High Dynamic Range (HDR) production is fast becoming the standard for major live sports broadcasts, with global events such as the Olympics, FIFA World Cup, Formula 1 races, and the Roland Garros tennis tournament now delivered in UHD HDR. This shift to HDR brings a remarkable improvement in visual quality, but the media and entertainment industry faces ongoing challenges in fully harnessing its potential.

In sports production, varying lighting conditions and sports disciplines demand unique calibrations for elements like skin tones, grass, or clay—requiring precisely set and maintained reference levels. Ensuring consistent reference levels and stable graphics across the production chain is crucial, especially in workflows that rely on a single HDR master and depend on regular conversions between HDR and SDR formats.

After content capture, network operators distribute either the HDR or SDR version based on the platform. Maintaining consistent conversions is essential. Static 3D Look-Up Tables (LUTs) are preferred for their simplicity and reliability, preserving stable graphics and natural midtones—like realistic grass on a soccer field. However, the need for different LUTs for each event, or even multiple LUTs within a single event, can result in inconsistencies.

Latest dynamic conversions techniques, built on content adaptation paradigms, introduce control points adjustments to guarantee that key reference levels, especially mid-tones, remain consistent throughout every conversion.  Looking ahead to potential future AI-based conversions tools will also have to address the challenges of ensuring reference levels consistency.

As for now, no signalization mechanism informs on how to drive the right conversion by selecting the most appropriate configuration at any point of the streaming chain and all these different conversion techniques are not interoperable without specific signaling.

This paper proposes an innovative metadata-driven method, formalized in the SMPTE Dynamic Range Conversion Characterization Metadata ST 2094-60 project, that characterizes all these necessary control points. The paper describes experimental results which demonstrate how to create, deliver and apply them in the production and delivery workflows. The proposed solution allows conversion consistency throughout the infrastructure and guarantees interoperability between any conversion technique.

By using metadata to automate conversions processes and improve reliability and quality, the solution will greatly enhance creative freedom for productions. 

In summary, the evolution of conversion techniques and metadata-driven workflows marks a substantial leap forward. By adopting these innovations, HDR producers can deliver stunning, reliable content that continues to meet and exceed global audience expectations.