Codecs

AI-based video compression has been discussed in research for years, but practical implementations are virtually nonexistent. In a recent conversation with Arsalan Zafar, CTO and co-founder of Deep Render, and Sebastjan Cizel, Head of Engineering, we explored the real-world performance of Deep Render’s AI codec real-world encoding and decoding performance, how its quality compares to HEVC, AV1, and VP9, and …

When my daughter owned a Tesla during her college years, one thing I never thought to do was play YouTube videos on the LCD panel. It turns out that was a good call—if I had tried to stream AV1 content, playback performance likely would have been choppy and disappointing. Why bring this up? Tesla recently announced that it was rolling …

This paper, entitled, C3: High-performance and low-complexity neural compression from a single image or video, and authored by a team from Google DeepMind (Hyunjik Kim, Matthias Bauer, Lucas Theis, Jonathan Schwarz, and Emilien Dupont), introduces DeepMind C3, a neural compression method designed to deliver low decoding complexity while achieving competitive rate-distortion (RD) performance. The authors are recognized for their contributions …

The white paper, authored by researchers from the University of Science and Technology of China, introduces a novel approach to addressing the limitations of pre-trained learned image compression (LIC) models. These models, while effective within the domains they were trained on, often falter when applied to new, domain-specific data. To overcome this, the researchers propose lightweight adapters that allow efficient …

It started with a simple observation: Fox plans to upscale a 1080p HDR feed to 4K for the Super Bowl LIX broadcast. The comment came from Dan Rayburn, who shared details about Fox’s planned encoding ladder, including a maximum bitrate of 15 Mbps for 4K delivery. This revelation prompted a lively LinkedIn discussion among industry experts, exploring the decision’s technical, …

A new AI-based codec proved 18% more efficient than VVC but substantial decoding requirements will limit short-term commercial application. Here’s a summary of the white paper.  In December 2024, researchers from Fudan University introduced M3-CVC, an AI-based video compression framework that combines large multimodal models (LMMs) for semantic understanding and conditional diffusion models (CDMs) for high-fidelity reconstruction. The framework employs …

I’m running some 1-pass vs. 2-pass comparison testing with x265. Unlike with x264 (see tests here), there’s a substantial performance penalty for running 2-pass with minimal quality benefit. I tried the no-slow-firstpass=1 setting, which only accelerated two-pass by about 14%. Here are the results from a single test file. I don’t want to process them all if there’s a magic …

I was reminded of this recently as I prepared for a talk on AV1 readiness at the upcoming United Cloud Tech Talk. Though quality is only a single factor, I wanted to nail the quality comparisons for the talk, but didn’t have time to produce all the iterations myself. It’s been a while since I benchmarked H.264, HEVC, AV1, and …

Most people reading this post want new codecs to succeed as soon as possible, and I do as well. But whenever I think about codec adoption, six pictures come to mind. Overview The first was the foundational adoption timeline released by the Alliance for Open Media (AOMedia) with the launch of AV1 in April 2018 (see above). Essentially, this posited …

Announcing session 2 of Streaming Summer Bootcamp: Learn to Produce H.264 Video. Thu, Jul 11, 2024, 11:00 AM – 12:30 PM. The session is to be produced on LinkedIn Live: Click to register. The free webinar will cover: • What H.264 is and its origins • What x264 is and why it’s important • How x264 compares to other H.264 …