Follow developments in AV1, AV2, H.264/AVC, HEVC/H.265, VVC/H.266, FFmpeg, cloud transcoding, hardware acceleration, bitrate optimization, and video quality measurement. StreamingMeme curates and summarizes the updates that matter to streaming engineers, product teams, and media technology leaders.
GNU and the Free Software Foundation (FSF), with Richard Stallman's endorsement, are consolidating around AV1 codec adoption due to its wider support among their community members. This decision prioritizes compatibility over the newer AV2 codec for the foreseeable future. The decision is part of an ongoing series discussing codecs and software patents, highlighting AV1 as the chosen standard for their video conversions.
A guide provides detailed instructions on using SVT-AV1 for AV1 encoding, covering its applications for VOD, storage, and live streaming. The guide highlights the importance of preset selection to balance encoding efficiency with computational demands for various use cases.
ByteDance's Multimedia Lab is seeking an SoC Digital Verification Engineer to develop advanced video codec hardware solutions, including FPGA and ASIC designs. This role focuses on designing hardware accelerators to support scalable video platforms and services for billions of users. The successful candidate will work on verification planning, environment development, test execution, and system-level verification for video encoding and processing hardware.
MediaKind provides detailed documentation on configuring individual services within its MK.IO Beam feed, specifically including the Live Encoder, Multiplexer, and Satellite Modulator. This guide offers engineers specific instructions for setting up contribution feeds using MediaKind's technology.
This article explains Variable Bit Rate (VBR) encoding, detailing how it adjusts bit allocation based on content complexity to improve quality for VOD and stored media. It contrasts VBR with Constant Bit Rate (CBR) and clarifies why VBR is typically used for on-demand content and pre-recorded streaming, while CBR is preferred for real-time applications like WebRTC due to VBR's unpredictable short-term bitrate spikes.
QSVEnc, a free software utilizing Intel QuickSync Video encoders, has released version 8.15. This update reduces GPU memory usage, optimizes processing, and introduces new video filters such as FSR1 upscaling and hqdn3d noise reduction. The software supports H.264, H.265, VP9, and AV1 encoding, catering to professionals focused on hardware-accelerated video processing.
This article provides an overview of VP8, Google's open-sourced, royalty-free video codec, highlighting its historical significance as the mandatory-to-implement codec for WebRTC. It details VP8's role in ensuring universal browser compatibility and its technical workings, comparing it to other modern codecs like H.264, VP9, and AV1. The piece emphasizes VP8's continued use as a fallback codec despite newer, more efficient alternatives.
GNOME 50.2 has been released, featuring an optimization that significantly reduces H.264 screencast file sizes by incorporating bit rate control in VA-API-based encoding. This update addresses previous issues where hardware-accelerated recordings were excessively large due to a lack of proper compression rate management. The improvement benefits content creators and streamers by enabling smaller video files for easier upload, download, and storage.
The Alliance for Open Media (AOMedia) has released the final specifications for its AV2 video codec, promising 30% more efficient data compression than AV1, with even greater benefits for 4K HDR video. This technical development aims to significantly reduce streaming costs for services. However, a recent lawsuit by Dolby against Snap Inc. regarding AV1 patent infringement raises concerns about the royalty-free status of both AV1 and its successor, AV2.
This article explains the technical differences between two Dolby Atmos formats: lossless TrueHD for physical media and lossy Digital Plus for streaming platforms. It highlights that current streaming bandwidth limitations and virtual soundbar technology prevent parity with high-fidelity physical media audio. These technical differences impact the actual immersive audio experience for consumers.
An AI agent has identified 21 zero-day vulnerabilities within FFmpeg, a widely used video processing component. Separately, the Chrome browser patched a record 429 bugs, highlighting ongoing security challenges in critical software used in streaming and web technologies.
Amazon Web Services (AWS) published guides detailing its video-focused services, AWS Elemental Inference and AWS Elemental MediaConnect. The Inference guide explains its AI features and integration into encoding workflows. The MediaConnect guide clarifies the functionality of its Flows, Router, and Gateway components for streaming professionals.
Microsoft Store has launched VideoShrink, a desktop application designed for local video compression. Utilizing HandBrake CLI and ffprobe, it optimizes file sizes while maintaining quality for various formats like MP4, MOV, and MKV, offering features like drag-and-drop and batch processing.
StreamEncoder PRO has launched a professional audio streaming encoder designed for broadcasters and radio stations. It supports simultaneous streaming via SHOUTcast, Icecast, HLS, RTMP, and RTSP from a single WASAPI capture, offering multi-instance profiles and unlimited outputs per instance. The software is built for broadcast workflows, including integrated audio processing and multi-channel DANTE support for running multiple stations on one machine.
Google's internal issue tracker shows a P3/S2 internal cleanup fix by wt@google.com for C++ compilation flags affecting Chromium and Aomedia codebases. The resolution includes multiple code changes related to fundamental streaming technologies.
The article asks for a comparison between H.264 and H.265 video codecs, exploring their respective strengths and weaknesses for various streaming applications. It implicitly discusses the technical considerations involved in choosing a video compression standard.
The `viser` crate offers a unified Rust library that simplifies video encoding optimization by providing a single facade for a dozen underlying crates, including FFmpeg and VMAF. This technical development helps developers building video processing applications manage dependencies more easily. It offers features like VMAF/PSNR/SSIM measurement, per-title encoding, shot/scene detection, and context-aware ladder generation.
Peacock will stream Telemundo's FIFA World Cup 2026 coverage in Dolby Vision and Dolby Atmos with Dolby AC-4, through a joint engineering effort between Dolby Laboratories and NBCUniversal. This initiative marks the first commercial deployment of Dolby AC-4 by a video streamer, highlighting advanced audio and video integration for live sports using native HDR production. The deployment aims to deliver a high-quality viewing and listening experience for all 104 matches.
The GNOME Project has released GNOME 50.2, a maintenance update to its desktop environment series. This update notably adds rate control parameters to the VA-API H.264 screencast pipelines, preventing the encoder from using its default bitrate.
The GNOME Project released GNOME 50.2, a maintenance update to its desktop environment. This update includes the implementation of rate control parameters for VA-API H.264 screencast pipelines, preventing the encoder from using its default bitrate. It also features various other fixes and improvements across core GNOME components and applications.
The Alliance for Open Media has officially released the definitive specifications for AV2, the successor to the AV1 video codec. AV2 promises to deliver 25-32% smaller file sizes compared to AV1, which could significantly reduce bandwidth consumption for video streaming platforms. However, its increased decoding complexity (roughly five times that of AV1) presents a challenge for widespread adoption without hardware acceleration or optimized software decoders.
The Alliance for Open Media (AOMedia) has officially released the AV2 video codec, promising a 30% efficiency gain over its predecessor, AV1. This new royalty-free codec aims to allow streaming services to either reduce bandwidth consumption or deliver higher picture quality. Major tech companies, including Google, Amazon, Apple, Netflix, and Microsoft, are members of AOMedia and are involved in the development and potential implementation of AV2.
ArXiv's recent submissions in image and video processing include a notable paper titled "KD-NVC: Accelerating Neural Video Coding Via Search-and-Distill Framework," which discusses advancements in neural video compression technologies. This research focuses on improving the efficiency of video encoding using AI-driven methods. Other submissions on the platform cover a range of AI applications in medical imaging and computer vision, though they are less directly relevant to streaming video technology.
NETINT Technologies highlights that the deployable performance of Video Processing Units (VPUs) in video infrastructure is critically dependent on system architecture, not just silicon efficiency. The article, authored by Mark Donnigan from NETINT, emphasizes that factors like PCIe topology, power delivery, thermal management, and mechanical design are crucial for sustaining real-world video workloads. It partners with Advantech for purpose-built platforms to ensure NETINT VPUs operate at full capacity under continuous 24/7 workloads.
AOMedia has released the AV2 1.0 open-standard video codec, targeting 30% better compression than AV1, with a reference encoder available for Linux users. Widespread playback support and hardware acceleration for AV2 are expected to take several years to materialize across different platforms and devices. The article details how to build and test the AV2 reference encoder/decoder on Linux and outlines the current lack of wider tool and hardware support.
AMD's rocDecode documentation details the codecs and hardware decode capabilities for various AMD GPU architectures, including H.265, H.264, AV1, and VP9. This information is crucial for developers and integrators building video processing solutions on AMD platforms, offering insight into maximum resolutions and supported bit depths per codec and GPU series.
Crocott has published an in-depth technical explanation of Adaptive Bitrate (ABR) transcoding for IPTV and OTT operators, detailing how it works, the importance of building effective bitrate ladders, and the differences between HLS and MPEG-DASH. The article highlights FastoCloud's GPU-accelerated ABR transcoding solution for HLS/DASH packaging and DVR, which integrates with CrocOTT middleware for live and VOD services, offering an affordable full ABR stack.
Delta Digital Video has released a whitepaper titled "JPEG-XS for Tactical ISR" that compares the ultra-low-latency codec to H.264 and H.265 for tactical ISR workflows. The whitepaper helps streaming professionals understand the optimal application of codecs in systems where low latency, deterministic behavior, and image quality are critical. It also highlights how DDV's products like the 5480E, VEC500, and 9600/9610 support JPEG-XS.
AMD's EPYC 8635P processor exhibits significant performance enhancements for video encoding and transcoding tasks, specifically for AV1 and JPEG-XS formats. Benchmarks show a notable upgrade over its predecessor, the EPYC 8534P, indicating improved efficiency for data center video processing operations. While also showcasing PostgreSQL performance, the core improvement relevant to streaming is in video encoding.
AWS Elemental Live has introduced enhanced graphic overlay features, supporting both static and motion graphics. Users can insert BMP, PNG, or TGA files for static overlays and HTML, MOV, or PNG sequences for motion graphics. The system allows control via REST API or SCTE-35 signals for dynamic graphic insertion during live events.
Bluenviron is considering adding per-title encoding capabilities to its MediaMTX media server, a feature that allows for dynamic bitrate management based on content complexity. This enhancement aims to optimize delivery efficiency for streaming content.
Hikvision has introduced Guanlan Encoding, an AI-powered compression technology that integrates a Large Vision Model into the H.265 pipeline to reduce video storage requirements by 30% to 50%. The technology uses precision Region of Interest segmentation to preserve detail on crucial objects like people and vehicles while compressing static backgrounds.
The article describes a new multi-channel audio and video transcoder capable of processing 4K, HD, and SD video streams and converting various formats (MPEG2, VP8) to H.264 or H.265. This hardware supports protocols like HLS, UDP, and RTMP, facilitating content delivery across satellite, internet, and local IP networks while reducing bit rates. The device can process between 5-6 4K programs or up to 70-76 HD/SD programs, with an advanced version handling over 100.
A discussion on Hacker News explores the technical viability of using H.264 as a high-bitrate intermediate format for remote transcoding systems. Users debate whether sending H.264 compressed files to specialized encoding hardware for further optimization into 'better' codecs is a sensible strategy. The core of the conversation revolves around efficient video compression workflows involving different codecs and hardware resources.
Discussions on MacRumors forums question the lack of native hardware support for the AV1 codec on Apple M2 chips. The conversation explores the feasibility of implementing AV1 support via software on iPad and Mac devices, considering the M2's 8-core configuration.
The Alliance for Open Media has released the final AV2 specification, an open and royalty-free video codec designed to achieve a 30% bitrate reduction compared to AV1. This new codec introduces advanced coding tools like data-driven transforms and larger coding units to improve efficiency, potentially allowing streaming providers to deliver 4K content with significantly less bandwidth. However, widespread adoption will depend on encoder implementations and hardware support, which historically takes several years.
Engineers are debating the necessity of hand-written assembly and SIMD in video decoders like dav1d to achieve high performance for codecs such as AV1 and the upcoming AV2. The core of the discussion is whether modern compilers are capable of generating sufficiently efficient code without low-level manual optimization for next-generation codecs. The article presents arguments for and against the "impossibility" of compilers matching hand-optimized assembly code for frame decoding.
This article provides a comprehensive guide to IPTV encoders, detailing their function in converting raw video into digital formats like H.264 and HEVC for internet-based television delivery. It covers how encoders work, their importance for bandwidth optimization and multi-device compatibility, and key features such as multi-channel encoding and low latency streaming. The guide also discusses different encoder types (hardware vs. software), setup procedures, best practices, and future trends like AI-powered encoding and AV1 adoption.
An engineer reported encountering corrupted JPEG output when attempting hardware-accelerated JPEG encoding on the MediaTek 520 MVPU using GStreamer v4l2 plugins. The corruption occurs for both file-based and live camera sources, though a test source pipeline works correctly. The issue indicates a potential bug or configuration problem within the platform's video processing unit or GStreamer implementation.
Trud.ca has released a narrated tutorial from Mile High Video 2026, offering an in-depth look at the design intent and compute allocation of the AV1 video codec. The tutorial uses empirical data to explain how specific coding tools influence efficiency and benchmarking across hardware and software encoders. It aims to build intuition around the codec's design rather than just detailing its specifications.
intoPIX and dSPACE have integrated intoPIX's TicoRAW, an implementation of the JPEG XS RAW ISO standard, into dSPACE's high-performance data logging platforms for autonomous driving development. This integration enables lossless compression of multi-sensor data streams exceeding 50 Gbit/s, significantly reducing storage costs and transfer times for AI training while preserving data fidelity. The solution addresses critical challenges in efficiently capturing and managing massive volumes of sensor data generated by modern test vehicles for ADAS and AD systems.
University of Texas researchers have developed SEAOTTER, a compression framework for cloud robotics that combines a sensor-embedded autoencoder with a learned JPEG transcode. SEAOTTER achieves significantly faster encoding and decoding than AVIF at high compression ratios, while also improving downstream computer vision accuracy and maintaining compatibility with existing JPEG infrastructure. The system leverages AI to optimize color transforms and quantization matrices for specific applications, producing standard JPEG files that can be efficiently decoded.
Researchers have proposed a Bitrate-Adaptive Distortion-to-Perception (BA-D2P) image compression framework that dynamically balances visual quality and reconstruction precision. This model uses hierarchical feature decoupling and multi-layer codebooks to optimize for perceptual quality at low bitrates and detail fidelity at high bitrates. It aims to integrate the strengths of both distortion-based and perception-based compression methods.
Researchers from JD.com and Shanghai Jiao Tong University developed AdaCodec, a predictive visual code for video Multimodal Large Language Models (MLLMs). This new approach encodes inter-frame changes as compact P-tokens, reducing visual token usage by 84.7% and significantly lowering inference latency. AdaCodec demonstrates improved performance across various long-video and general video-understanding benchmarks compared to traditional per-frame RGB encoding.
EaseUS has released VoiceWave, a real-time AI voice changer for live streamers, specifically integrated with platforms like Kick, OBS Studio, and Streamlabs. The software offers over 400 AI voice effects, ultra-low latency processing, and soundboard capabilities, designed to enhance interactive audio during live broadcasts. It supports integration with over 2000 platforms including games and chat apps.
Redis released version 8.8 with significant performance improvements across various data types and operations. Key enhancements include a 68% boost in MGET throughput, 83% improvement in stream processing (XREADGROUP), and up to 60% faster persistence and replication. These optimizations aim to enhance back-end performance for real-time applications and data distribution.
An OBS Studio user discovered that copying the `obs-qsv11.dll` plugin from version 27.2.4 into modern OBS installations enables hardware H.264 encoding on Intel Sandy Bridge processors. This workaround addresses the incompatibility of newer Quick Sync encoder APIs with legacy Intel graphics drivers, significantly reducing CPU load on older systems. The user has inquired whether OBS plans to provide an official fallback for older Quick Sync architectures.
Nerdleveltech is recruiting a Full Time Video Streaming Engineer in Los Angeles to develop low-latency infrastructure. The role focuses on optimizing video pipelines and integrating Chromium Embedded Framework for real-time media delivery, supporting codecs like AV1 and VVC. This position requires expertise in GStreamer, FFmpeg, Rust, and C++.
Adobe Premiere Pro users are requesting the addition of a Constant Rate Factor (CRF) option for H.264 and H.265 video exports. Users currently rely on third-party tools like FFmpeg with custom scripts to achieve consistent visual quality, circumventing Adobe's existing bitrate-based export presets which can result in inconsistent quality.
Developers for Radarr and Sonarr are implementing initial support for the H.266/VVC video codec following compatibility with FFprobe 8.1. This update will enable these media management tools to correctly identify and parse VVC metadata, addressing a previous issue where movies encoded in h266 caused display errors. A Pull Request for VVC/H266 support has been created for Sonarr, which is upstream for Radarr.
