LCEVC-enabled video messaging

How can video messages dynamically deliver the best experience for each recipient, based on device capability and network conditions?

MPEG-5 LCEVC alongside your existing codecs such as AVC and HEVC provides two levels of quality directly from the sending device for an optimized adaptive experience.

The challenge

Video messaging is increasingly processed entirely on the sender’s device. Driven by privacy, latency, and architectural requirements, encoding happens at capture time and the resulting file is what every recipient receives, typically with no cloud transcode step to adapt the content after upload.

This pushes services into a lowest-common-denominator delivery model. One video file must work across a fragmented mix of devices, operating systems, screen sizes, and network conditions. In practice, that often means sending a low-resolution version, typically 360p or 480p at around 1.3 to 1.7 Mbps, because at the moment of sending, services have to prioritise fast, reliable delivery over higher visual quality.

Once that choice is made, it’s effectively locked in. A recipient with a flagship device on Wi-Fi gets the same low-resolution file as someone on a budget phone over cellular, with no practical way to selectively improve quality later for capable devices without re-encoding and resending the content.

The solution

MPEG-5 Part 2 LCEVC enables a scalable two-layer encoding structure on the sender device, working alongside the device’s existing hardware encoder, whether AVC (H.264) or HEVC (H.265). It is deployable in software at the app level on both iOS and Android, so it can be rolled out through a standard app update with no dependency on device firmware or OS-level codec support.

On the sending side, the hardware encoder produces a universally decodable base stream as it does today. LCEVC then encodes the residual detail into a second, lightweight layer. The two layers can be packaged together for recipients with LCEVC-enabled apps, while the base stream remains a fully standalone stream for universal compatibility.
On the receiving side, devices with LCEVC-enabled apps decode both layers to reconstruct higher quality video. Devices without LCEVC support simply decode the base stream with no special fallback logic. If a recipient device cannot decode or display higher resolution, services can deliver only the base stream, reducing power consumption and improving reliability.

This provides a clear rollout path for messaging services. You can start with the client versions and device tiers you control and expand coverage over time. There is no flag day and no need to change capture or send workflows. Every recipient continues to get at least the same experience they get today, while those with updated apps get a better one.

LCEVC scalable mode to enhance a HEVC 1080p HDR channel to UHD HDR

Results and benefits

In video messaging, services typically deliver a single baseline encode, for example 360p or 480p at around 1.3 to 1.7 Mbps, to ensure consistent playback across recipients.

With MPEG-5 LCEVC, that base stream can be retained or even reduced in bitrate to accelerate Time-To-First-Frame (TTFF) and reduce transmit power consumption. An enhancement layer can then be transmitted to capable devices as quickly as network conditions permit to enable a higher-resolution experience such as 720p at under 3 Mbps. The base stream is decoded first, with enhancement applied dynamically when received and supported.

In testing on a representative 50-clip user-generated content set, using a fixed 360p base at 1.5 Mbps with LCEVC to reconstruct 720p:

The enhancement layer delivered a clear and visible quality improvement over the base stream.
In all cases services deliver the existing baseline service, avoiding any regression across their existing user base (i.e. there is only upside).
The scalable approach means that the base stream can be delivered first to minimize time to first frame (TTFF), with the LCEVC enhancement layer decoded automatically as it arrives, based on device capability and available bandwidth.

The primary benefit is improved quality of experience and delivery flexibility, without changing how video messages are captured, sent, or played back across the existing device base.

LCEVC quality boost examples – YT_Sports_43e2

Why it matters

In video messaging, compression decisions are made at the edge, on the sender’s phone, at the moment of capture. Those decisions directly shape the user experience, and they cannot be revisited for different recipients. This is fundamentally different from cloud-based video workflows where content can be re-encoded, multi-bitrate packaged, and adapted per viewer after upload.

LCEVC introduces flexibility into this otherwise fixed pipeline. It allows services to maintain a universal baseline for compatibility while adding scalable quality for recipients whose devices and networks can support it. And because it works alongside whatever base codec the device already supports, it does not require waiting for a single next-generation codec to reach universal availability. Deployment happens through app updates, on a timeline the service controls.