Exploring Workflows for Real-Time HDR-SDR Conversion

Due to the growing demand for high quality imaging in the movie industry as well as in broadcasting environments, research on high dynamic range (HDR) technologies is enhanced more and more. The incentive for the great interest in HDR content is particularly based on properties such as natural reproducibility of a scene containing the approximate dynamic range of real-world brightness distribution, a broader color spectrum and a more intense spaital depth in the image. Especially in the broadcast-related scope there is still potential to further develop and improve HDR technology. As not every end user has an HDR-compatible device as of now, methods are needed to map high dynamic range content to standard dynamic range (SDR) including transfer function and color gamut. Considering usually demanded real-time HDR-SDR conversion workflows, conventional mappers often produce an unaesthetic and unnatural look due to color hue shifts, flat brightness distribution or unnatural looking skin tones, which causes loss of the creative intent of images especially but not only in context of intense lighting effects. — On that account, the research team has set themselves the goal of analyzing existing mapping algorithms for a real-time broadcast use case in a show context. To properly test and apply the used algorithms on HDR video footage, a new dataset of HDR show content is produced. The insights of this comparison are depicted and explained.

Published

2023-10

Content type

Original Research

Keywords

HDR, High Dynamic Range, PQ, SDR, Dynamic Range, Brightness, Live Broadcast Production, Television, Broadcast Engineering, Film, Motion Picture, HDR-SDR Conversion

DOI

10.5594/M001993

ISBN

978-1-61482-964-5