speaker Detection
A new feature has been released that automatically detects the number of speakers in audio and video files. This pre-analysis eliminates the need to manually count speakers as the system now automatically identifies characters, simplifying the media ingestion process.
Autocasting
This new feature automatically analyzes the original audio and suggests the most similar synthetic voice. This helps users make more informed decisions when choosing a voice that best fits their content.
Automatic Inference
A new option has been added that allows users to automatically infer voice tracks immediately after asset and language preparation and to specify the number of alternate tracks generated. This feature can improve efficiency by reducing wait times for audio content reviewers. If desired, automatic inference can be turned off completely.
New SRT Script Import for Voice Over Projects
Previously, media files were analyzed based solely on their audio to set in/out timecodes and segmentation. With this new feature, users can select an SRT file and choose to import the speaker information, timecodes, and segmentation it contains.
New Character Selector and Enhanced Functionality
The new character selector is more user-friendly and allows users to apply character changes to individual lines. Additionally, it introduces the ability to add characters even after media ingestion, offering greater flexibility and control over project customization.
User Assignments
Project owners can now assign language content reviews to specific users, such as editors or project managers. Editors are restricted to the languages and projects they are working on, making it easier to assign clear responsibilities and ensure security.
Adding Languages to Running Projects
With this update, users can add languages to existing projects and batches to meet evolving language requirements.
v2.4.2
Release Date: 2024-10-24
Editor Improvements
This release is all about new features that streamline the workflow for audio content reviewers.
Mark Alternative Takes as Favorite
This new feature allows users to mark generated takes as favorites, making it easier to save potential candidates for comparison.
Warning for Overlapping Chunks
Users now receive a warning in the Editor lines and an indicator in the Timeline when audio chunks overlap in the target language. This helps identify lines that may need speed adjustments or regeneration for better alignment.
v2.4.1
Release Date: 2024-10-08
Bug Fixes & Performance Enhancements
This release is focused on addressing bugs and improving performance:
Timeline Controls for Audio Files in Dubbing Projects
An issue where timeline controls were not visible when users uploaded audio files to Dubbing Projects has been resolved. With this fix, all controls are now displayed correctly.
Enhanced Handling of Unsupported Characters in Script Files
Previously, uploading a script file containing unsupported characters would cause the batch creation to hang indefinitely. Now, users are notified of such an error in the corresponding line of the Preview page, allowing them to correct it manually.
Consistent Decimal Notation on Billing Page
An inconsistency on the Billing page has been addressed where both commas and periods were used as decimal separators. The platform now consistently uses periods for all currency values and decimals.
Minor Style Enhancements for Batch Tables
Several minor style enhancements have been made to the batch tables, improving their overall appearance and formatting.
Preserved Prosody when Changing Voices of the Same Gender
An issue has been fixed where switching to another voice of the same gender caused the prosody of the text to be lost. With this fix, prosody settings are now preserved, ensuring a consistent audio experience when changing voices.
v2.4
Release Date: 2024-09-05
New Features & Additional Library Voices
Additional Library Voices
A new set of 12 voices has been added to the library, offering users a broader selection to choose from.
Adding and Deleting Lines during Source Review
Users can now add or delete source lines during the review phase.
Independent Language Progression
Users can start working on each target language independently, even if all target languages haven't been uploaded yet.
Script Downloads: Line-Based and Subtitle-Based for SRT, VTT, and TTML Formats
Users now have the flexibility to download scripts in either the default line-based segmentation or subtitle segmentation formats.
Last Edit Info on Takes
Details of the last edit on each take, including the username and the date and time, will now be displayed.
Comments on Takes
Collaborative comments can now be added to takes.
Full Text Display
Users can now click on a line to expand and view the full source and target text on the Preview Page, instead of hovering to see it in a tooltip.
Glossary Audio Samples
Users can now listen to audio samples of terms added to the glossary.
Merging and Deleting Audio Splits
Users will now be able to merge and delete audio splits in the Preview timeline. After splitting an audio chunk, users have the option to merge it back. Moreover, it will also be possible to delete splits of the audio and restore the original.
Integration of Background Music and Vocal-Only Tracks
Users now have the option to upload additional tracks, such as background music and vocal-only tracks, which are integrated into the Asset Management page and available as source tracks in the Timeline.
Source Audio Segmentation Delimiters
Vertical delimiters are now displayed in the Preview Page timeline to improve sound syncing.
Enhanced Search Bar on the Preview Page
The search bar is now extended across all pages, rather than being limited to the selected page.
Removal of Grayed-Out Characters from the Casting Page
Characters created during asset ingestion without any assigned lines will no longer be visible on the casting page.
v2.3.1
Release Date: 2024-07-25
Bug Fixes & Performance Enhancements
This release primarily focuses on bug fixes and performance enhancements, specifically including:
Faster Batch Creation
The process for creating target lines and assigning voices after casting or voice change is now significantly faster.
Bug Fix for Glossary Matching
A bug causing an erroneous warning message for glossary matching when the glossary was loaded before asset ingestion has been resolved. The warning no longer appears when there are no mismatches.
Improved Acronym Management
Acronym management has been enhanced, with automatic recognition now limited to words consisting of three uppercase letters.
Source Review Page Optimization
The Source Review page now includes a Line ID and a Search Bar. The search function now scans all pages, not just the displayed page. Additionally, changes made in the Source Review are accurately reflected in the Preview Page (starting offset).
Input and Output File Formats
Input and output file formats are now aligned and include: TXT, SRT, VTT, TTML, DOC, and DOCX.
Starting Time of Source Audio in Preview Page
The starting time of the source audio segment is now displayed under the source text when a line is selected.
v2.3
Release Date: 2024-06-28
New Features & Additional Languages
Additional Languages
New Languages: Simplified Chinese, Japanese, Korean, Turkish, Hindi, and Portuguese (Portugal).
Currently the supported output languages are: US English, UK English, French, Italian, German, Spanish (Spain), Spanish (LatAm), Portuguese (Brazil), Portuguese (Portugal), Polish, Simplified Chinese, Japanese, Korean, Turkish, and Hindi.
Timeline in the Production area
Now you can streamline the generation of long content with multiple speakers (podcasts, e-learning content…).
Multiple Take Generation
During batch creation, three takes are automatically generated, and ready to be played when you enter the Preview Page.
Zoom in Timeline and Fade-in & Fade-Out for Audio Chunks
Now you can zoom in the timeline to accurately sync or split the target audio chunks. Additionally, fade-in and fade-out are automatically applied when you split the audio chunks.
Adding Terms to Glossary directly from the Text Editor
Now you can add new terms to the glossary directly from the Text Editor on the Preview Page.
Manually Editing Start and Stop Times from the Review Page
In addition to editing the start and stop times of each audio chunks by dragging the edge of the chunk, now you can manually edit these values in the dedicated field next to the text.
v2.2
Release Date: 2024-05-02
New Features
Two-Step Asset Ingestion & Asset Management Page
Users now have the option to upload source text and target translations asynchronously, with all assets easily managed on the Asset Management page.
Source Transcript Review
Users now have the option to review and edit the source transcript, and adjust the start and stop of the each individual audio chunk.
Glossary Update
Users can now upload a glossary in Excel file format, allowing them to create and/or update the glossary with just one click.
Split Audio Feature
Users now have the option to split individual audio chunks in the timeline, allowing them to extend the duration of internal pauses within a segment.
v2.1
Release Date: 2024-03-11
New Features
Video Integration in the Timeline
Video functionality has been incroporated in audio/video batches. The video player follows the timeline. It is possible to expand the video to full screen, having a view of only the video together with the timeline.
Rework Line
Users now have the ability to rework bounced lines. When a bounced line is chosen for rework, it will no longer appear in the bounced list but will instead be found in the ingested list with a "Rework" label.
Adding pauses in timeline
It is now possible to move multiple segments together in the timeline, facilitating the addition of pauses within the timeline structure.
Resyncing Timeline
By selecting the resync button on the video, all audio segments will be realigned to their original positions based on the start time of the source segment.
Scrolling pages while playing the timeline
During playback, the Preview page dynamically scrolls and changes pages based on the audio being played. Similarly, clicking on an audio segment in the timeline will prompt the timeline to adjust accordingly. If a user selects an inferred line in the Preview, the timeline will shift to the corresponding audio segment.