๐Changelog
What's new in the Checksub Editor
Last updated
What's new in the Checksub Editor
Last updated
Features: -- Introduce term bases for translations and dubbings. With the new Term Bases you can make your translationmore consistent from one project to an other. We're the first one to introduce a Term Bases with pronunciation capabilities. Now you can decide how to translate a word and how to pronunce it when you generate a dubbing. We're automaticaly change the way to write the word when we generate the dubbing.
When it comes to translation features, we're going one step further than a basic search and replace function. We'll be able to modify your word even if it's plural or masculine and feminine. This approach ensures stable quality and avoids the need to add terms in all their written versions.
-- Allow to use multiple languages in API requests (languages_target) -- Allow to export in other subtitle formats in the API (itt, vtt and others)
Updates -- Improve the accuracy of the diarization
New feature: โ We're happy to introduce a new amazing feature. You can new give a prompt and timing parameters to improve the quality of your transcription, translation and subtitles. We give you more information about this feature that will allows you to improve the quality of your project in this section.
Bugfix: -- Prevent holding the Ctrl key from triggering the splitting shortcut (double Ctrl press)
New feature: -- Auto-download exportations Bugfix: -- Fixed a bug that led to a crash for shared links when the assets of the project were not yet processed
Select "Create voices from a text" and you will be able to import a script or write it sentence by sentence in the editor.
If you wish, you can add a video to your project to merge your voice-over with the video. To do this, click on the video upload icon in the top center or the image icon, which lets you reuse a video already added to your library.
For those who want the best quality, it's sometimes necessary to regenerate a sentence. But what if, after several generations, you want to go back to a previous version? Until now, this has been impossible for the vast majority of voice generation solutions.
Checksub now offers versioning of generated voices. You can generate several times the voice and always have the option of reverting to an earlier version if you prefer.
-- Allow Emotion Recorder for sharing links -- Remove variables from API voice generation when regenerating a voice -- Support presigned URLs for uploading a video Bugfixes: -- Fix an error in the speed section of voice blocks -- Fix an issue with the dashboard automatic update -- Fix an error occurring when trying to save changes to voice blocks with a sharing link
It is now possible to use voice cloning, library voices or not to dub a speaker within the same project. This opens up new possibilities, for example, if you have permission to clone the voice of one person but not another in a video. You can also choose to dub the voice of one person but not the other.
You can choose the dubbing parameters for each caller in the project.
When you generate automatic dubbing, it can happen that a sentence doesn't have the intonation you want. In such cases, it's difficult to adjust the result to obtain the rendering you have in mind.
Rather than having to play with sliders to adjust the emotions and prosody of your sentence, we now let you record your voice with the intonation you want, right from the editor.
We will then copy the emotion of your voice and apply it to the voice used in your project.
PS: This option is only available with cloned voice at the moment.
New features: -- Add Hindi premium voices -- Update the Ctrl+A behavior for subtitles -- API: Get the api_request_id from a project_id
Improvement: Increase the handler to resize the timeline Allow to replay the video when the playhead has reached its end Avoid to jump to the beginning of the editor after deleting a cue Improve the tooltip for low recognition New features Add buttons for zooming in/out the timeline
Bugfix: Removed xh-ZA from translation languages
Bugfix: -- Prevent auto scrolling to the active cue on certain events, causing a black screen Improvement: -- Update the Watermark when we export a video with the free trial
Bugfix: -- Prevent the payment button for subscriptions being blocked for a trialing subscription without credit card
Bugfix: -- Fixes a bug that could lead to lost edits when going backward and forward in the browser (or with the two-finger swipe on a Mac trackpad)
Modify multiple speaker IDs in bulk in the subtitles editor. Now you can select and edit in bulk the speaker id of multiple subtitle block. This comes in very handy if you want to generate a dubbing afterwards.
Compare the translation with the original text in the dubbing interface Improvements: -- Some improvements concerning the opening/closing behavior of the source text and the right panel
-- Remove the livechat bubble from the editor -- Improve the timeline according to the new design of the editor -- Improve Somali transcriptions Features: -- Add new features to the external API (list projects, list exportations, start an exportation)
Allow to start a free trial without payment details
A new UI for the Checksub editor ๐ We hide the timeline per default, we're using a overlay menu to edit your subtitles and dubbing, and we clarify the subtitles editor and Dubbing editor. Features: -- Allows to "play" subtitles even when there is no video/ audio asset Bugfixes: -- Fixes the bug that could happen when replace "something by nothing" in Search & Find -- Fixes a bug that lead to the audio not being updated when replacing an asset
Now you can use the new tabs to switch from Subtitles to Dubbing
๐ Refine translation to improve your dubbing (Enterprise Option)
Now we're able to pick the best translation for your dubbing. What's the best translation for dubbing? Like the movies for dubbing you don't need a word to word translation. You need to take into account the duration of the initial sentence to pick the best translation. Now thanks to our new AI we will automatically pick the best translation to generate a perfect dubbing.
A better voice cloning
Now we implemented a more stable voice cloning technology for AI-Dubbing.
A new way to adjust the speed of your dubbing
You can now adjust voice speed more precisely. Click on the Adjust speed block in the right-hand menu.
If the voice block in the timeline doesn't have enough space to say the entire sentence, an orange border is added. You can then modify the text, move the block or adjust the speed of the voice to fix this.
More languages supported by Checksub ๐
Adds new languages for transcriptions, for example, Somali and Luxembourgish
Shows a clearer error message when already regenerating a voice block
Fixes an issue with the alignment of the loader for regenerating blocks
Fixes an issue with a disappearing video element when moving and releasing it
Now you can Clone your voice and generate a foreign dubbing ๐คฏ
Allow to clone voices in 8 languages
New live chat code
We improved the synchronization mechanism between video, subtitles, and even voices. Now you can see the subtitle ID when you select a subtitle in the video area.
New synchronization mechanism between each element
Added the index in the timeline and the video area
Fix an issue to close the right panel
Fix the buttons in the header for small screens
We added 200 new AI voices for dubbing. We now allow you to delete one subtitle with one click. We now make the subtitle ID visible in the editor to make the editing easier and faster. We optimize the workflow to create new projects by using the subtitles settings of the last project.
General improvements & features:
Preselects the last used languages
Editor improvements & features:
When pausing without active cue, shows the cursor at the end of the previous cue
Shows the cue id in the text editor and selector
The active cue can be deleted from the text editor by clicking on the trash bin icon
Updates generating blocks every 5 sec
Adds more than 200 voices
Editor bugfixes:
Fixes a bug concerning the cursor position when pausing between words
Improve the AI-Dubbing with an audio isolation feature. We can separate your original voice from the audio background (music, ambient noise,...). For example, you can delete the original voice and keep only the music.
Bugfix:
General improvements & features:
Improves the language input
Simplifies the process to generate new subtitles
Uses the subtitle settings from the last subtitle of the last project version with the same format (horizontal or vertical/square) for the initial splitting
Makes the review of the transcript optional
Makes audio separation optional
Editor improvements & features:
Blocks scrolling in the right panel when a dropdown is open
Brings back the original style of the line at the top of the timeline
Places dropdowns always on top so that they are always visible
Changes the colors of the audio separation assets
Editor bugfixes:
Only deletes old audio files when the regeneration of a voice block was successful
Saves unsaved changes before forwarding when the โReturn to dashboard buttonโ is clicked
Forwards the user to the dashboard when creating a transcript from the editor and reloads the editor when adding translations
Allows to export SRT from a sharing link
Bugfix:
Fixes a bug with Karaoke animations (font size is shown correctly)
Fixes an inconsistency between the dashboard and editor concerning the permissions to export
Transcription improved by 20%
This new version improves our speech-to-text transcription for English, French, Danish, Swedish and German.
It means fewer edits for you to get perfect captions. โ๏ธ
New AI-dubbing voices in Chinese
Shows a country selection for voices
Adds voices for Chinese
We added 14 new languages
Bashkir, Basque, Belarusian, Esperanto, Estonian, Galician, Interlingua, Marathi, Mongolian, Tamil, Thai, Uyghur, Vietnamese, and Welsh.
That's why we invented the emotion recorder ๐จ.
