“Lightweight” Doesn’t Mean Less

Why hybrid IP is the destination, not a detour

Jacob Zuo, Founder and CTO

For a decade, two curves in broadcast production have moved apart. Demand keeps climbing — more formats, platforms, and live events with shorter cycles. Resources run the other way — advertising shifts to streaming, budgets tighten, audience attention fragments. This is not cyclical but structural, which closes the old escape route of overbuilding for revenue that no longer comes. Broadcast is entering an era of resource discipline: every dollar of production capability must answer for its utilization. The industry assumes only two answers — spend more and stack equipment, or scale down to “good enough.” We believe that “lightweight” production claims the trade-off is false.

What lightweight is not, and what it is 

Three misconceptions travel with the word “lightweight”. It does not mean smaller, cheaper gear; that is going down-market — you save on procurement and lose reliability and signal quality, which is weaker, not lighter. It does not mean cutting professional capability; fewer camera positions and rougher output walk away from the problem rather than solve it. And it does not mean going all-IP or all-cloud; pushing every signal onto ST 2110, or lifting the whole system into the cloud, only swaps cabling complexity for network engineering, and hardware maintenance for software plus bandwidth costs. The load on the operator does not lift, and sometimes it grows.

So what is it? A system’s total complexity never disappears — it only moves. Lightweight production redistributes it: away from the venue, operator, and cabling; into the architecture, software, and network. Smaller equipment is a by-product, not the method. The real test is whether the operator’s complexity has fallen — fewer devices, simpler paths, faster response to change. Traditional infrastructure carries an original sin: built for peak, its capability locked into fixed rooms, paths, and single-purpose devices that mostly sit idle. Move complexity into the architecture and capability stops being a fixed asset and becomes a resource that can flow and recombine on demand. Those are the two faces of lightweight: lower complexity, the cause; and higher utilization, the measurable result.

Three markets, one shift 

North America, Europe, and China sound different but are three faces of one change. North America talks REMI, remote production, and ST 2110 — decoupling operators, equipment, and venues over IP; strip the acronyms and it is resource fluidity. European public media talks about moving beyond linear workflows — one team and system serving many platforms at once, forcing complexity off people onto architecture. China named the shift and wrote it into policy: in April 2026 the National Radio and Television Administration issued its Technical Application Guide for Lightweight Production and Playout Systems (2026 Edition), setting the direction as lightweight, integrated, and intelligent — the goal is to not lower standards but more adaptable professional production through architectural change. It is the first time “lightweight” has crossed from vendor shorthand into an officially defined roadmap.

A judgment some readers won’t like: hybrid is the destination, not the detour. 

The mainstream story treats hybrid IP as a temporary way station to all-IP, and SDI as technical debt. Our view is the opposite. First, SDI has not earned retirement on critical links: its determinism, zero configuration, and fault tolerance are still the cheapest reliability on those links — technology does not retire what is still the best solution. Second, the economics of full ST 2110 — PTP timing, non-blocking fabrics, dedicated network teams — never close for most mid-sized broadcasters and corporate users; a road only the top of the pyramid can walk is not the industry’s direction. Third, and most important, the question is wrong: the endgame is not one protocol winning but protocols ceasing to matter. Once an architecture layer treats SDI, NDI, SRT, and ST 2110 as interchangeable resources to route and reuse the question of, “which protocol” drops from strategy to engineering detail. The next competition is at the architecture layer, not the protocol layer — and whoever schedules a heterogeneous signal world as one ends the protocol war not by winning it, but by making it irrelevant.

Three layers of getting there 

Signal layer — connect, don’t replace: bring existing SDI resources into the IP domain at minimal cost so they become schedulable objects, not scrap; the test is whether old operating habits survive. Function layer — modules, not more boxes: instead of a dedicated box per requirement — each with its own power, management, and signal path — functions like encoding, transcoding, NDI processing, multiview, and remote receiving become service cards combined on demand inside shared infrastructure, embodied by Kiloview’s RF02: up to 18 cards in a 2RU chassis, assembled around the workflow. Capability layer — software-defined, not hardware-bound: once a compute card lives in the same framework (such as our FXC-A8800 in the RF02), recording, monitoring, graphics, intercom, and future AI workloads no longer need their own hardware tier. Functions stop being equipment you buy and become software you deploy.

One proof: Beijing Radio and Television (BRTV). Before its upgrade, Beijing Radio and Television ran 16 studios as SDI islands, sharing little, coordinated by hand. The upgrade kept the SDI foundation, integrated studios through a modular framework, pooled signals through centralized routing (Kiloview NDI CORE MAX), and brought field and OB feeds into one scheduling system via bonded encoders and IP converters. Three numbers followed. Studio utilization rose over 80% — fixed asset turned flowing resource. Operating and maintenance costs fell over 50% — complexity redistribution on the books. Bringing a field signal into production now takes minutes — recombination speed, once adding a signal is no longer a cabling job. After the upgrade: SDI still there, studios still there, and habits still there — running stably in hybrid. Hybrid is not an unfinished road; it is the destination.

Light means wasting less. The gap between demand and resources will not close — which is why the yardstick is changing. Within five years, what marks a production system as advanced will not be which protocols it supports or how many signals it carries, but its utilization and its speed of recombination. Light has never meant owning less — it means wasting less: less idle capability, fewer stranded signals, fewer single-purpose devices, with less complexity on the operator. That is the path we have chosen: not betting on any protocol to win, but building the architecture layer that schedules every protocol and every existing resource as one. When the industry finally understands “lightweight,” it will not describe a smaller device — it will describe a way of producing that wastes less.

---

This article was written by Jacob Zuo, which contributed to the drafting of the National Radio and Television Administration’s Technical Application Guide for Lightweight Production and Playout Systems (2026 Edition).