Why Video Conferencing Is the Toughest RTC Challenge and How We Overcome It

RTC’s Unique Challenges in Meeting Scenarios

Video conferencing is considered the most demanding RTC scenario because it requires handling large numbers of participants, free microphone and camera control, multiple layout options, screen sharing, and must work under weak networks, low‑end devices, noisy environments, and poor lighting.

Complex Lighting Video Experience

We introduced a camera pipeline that adds automatic focus, exposure, and a face‑aware exposure algorithm (FaceAE) to optimise brightness for faces in back‑lit or low‑light conditions, achieving better visual quality without extra performance cost.

Screen Sharing Optimization

We developed an intelligent encoding mode that automatically detects the type of shared content (static documents vs. video) and switches between high‑quality and smooth‑playback profiles, eliminating the need for users to manually adjust settings.

Multi‑Grid View Enhancement

To support up to 49‑grid layouts we designed a 10‑level Simulcast system that aggregates streams into four tiers based on subscription weight, providing smoother resolution downgrades and improving average resolution by up to 16% while keeping latency stable.

Key Technologies for Meeting Control

Reliable, low‑latency signalling is achieved by reusing the RTC media transport path, delivering 98.6% of signalling messages within 200 ms and maintaining 100% delivery even under weak network conditions, ensuring audio‑video state consistency.

New Solutions for Complex Algorithms on Web

Edge rendering moves heavy video‑processing (beauty, virtual backgrounds, etc.) to the cloud edge, adding only ~30 ms latency and keeping client CPU usage minimal, enabling high‑quality effects on low‑end browsers.