Compression Efficiency Beyond the Codec
V-Nova develops compression and visual data technologies used across broadcast, streaming, and AI workflows. Its best-known technology, MPEG-5 LCEVC (ISO/IEC 23094-2), is a standardized enhancement layer that sits on top of any base codec. Rather than competing with H.264, HEVC, VVC or AV1, LCEVC encodes a lower-resolution base stream with the existing codec and adds a lightweight enhancement sub-layer that restores full-resolution detail on playback.
The practical consequence is that quality improves and bitrate drops without forcing a codec migration or a device replacement cycle. In MPEG verification testing, LCEVC delivered approximately 46% bitrate savings when enhancing H.264/AVC at UHD and 31% when enhancing HEVC at UHD. LCEVC is a core component of Brazil’s DTV+ (TV 3.0) standard, now moving into live commercial broadcast, and has been used in live streaming trials by Globo around the FIFA World Cup and Carnival broadcasts.
V-Nova also develops SMPTE VC-6, a hierarchical low-complexity format aimed at production, edge and AI workloads, and PresenZ, a 6DoF immersive format. The company holds more than 1,500 international patents and contributes to multiple international standards.
NETINT and V-Nova first collaborated in 2020, integrating V-Nova’s LCEVC software library with NETINT’s Codensity T408 video transcoder, with a stated intent to explore LCEVC in future NETINT silicon.
Role in the VPU Ecosystem
Most partners in the VPU ecosystem operate at a layer adjacent to processing: transport, orchestration, infrastructure. V-Nova‘s work sits closer to the codec itself.
LCEVC splits an encode into two parts. The base layer is a standard, fully compliant stream in a conventional codec, encoded at reduced resolution. The enhancement layer is a compact, low-complexity residual that reconstructs full-resolution detail at the decoder. Because the base layer is ordinary H.264, HEVC or AV1, it is the same category of work that dedicated video silicon is designed to do efficiently: fixed-function, high-density, low-power encoding.
That structural fit is why the two approaches are discussed together. VPUs address the compute cost of encoding. Enhancement-layer compression addresses the delivery cost of bandwidth. These are usually traded against each other, and operators facing both pressures tend to end up evaluating both categories of answer.
NETINT and V-Nova have a collaboration history on hardware-accelerated LCEVC dating to 2020, and LCEVC remains a technology NETINT tracks as the standard moves into national broadcast deployment. Operators evaluating LCEVC as part of a longer-term distribution strategy are welcome to start that conversation with either company
Company Type
Video compression and visual data technology company; standards developer and IP licensor
Core Focus
Enhancement-layer compression (MPEG-5 LCEVC), hierarchical visual data formats (SMPTE VC-6), codec-agnostic efficiency
Primary Workloads
Live streaming, linear and OTT distribution, VOD libraries, next-generation terrestrial broadcast (DTV+ / TV 3.0), AI and computer vision pipelines
Customer Profile
Broadcasters, streaming platforms and OTT services, network operators, encoder and SoC vendors, device manufacturers
Who to Talk to About VPUs at V-Nova
Connect with the V-Nova team to discuss MPEG-5 LCEVC, DTV+ and next-generation broadcast, and where enhancement-layer compression fits an existing distribution chain.
Where the Technologies Align
Dedicated silicon reduces the cost of producing video. Enhancement-layer compression reduces the cost of delivering it.
- LCEVC is codec-agnostic by design. It layers over standard H.264, HEVC and AV1 output, so it does not require a change to the base codec, the player estate, or the distribution chain.
- The base layer is conventional encoding. Encoded at reduced resolution, it is the same class of workload that dedicated video processing hardware handles efficiently, which is why the two approaches are structurally compatible.
- The efficiency gains address different costs. Published MPEG verification testing reports approximately 46% bitrate savings over H.264/AVC and 31% over HEVC at UHD. Those are delivery-side savings, distinct from the compute-side savings of moving transcoding off the CPU.
- No device replacement cycle. LCEVC decodes in software on existing devices, in browsers via JavaScript and WebAssembly, and in silicon where SoC support is present.
- Standards momentum is real. LCEVC is a core component of Brazil’s DTV+ (TV 3.0), enabling UHD HDR delivery within existing spectrum, and the same efficiency logic applies to bandwidth-constrained OTT distribution.
Every few years the industry settles on a new codec, and every few years the same sequence plays out. Encoders arrive first. Players follow. The installed base of devices takes the better part of a decade. By the time a codec transition is complete enough to build on, the bandwidth it was meant to save has already been absorbed by higher resolutions and more concurrent streams.
LCEVC does not ask anyone to wait for that. It encodes a base stream in a codec that is already deployed everywhere, then carries a compact enhancement layer that restores full-resolution detail at playback. The efficiency applies to the codec being run today rather than the one a device base might support several years from now. MPEG verification testing puts the gain at roughly 46% over H.264 and 31% over HEVC at UHD, and Brazil specified it in DTV+, its next-generation national broadcast standard.
That reasoning will look familiar here. Video keeps getting more expensive to produce and to deliver, and the standard responses are to add more general-purpose compute or to wait out a codec cycle.
Purpose-built silicon and enhancement-layer compression are two expressions of the same instinct: change the architecture instead of scaling the brute force.
News, Articles and Videos
Featuring V-Nova
When Standards Become Strategy
Every codec transition promises savings and then makes you wait a decade for the device base. This looks at why an enhancement layer changes that calculation, what the verification numbers actually show, and what Brazil’s national rollout means for everyone else.
V-Nova and NETINT Partner to Boost Transcoding Quality and Speed
The announcement that started the relationship. V-Nova’s LCEVC library met NETINT hardware transcoding, with both companies setting out where enhancement-layer compression could go next on dedicated silicon. Worth reading as the origin document, and for what it anticipated.
V-Nova at IBC 2026
Brazil’s DTV+ has moved from validation into live broadcast, and V-Nova’s IBC showcase is the clearest look at what that took. Covers the LCEVC device ecosystem, SMPTE VC-6 for AI and visual data workflows, and what ships next.
The Next Layer of Video Efficiency
The published evidence on LCEVC, including the studies that argue against it. Five years of subjective testing from MPEG to Brazil’s TV 3.0 rollout, the adversarial PSNR results, and the precise operating points where the savings stop holding.
Coming Soon (08/28/2026)
Learn more about V-Nova and its MPEG-5 LCEVC and SMPTE VC-6 compression technologies.
