MCR MV Sebiu Labs Early alpha

MCR MV

Master control, from one box to a whole facility. One system for live production, master control and FAST channels, with captions, translation, AI voices, comms and replay built in, all run from a web browser.

Test the early alpha → ↓ Feature guide (PDF, 10 MB)

If you would like to test an early alpha release, please get in contact: contact@sebiulabs.co.uk

Watch: never go to black

One system, every job

Most facilities buy these as separate products, each with its own logins, alarms and support. MCR MV does them with one panel, one alarm bus, one as-run log and one API.

Production switcher

Preview and programme with cut, dissolve, wipes and stingers, 8-box DVE, chroma key, scenes, macros and a second M/E bank.

Graphics

Six keyed layers, 23 animated templates, your own brand and fonts, web pages and NDI with alpha, and live data, run from a one-press Board.

Instant replay

Mark the last few seconds, every angle on the same frame, any speed, true slow motion from high-frame-rate cameras, and reels.

Multiview

Broadcast layouts with tally and meters, in the browser over WebRTC, on a gallery screen over NDI, or as an output of its own.

Master control

Primary and backup failover, self-healing inputs, EBU R128 or ATSC A/85 loudness, alarms in plain English and an as-run log.

FAST and playout

Playout decks, a rundown from the library, traffic, SCTE-35 breaks, house-format conform, programme guide exports and HLS with ABR.

Captions and translation

Live CEA-608/708, WebVTT, HLS subtitles and burn-in, with 28 languages in the recognition catalogue, following the speaker when the language changes.

AI voices

A dubbed track per language, mixed like an interpreter under the speaker, on the desk and on every output, and labelled as AI.

Audio desk

A strip for every source, buses, follow mixes, international clean and a 5.1 programme with AC-3 delivery.

Recording and highlights

Programme and ISO recording, quick clips, and a Highlights page that says why each moment was picked.

Sources and outputs

SRT, RIST, RTMP, NDI, HLS and WebRTC in, browser guests from a link, and SRT, RTMP(S), HLS and NDI out, added or dropped on air.

Scale out

Start with one box, then add render, transcode, AI, ingest and NDI matrix nodes. Logins, roles, an API and a Companion module are built in.

Inside the panel

The switcher: preview and programme, the source bus, transitions, DVE and TAKE.

Multiview: every source with its label, tally and audio meter.

Graphics Play: the Board, one press per graphic, beside programme and the layers.

Graphics Make: build a DVE scene without touching air.

Instant replay: marked moments, two channels cued, frame-accurate angles.

Captions: every language with its captions, dub and burn-in on one row.

Highlights: what the channel found, what counts, and why.

The audio desk: a strip for every source, with buses and programme processing.

Runs on CPUs, with AI where you choose

Switching, graphics, the audio desk, recording, replay and every output run on ordinary CPUs, in the cloud or on your own servers. Operators and remote guests install nothing.

Speech recognition, translation and voices run locally on an NVIDIA graphics card, or through cloud providers with your own API keys, chosen per service. With every service local, programme audio never leaves your network.

Locally, translation from English covers 7 languages and voices cover 8, so a fully local dub works in five: Spanish, French, Italian, Portuguese and Mandarin. Other languages use cloud services. Recognition accuracy has been measured for English and Spanish so far.

Read the whole guide

The feature guide covers every feature area in 42 pages, with what is still being proven marked as it is.

↓ Download the feature guide Get in contact

Screens are from a demo channel; pictures, names and text in them are examples. Pictures from Wikimedia Commons: “Dawn of Darkness — Live Closing Passage” and “Colorful Autumn — Stage Excerpt”, Mortaz Band, by Mostafa Meraji (CC0); “Aerial imagery of Zug Switzerland” and “Cars drifting around a race track at twilight and night as viewed from FPV drone” by anotherview (CC BY 3.0); “Close-up fireworks during Lunar New Year celebration in Zhongshan, China, 2026” by JaydenChao (CC BY 4.0); “Rome marathon, via Galvani, 2017” by Husky (CC BY 4.0). Video pictures: Big Buck Bunny, Caminandes © Blender Foundation, CC BY.

NDI® is a registered trademark of NDI, Inc. AC-3 is named by its standard designation. Other product names are trademarks of their owners; MCR MV is not affiliated with or endorsed by them.