The developers behind the open-source video editor Kdenlive have brought AI learning into 2023's new release with the addition of openAI's Whisper speech-to-text engine. This and a handful of other new tent pole features make the newly released 23.04 version of Kdenlive even more exciting.
Kdenlive 23.04 Is Here
Kdenlive announced some eagerly anticipated new features in the software's latest release. It's been an eventful year for the developers, with 23.04 being the first major release since their successful fundraising campaign earlier in 2022.
They had some ambitious self-directed goals to hit with this release, including nested timelines, improved speech-to-text, on the fly English translation, and the usual host of bug fixes and code improvements.
Download: Kdenlive 23.04 for Windows, Mac, and Linux (Free)
The New Features to Try in Kdenlive 23.04
There is a lot to be excited about with this release. With the biggest and most anticipated being the announcement of nested timelines and the addition of the OpenAI Whisper speech engine.
Nested Timelines
Nested timelines aren't anything new, editors have been using nested sequences in Premiere Pro for some time. Now, with 23.04, Kdenlive users are able to similarly streamline their work.
Essentially, a nested timeline allows an editor to work on their project in smaller chunks (called sequences) rather than one long, often ungainly, timeline. Before nested timelines came around, an editor could edit a particular sequence, finalize it by rendering it out, and then reimport the now completed sequence back into their project.
But that came at a cost. Finalizing the sequence is exactly what it sounds like. Once that sequence was rendered out and reimported, it could no longer be edited.
Nested timelines fix that. Finalizing and then reimporting each individual sequence is no longer necessary. Separate timelines can be created for each sequence, with a main timeline consisting of those sequences rather than the individual clips.
The individual sequences not only remain editable in their own nested timeline, but any changes made to them will automatically reflect in the main project as a whole.
OpenAI Whisper Speech-to-Text
Whisper is a voice-to-text tool created by OpenAI, the team responsible for ChatGPT. Free and powered by deep learning and neural networks, the project promises a more natural, and therefore more accurate, language processing system for use in speech-to-text applications. The developers of Kdenlive have added it into 23.04 to support the automatic subtitles feature.
While Kdenlive has an existing subtitle engine, in VOSK, the addition of the Whisper speech-to-text protocol promises a marked improvement. One of the main selling features of Whisper is not only its increased accuracy but its ability to translate audio to English on-the-fly.
If an editor has a scene that contains a foreign language that requires subtitling, theoretically, Whisper should be able to do it without missing a beat.
The developers have made an effort to make Whisper simple to configure by simply downloading the required files through the configuration screen. And as simple as possible to use by creating a subtitle track in your project and then selecting the magic wand and following the prompts.
Timer
The new Timer effect received probably the least hype in the announcement of 23.04. It's a useful feature that looks deceptively simple, but in reality, has been a somewhat complicated manual procedure up until now.
Like most NLEs, of course, Kdenlive includes the ability to view the project's time code, but it is not customizable and syncs only to the project as a whole. Creating a timer that would automatically count up or down and be synced to a single clip or a single track was an in-depth process involving a number of different steps.
With this new update, the Timer can be added like any other effect with a drag-and-drop, and can be edited quickly and easily through the Effect/Composition Stack next to it.
Adjustable parameters include font, size, color, position, and anything else one would expect.
What's Next on the Roadmap?
As well as the new features, Kdenlive 23.04 includes the usual bug fixes and polish for 2023's release. Those include improving the performance of clips in the timeline, better importing of clips with large file sizes, improved subtitle handling, and a host of new transitions for editors to play around with.
Developers noted in their announcement that for the next release, they were planning on working on the effects workflow while working towards improved GPU support.