Apple and DASH Forum target sub-ten-second glass-to-glass latency for live streaming

Reducing the delay between the venue camera and the home screen is now a commercial requirement for high-value sports rights. As betting and interactive features become central to monetization, the current 40-to-80-second lag seen in major events like the Super Bowl creates a synchronization gap that alienates viewers. This technical compression forces a shift in advertising, moving from just-in-time auctions to pre-fetched creative to fit within shrinking decision windows. The industry must now balance this need for speed against the stability provided by larger buffers. Watch for the finalization of the Media over QUIC specification to see if sub-second broadcast scale becomes viable.

The DASH Industry Forum has been working to standardize low-latency streaming approaches since its founding, with Apple's Low-Latency HLS and the CMAF-based chunked transfer encoding approach representing the two dominant paths. In 2024, the DASH Industry Forum published updated guidelines for low-latency DASH implementations, specifying segment durations, chunk sizes, and signaling requirements that enable sub-10-second end-to-end delivery at scale. Apple, meanwhile, has continued iterating on LL-HLS since its introduction in 2019, with the protocol now supporting partial segments and blocking playlist reload to reduce latency without requiring a separate protocol stack. The convergence of these approaches under CMAF packaging means that a single encoded rendition can serve both ecosystems, a point the DASH Forum has emphasized in its interoperability work.

On the commercial side, the IAB Tech Lab has been developing standards to address the advertising implications of low-latency streams. The IAB Tech Lab's Server-Side Ad Insertion working group released updated specifications in 2025 that account for reduced decision windows in live environments, including pre-fetch mechanisms and SCTE-35 signaling adaptations for sub-10-second contexts. Magnite, one of the largest independent sell-side platforms, has been testing server-side ad insertion workflows that can operate within compressed latency budgets, a capability that becomes essential as sports rights holders demand parity between broadcast and digital ad experiences. The challenge is that traditional just-in-time ad auctions require 200-500 milliseconds of decision time, which consumes a significant portion of the available buffer in a sub-10-second pipeline.

Technical benchmarks from independent testing have shown that the gap between theoretical minimums and production reality remains significant. Mux's annual latency report found that median glass-to-glass latency across major streaming platforms was approximately 30 seconds in 2025, with the fastest implementations reaching 8-12 seconds under optimal conditions. Phenix Real Time Solutions has demonstrated sub-500-millisecond delivery using WebRTC-based approaches, but those solutions have not yet achieved the CDN-scale economics needed for events with millions of concurrent viewers. The Media over QUIC working group at the IETF is developing transport-layer specifications that could bridge this gap by enabling multiplexed, encrypted delivery over UDP without the head-of-line blocking that affects TCP-based protocols, potentially making broadcast-scale sub-second delivery viable within the next two to three years.