Profuz Digital showcases AI-Orchestrated Profuz LAPIS And Workflow-Driven SubtitleNEXT

European developer Profuz Digital will present at IBC Stand 8.A88, the latest version of subtitling platform SubtitleNEXT and new capabilities in the powerful AI-enabled Profuz LAPIS digital media asset management system, from 12 – 15 September at The RAI in Amsterdam.

Already adopted by many organisations such as Canal+ France, Council of Europe, Bulgarian National Radio and others, Profuz LAPIS delivers a revolutionary solution to the industry, loaded with every tool and feature required to manage multiple TV channels. It is designed to organise media assets that include movies, documentaries, series, clips, and promotions for categorisation, advanced searching, filtering, meta data tagging and sorting, by effortlessly providing centralised, secure, and reliable efficiency in one powerful elegant solution, streamlining workflows, and boosting productivity.

Profuz LAPIS is more than just another digital media asset management system that integrates AI—it is a sophisticated orchestration platform for AI-powered tasks. Instead of simply running different AI engines independently, LAPIS intelligently choreographs an organisation's entire workflow by distributing and coordinating tasks across multiple AI processing engines.

Traditional AI-powered media asset management systems tend to treat AI engines as separate tools, resulting in a more fragmented outcome that requires time-consuming manual intervention. LAPIS, however, harmonises AI capabilities into a unified workflow, where data is processed in layers and enriched at every stage. LAPIS brings cumulative benefits rather than working in isolation, so that each AI engine’s output enhances overall results. LAPIS combines data from multiple sources — such as speech recognition paired with face detection in order to create richer and more accurate insights.

The new 'Multi-AI Processing Integration' feature in LAPIS manages seamless connection with various AI engines for transcription, translation, object detection, face recognition and more. Delivering optimised performance by distributing workloads efficiently, LAPIS minimises processing time while maximising accuracy, ensuring organisations work faster and smarter.

The new 'Smart Task Distribution' in LAPIS selects the most suitable engine based on metrics such as language, category, and genre. It eradicates the need to rely on one single AI engine, and instead, LAPIS now analyses the nature of each task, before assigning it to the most suitable available engine. For example, “speech-to-text” tasks may be directed to a particular specialised engine, while another engine might manage more complex translations.

Other new features and capabilities in LAPIS to be showcased at IBC include - subtitle translation with intelligent time slot management; voice synthesis obtained from timed texts; speaker identification within speech-to-text; multiple optional servers for AI and media processing such as transcription, translation, object detection, face recognition, text extraction, conversion, transcoding and others, that include intelligent distribution of tasks. Other functionalities include image, audio and text file editors, the volume meter and slow motion control tools in the media player, and a brand new timeline media editor plus much more.

SubtitleNEXT is an affordable timed-text software platform that is intuitive and manages subtitles and audio description. The latest version (5.16) will be demonstrated at IBC, packed with an extensive host of stunning new features. 'Batch translation' of subtitle files is now made possible in the latest version of SubtitleNEXT, allowing subtitlers to automatically translate multiple subtitle files in bulk. Another significant addition is the new text splitting mechanism enabling the correct translation of subtitles into all languages.

The Representational State Transfer Application Programming Interface (REST API) for integration with other platforms is now available too. REST API allows different systems to talk to each other over the web. It exposes multiple endpoints for interaction with SubtitleNEXT and allows control of the operations mode, speech recognition, subtitle files' status, conversions and output, media source and playback and more.

Another significant upgrade for broadcasters includes SubtitleNEXT’s support for SDI and NDI-based workflows. The new Decklink SDI input and output capability now makes it possible for SubtitleNEXT users to experience low latency high-quality graphics, teletext and closed caption subtitles. Graphical subtitles can be transmitted as fill and key, burnt over pass-through video with hardware or software keying. Teletext (OP-47) or closed caption subtitles can be inserted in the pass-through video signal. NDI input for video preview and overlay and output with fill and key, or mixing over incoming feed, allows real-time video streaming with low latency using regular IT Infrastructure, making it easier to integrate subtitles into live broadcasts without the need for additional hardware.

The new Volume Meter enables audio level monitoring of specified audio sources within the SubtitleNEXT application, resulting in cost-savings with no need for separate monitoring applications. SubtitleNEXT offers improved .890 format handling and also supports TTML based formats and profiles such as SMPTE-TT, IMSC 1/1.2, EBU-TT and others.

SubtitleNEXT has a proven track record of success with organisations already embracing it, including Proximus Media House, London’s Wilco Media, Euro Media Group (EMG), Polsat, Bulgarian National Television, IT Pros Italy, Digitalmeister, Leinhäuser Language Services, University of Warsaw, Ghent and KU Leuven University, and others.

subtitlenext.com