Changelog #1 - Multistreaming for Deep Live Hub

Welcome to the first episode of our monthly changelog newsletter. Get a quick and clear overview of all technical updates and platform changes. Looking for customer stories and more entertainment? subscribe to our editorial Newsletter via mail and for deeper technical insights, visit our documentation at docs.deepva.com.

DEEP LIVE HUB

We have updated the Deep Live Hub of our DeepVA composite AI platform and are pleased to introduce the latest improvements and features. This update focuses on usability and significantly increases the flexibility and security of this module. As our single solution for all live streaming functions, we are currently not only bringing the file-based Deep Media Analyzer capabilities to the Deep Live Hub, but are also constantly improving the current toolset for our broadcast and live event customers.

New function: multi-streaming for more destinations

A major novelty is the support of multiple RTMP pushers at the same time. This allows you to broadcast live streams in parallel to platforms such as YouTube Live, Facebook Live and Twitch at the same time - without the need for external tools. This function is directly integrated and can be used by now, ensuring efficiency and saving you time and money for restreaming.

Use Case: An event organizer can now broadcast a conference to several platforms simultaneously by starting the stream to our platform once and we are automatically distributing it to all desired destinations. This guarantees maximum flexibility, more usability and less additional tools.

More user-friendliness thanks to reusable subtitle settings?

Settings for subtitles, whether “Burned-In” or “Closed Captions”, can now be saved and reused in later projects. This saves time when editing similar projects and ensures that subtitles and formatting always look the same.

Improved security with RTMPS and authentication?

For more security, the Deep Live Hub now supports the RTMPS protocol. In addition, the user name and password can be transmitted directly as part of the URL or in the parameters, which simplifies setup and better protects your streams from unauthorized access.??

Stability and speed at focus

The components of the streaming pipeline have been technically revised and stabilised, while the player in the editor is now more reliable, especially during longer streaming sessions. A new error handling routine also allows us to better handle network problems, so that editing is possible even with clients that are not well connected, such as mobile device at a trade show.

Improvements: Optimized file management and more stable player?

In batch workflows, transcribed files such as SRT, AMT and WebVTT can now be easily deleted. This ensures clearer project management and reduces unnecessary data waste.??.?

Bug fixes:

• Stability problems with live streaming have been solved.??

• Pop-up windows can no longer be closed accidentally.??

• The quality and speed of the SRT have been optimized.??

• The editor remains stable even if the streaming worker crashes.??

• Video pull configuration issues have been resolved, and the platform code has been standardized to ensure consistency.?

DEEP MEDIA ANALYZER

Improved language support and new models

The Deep Media Analyzer now supports Semitic languages such as Arabic and Hebrew, which are written from right to left. This is now also working in the transcript editor.

Face Recognition Update: A new model has been released (version V36), which enables the recognition of additional personalities from politics, society and sport.??

Outlook: Further functions in the new year?

Look out for more exciting features next year - the next releases and new features are already in testing, including transcription based composite AI workflows for creating new exports and advanced metadata generation. Happy holidays from all the DeepVA team!

All DeepVA changelog updates are also available here:? https://docs.deepva.com/changelog/?