Editorial: Video Content Production and Delivery Over IP Networks and Distributed Computing Facilities

Matteo Naccari, Fan Zhang, Saverio Blasi, Thomas GuionnetFrontiers in Signal Processing 2022 · DOI · OpenAlex

Abstract

This editorial highlights a pivotal shift in the media industry: moving from expensive, bespoke hardware-based video production to flexible, software-defined workflows powered by IP networks and the cloud. It showcases four key research contributions that introduce high-throughput codecs (HTJ2K), low-complexity enhancement layers (LCEVC), volumetric data standards (V3C), and new benchmarking datasets to make this transition possible.

Introduction: The Migration to IP-Based Production

For decades, professional video production relied on "bespoke" infrastructure—specialized hardware, dedicated SDI cables, and rigid physical studios. While reliable, these systems are incredibly expensive and difficult to scale. The industry is currently undergoing a massive migration toward IP-based production and distribution. By using standard Internet Protocol (IP) networks and cloud computing, broadcasters can replace hardware stacks with software running on general-purpose servers.

However, this shift creates a technical bottleneck: bandwidth and latency. Traditional compression standards (like the original JPEG 2000) were designed for hardware (FPGAs/ASICs) and do not perform efficiently on standard CPUs or GPUs. To make cloud-based production viable, the industry requires new "mezzanine" (intermediate) codecs that are software-friendly, ultra-low latency, and capable of handling the massive data rates required for 4K, 8K, and immersive 360° video.

Emerging Compression Technologies

This editorial synthesizes the approaches of several featured research works focusing on next-generation codecs:

High Throughput JPEG 2000 (HTJ2K)

HTJ2K is a software-friendly extension of the JPEG 2000 standard. It is designed to significantly speed up the encoding process while maintaining backward compatibility. Optimized for execution on modern CPUs and GPUs, it serves as a viable candidate for real-time IP-based production.

MPEG-5 Low Complexity Enhancement Video Coding (LCEVC)

LCEVC utilizes a "pyramidal" approach where a base layer is encoded at a lower resolution using any existing codec, and LCEVC adds enhancement layers to restore detail. This strategy allows for higher quality with lower computational overhead compared to single high-resolution streams.

Visual Volumetric Video-based Coding (V3C)

As the industry moves beyond 2D screens, V3C provides a family of standards designed specifically for the capture, storage, and transmission of volumetric data (3D/AR/VR). This ensures interoperability across different immersive ecosystems, preventing the creation of "walled gardens" of proprietary formats.

Benchmarking and Datasets

A significant resource highlighted for the research community is the BVI-CC Dataset. Introduced by Katsenou et al., this comprehensive dataset is designed for research on video compression and quality assessment. It includes a variety of video content along with objective and subjective quality scores, specifically tailored for modern streaming scenarios like adaptive bitrate (ABR) delivery. Public access to such datasets is essential for streamlining the design of novel compression schemes.

Key Insights and Results

The research featured in this editorial leads to several critical conclusions:

  • Software Optimization: Future mezzanine compression must be optimized for general-purpose computing. Codecs that cannot run efficiently on standard cloud instances (CPU/GPU) will not survive the transition to IP.
  • Performance Gains: Studies confirmed that H.266/VVC combined with LCEVC provides measurable improvements in both objective and subjective quality over the previous H.265/HEVC standard.
  • Real-World Adoption: LCEVC has already been appraised and adopted as one of the two primary compression technologies for the future generation of digital TV in Brazil (TV 3.0), demonstrating its readiness for large-scale deployment.

Limitations and Future Work

Despite these advancements, challenges remain. Ensuring seamless interoperability across all AR/VR ecosystems using V3C is a complex task requiring further protocol refinement. Additionally, while current papers provide resource estimates for CPU/GPU/FPGA, real-world performance in congested or "noisy" IP networks requires more extensive field testing. Finally, the community continues to need more diverse, publicly available datasets to keep pace with the rapid evolution of adaptive bitrate and immersive video technologies.

Summary updated

Abstract

EDITORIAL article Front. Signal Process., 15 July 2022Sec. Image Processing https://doi.org/10.3389/frsip.2022.975838
BibTeX
@inproceedings{naccari2022editorial,
  author = {Matteo Naccari and Fan Zhang and Saverio Blasi and Thomas Guionnet},
  title = {{Editorial: Video Content Production and Delivery Over IP Networks and Distributed Computing Facilities}},
  booktitle = {Frontiers in Signal Processing},
  year = {2022},
  doi = {10.3389/frsip.2022.975838},
}
Figure