Quality Advancements and Automation Challenges in File-Based Conversion: Noise-Reduction, Deinterlacing, High Frame Rates, and Compression Efficiency

With continued expansion of IP infrastructures and better compression schemes, the age of explosive growth in video online content has arrived. Bigger (UHD/4K/8K), faster (HFR) and better display technologies offer the potential for a truly immersive viewing experience. However, the shortcomings inherited in archive material, such as noise, and other artifacts, are increasingly visible. All content needs to compete for bandwidth and viewership, so high quality pre-processing prior to compression is becoming more important. Pre-processing can help to achieve the best possible picture quality while saving significant compressed bit-rates, in particular for remastering. This paper shows detailed compression efficiency results relating to deinterlacing, noise reduction, and simulated motion-blur. Additionally, an overview on an innovative clustered server system is given, with emphasis on advancements in flexibility, automation, efficient scalability, and latency. Various clips showing processing quality are available on request.

Published

2014-10

Content type

Original Research

Keywords

HFR, Noise reduction, Motion-compensated, Deinterlacing, Frame-rate conversion, Remastering, File-based, Standards conversion, Up-scaling, GPU, IP-based, UHD, 4K, SD, HD, AVC/HEVC Compression, Online distribution, Workflow, Low latency, Scalability, Flexibility, Automation

DOI

10.5594/M001566

ISBN

978-1-61482-954-6