“HEVC showed that compression efficiency is only one part of the deployment equation. Licensing structure and predictability also affect the business case.”
Jan Ozer’s recent paper, “Codecs for CFOs,” does the video industry a service. It moves codec strategy out of the engineering lab and places it where it belongs: inside the financial model of the business.
His central argument is uncomfortable but correct. A codec decision is no longer simply a comparison of bitrate, quality and encoding speed. It is a capital-allocation decision involving infrastructure, device reach, operational complexity, customer experience and increasingly material licensing exposure.
HEVC is the clearest demonstration of that reality. It is an excellent technology and an undeniable commercial success, particularly for 4K and HDR. Yet it also taught the industry that a technically mature standard can still make the complete economics hard to model with confidence. Patent pools, bilateral agreements, evolving distribution programs and enforcement actions have made it hard for many services to determine the complete cost before deploying.
The industry does not need another debate about whether patent holders deserve to be paid. They do. What technology buyers need is something much more practical: the ability to understand who must be paid, what triggers payment and how large the obligation can become.
The antidote to licensing uncertainty is not necessarily “free.” It is clarity.
Licensing is part of the product
For too long, licensing has been treated as a legal appendix to codec development. Engineers evaluate compression performance; product teams assess device coverage; and only later does somebody invite finance and legal into the room.
That sequence no longer works. The licensing architecture of a codec is part of the product itself. If its cost cannot be modeled before deployment, then its technical efficiency is harder to be accurately valued either.
This is where MPEG-5 LCEVC offers a fundamentally different proposition.
LCEVC is not a replacement codec. It is an ISO-standardized enhancement layer that works with an existing base codec, including H.264, HEVC, AV1 and VVC. The base video is typically encoded at a lower resolution, while lightweight residual layers restore detail at the target resolution. The architecture improves compression and processing efficiency without forcing an operator to discard the codec infrastructure it already owns. The technical model is described here.
That distinction matters commercially as much as technically.
A service can, for example, enhance its established H.264-based delivery with LCEVC while planning a migration to HEVC or AV1 as an entirely separate delivery format. This can materially improve the efficiency of an established H.264 workflow while retaining the compatible H.264 base.
Independent evaluations show the point across different use cases. In 1080p60 live gaming, LCEVC-enhanced x264 delivered approximately 42% VMAF bitrate savings over x264, while LCEVC-enhanced x265 achieved around 39% over x265. In a separate full-ladder live sports streaming evaluation, Jan Ozer found that an LCEVC-enhanced x264 ladder reduced weighted bandwidth by 22% while improving both VMAF and subjective quality scores. The precise results depend on content, encoder settings and workflow, but together they illustrate the same thing: LCEVC can provide another route to efficiency without requiring a wholesale replacement of the base codec.
Optionality has financial value
Because the LCEVC enhancement sits above the base codec, operators gain something that conventional codec migrations rarely provide: optionality.
Devices updated to support LCEVC receive the enhanced reconstruction. Devices that do not recognize it can still decode the underlying base representation. That enables staged deployment by device cohort, application or workflow instead of demanding a synchronized change across encoders, players, silicon and the installed device base.
This is not a claim that every implementation becomes effortless. Players must still be integrated and tested, and the commercial outcome must be measured against actual viewing patterns. Ozer is right to warn that laboratory efficiency does not automatically translate into equivalent CDN savings. Operators must evaluate their own content, ladders, device mix and audience distribution.
But a deployment that can build on an existing base codec is structurally different from one requiring a second, parallel codec estate. It gives engineering teams more control over when and where adoption occurs and gives finance teams a way to invest incrementally.
A licensing model you can plan around
LCEVC is a patented technology. As a product person, predictability of the licensing obligation is as important as the fee structure. Because the LCEVC programs are published, services and device implementers have visibility into the royalty structure, thresholds and caps.
At product level, the model is relatively simple. For consumer devices, licensing is based on the number of LCEVC-enabled devices shipped, with published volume tiers and annual caps. For video distribution services, the royalty scales with the size of the service, using published measures such as users, subscribers or revenue; smaller services fall within a waiver tier, and royalties for free-to-air broadcast distribution are waived.
This does not remove the licensing considerations associated with the rest of the video stack. If, for example, LCEVC is used with HEVC, the HEVC licensing question remains. The advantage is that the LCEVC part can be understood separately: there is a published route to a license from a single entity across consumer-device and video-distribution use cases.
For me, that is the useful distinction. I do not need licensing complexity to disappear; I need enough visibility to model it before committing to a product architecture.
What LCEVC does, and does not, solve
If an operator uses HEVC as the base codec beneath LCEVC, it must still address its HEVC licensing obligations. LCEVC does not make those disappear.
What LCEVC provides is choice.
An operator can evaluate LCEVC over H.264 as an alternative route to greater efficiency. It can enhance an existing HEVC deployment and extend the return on that investment. It can also use LCEVC to reduce the complexity of adopting AV1 or VVC. In each case, the LCEVC cost can be incorporated into the model in advance rather than discovered after the architecture has been committed.
That is the real antidote to codec uncertainty: not the promise that intellectual property has no price, but the confidence that the price and its triggers can be understood.
The next generation of codec decisions will not be won by the most impressive headline compression ratio. They will be won by technologies whose complete economics, compression, compute, deployment and licensing can be written down before the first production stream is delivered.
For CFOs, that is not a small distinction. It is the difference between taking a calculated investment decision and signing a blank cheque.
