Preliminary VMAF and PSNR Scores for New Xilinx Transcoding Appliances

Live streaming at scale is the perfect application for hardware transcoders that deliver highly-affordable encoding density. Xilinx just announced a series of transcoding appliances targeting these live streaming use cases built around the new Xilinx Real-Time (RT) Server reference architecture. I’ve been benchmarking output quality with two card-based encoders deployed in these appliances, the Alveo…

How to Install Python 3.8 and Bento 4.

When I first started using Bento4 it ran only on Python 2.7x and earlier. For this reason, the installation instructions in my book Learn to Produce Videos with FFmpeg in 30 Minutes or Less describe how to install Python version 2.7. At some point, however, Bento upgraded to Python 3.x, obsoleting those instructions. I’ve upgraded…

AV1 Turns 2.0

Just a quick announcement that the Alliance for Open Media has launched version 2.0 of the AOMedia AV1 encoder that you can download here. According to the Phoronix website, from which I grabbed this news: “Libaom 2.0 is the first release since the original 1.0 release back in mid-2018 after the AOMedia codec working group…

New Five-Star Review for FFmpeg Book

One thing that never gets old is a five-star review on Amazon. I just received another one for my book, Learn to Produce Video with FFmpeg: In Thirty Minutes or Less (2018 Edition), which is available on Amazon ($34.95) and for PDF download ($29.95). In this review from a verified purchaser, the customer said: “I…

Encoding VP9 in FFmpeg: An Update

This is a long post only of interest to those attempting to optimize their VP9 encodes. The three key takeaways are 1) use the command script shown on the bottom of the page, 2) a speed setting of 2 offers the optimal quality/performance tradeoff, and 3) the row-mt setting improves performance significantly with zero quality…

VMAF is Hackable: What Now?

Just a quick note to let you know about some recent findings relating to the Netflix VMAF metric. By way of background, VMAF launched in June 2016 as the metric that powered Netflix’s per-title encoding engine, replacing PSNR. The fact that Netflix created and used VMAF gave the metric tremendous credibility and because Netflix open-sourced…