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.
AWS Deadline Cloud now supports persistent Amazon EBS volumes for Service-Managed Fleets, replacing previous ephemeral-only storage. This enhancement allows workers to retain Conda environments, shader caches, and asset collections across worker lifecycle events, which reduces startup times for visual effects and animation workloads. The persistent storage is available in all AWS regions where Deadline Cloud is offered and is priced the same as existing Service-Managed Fleets EBS pricing.
Benchmarking results indicate that the Linux 6.18 LTS kernel performs worse in SVT-AV1 and Kvazaar H.265 video encoding tasks compared to more recent kernel versions. This performance disparity is attributed to improvements in the AMD P-State driver found in newer Linux releases when tested on Threadripper workstations.
Artificial intelligence is significantly transforming voice recording by automating audio cleanup, streamlining script preparation, and enabling voice cloning. These advancements empower streaming content creators to scale localization through automated translation and reduce post-production time for audio.
Amazon Web Services (AWS) has announced support for Kubernetes version 1.36 on Amazon EKS and EKS Distro. This update introduces features such as pod-level vertical scaling without restarts and resource health reporting to identify hardware-caused crash loops, which can benefit streaming and cloud workflows. The new version is available in all AWS regions where EKS is present, including AWS GovCloud, for new clusters and existing cluster upgrades.
NVIDIA and Adobe announced the RTX Spark partnership to accelerate Premiere Pro workflows. The collaboration will leverage NVIDIA's AI-driven GPU hardware to offload rendering, encoding, and AI tasks, aiming to reduce bottlenecks in high-resolution 4K and 8K post-production workflows. This integration is expected to result in faster exports, smoother editing, and improved AI features like masking and auto-reframing for Premiere Pro users.
NVIDIA's App beta 11.0.8 update expands ShadowPlay recording capabilities, enabling 4K at 240 FPS for dual NVENC GPUs (RTX 50/40 series) and 1440p at 240 FPS for single NVENC GPUs. This update unifies NVIDIA's software ecosystem, integrating features traditionally found in the Control Panel and GeForce Experience, with bug fixes and stability improvements.
Brown University researchers have developed PackUV, a new processing method that maps 3D Gaussian Splatting data onto 2D image frames. This innovation allows high-resolution volumetric video to be compressed and streamed using existing codecs and internet infrastructure. The breakthrough enables interactive 3D viewing on standard consumer devices like smartphones and smart TVs, solving previous challenges of storing and streaming terabytes of 3D data.
Mux has launched Robots Directives, a new feature designed to automate video platform workflows for Mux Video assets through declarative configuration. This feature manages asset dependencies and task execution, eliminating the need for additional customer infrastructure. The directives ensure desired workflows are executed once assigned to an asset.
Samsung has introduced a new APV (Advanced Professional Video) codec designed to improve video recording performance on smartphones. This technology aims to deliver higher bitrates and reduce compression artifacts for high-end mobile video production. The article notes this could be a significant upgrade for smartphone video recording.
Google Ads documented its recommended video ad specifications, advising the upload of .MPG (MPEG-2 or MPEG-4) format videos. The guidelines also list other supported video formats including .WMV, .AVI, .MOV, .FLV, and .MPEG-1.
This article, from the Japanese site Hatena, contains machine code which mentions parsing and forwarding metadata, and Vulkan compute-based operations related to LCEVC support. The context suggests a technical update or change log for software.
Delta Digital Video (DDV) has published a whitepaper detailing the application of JPEG-XS codec technology for tactical Intelligence, Surveillance, and Reconnaissance (ISR) platforms, highlighting its benefits for low-latency, deterministic video delivery over H.264/H.265. The paper positions JPEG-XS as crucial for mission threads requiring forensic fidelity and machine vision performance, while DDV's Model 5480E encoder and VEC500 OEM module support both JPEG-XS and legacy codecs to bridge current and future military IP video standards. It also specifies how codec selection should be an architectural decision based on mission requirements rather than a procurement decision per platform.
The Alliance for Open Media (AOMedia) website serves as a central hub for information regarding open-source digital media technology, primarily highlighting the AV1 video codec. It features recent news about advancements in AV1-related software, such as dav1d, Libavif v1.4.0, SVT-AV1 4.0, and the release of reference software for Polygonal Mesh Coding Standard. The site also announces new members joining the alliance, including Immersive Machines, Green Field Sound, and Qencode.
Eduardo García-Mádico Portabella of EGP Sistemas has developed a new VA-API driver, "rockchip-vaapi," for the Rockchip RK3588 and RK3576 SoCs. This driver bridges libva to librockchip-mpp, enabling hardware-accelerated video decoding of H.264, HEVC, VP9, and AV1 codecs in applications like Firefox, at resolutions up to 8K, with zero-copy DRM PRIME 2 surface export. The project was developed with assistance from an AI model, Claude Sonnet 4.6.
Nokia published a blog post outlining how Versatile Video Coding (VVC) can facilitate multiview live streaming. The article details VVC's use in a single-player multiview approach, allowing personalized viewing experiences by clients merging multiple camera feeds. VVC provides advantages over HEVC for this application, such as independent subpictures and simplified bitstream merging.
AetherEngine is an open-source video player engine for Apple platforms (iOS, tvOS, macOS) that integrates FFmpeg for demuxing and AVPlayer/VideoToolbox for native playback. It supports a wide range of codecs including HEVC, H.264, AV1, and VP9, along with advanced features like Dolby Atmos passthrough, HDR10, HDR10+, Dolby Vision, and HLG with HDR-to-SDR tonemapping. The engine handles complex playback requirements by dynamically switching between a native AVPlayer HLS-fMP4 pipeline and a software-based pipeline using `AVSampleBufferDisplayLayer` for codecs not natively supported by AVPlayer.
The article discusses the growing adoption of the AV1 video codec, replacing H.264 as the industry standard due to its superior compression efficiency for 4K and 8K video. It highlights AV1's open-source and royalty-free nature, which facilitates widespread deployment by major technology companies and enables enhanced capabilities for various video applications, including streaming and surveillance, in 2026 and beyond.
YouTube has published its recommended upload encoding settings for videos, detailing specifications for container format (MP4), audio codecs (AAC-LC, Opus, Eclipsa Audio), video codec (H.264), frame rates, and bitrates for Standard Dynamic Range (SDR) and High Dynamic Range (HDR) content. The guidelines also specify color space recommendations for SDR uploads, primarily BT.709, and outline how YouTube standardizes or converts color space values during processing.
@mediabunny/server is a new package that polyfills browser-environment media processing features for server-side environments like Node.js, Bun, or Deno, enabling full Mediabunny functionality on the server. The package facilitates server-side video and audio processing, including encoding/decoding for codecs like H.264, HEVC, VP9, and AV1, as well as video frame transformations, using NodeAV as a thin wrapper around FFmpeg's C API. It aims to integrate media processing natively with TypeScript, reducing the need for external CLI calls and offering features like automatic hardware acceleration, multithreading, and zero-copy decode/encode paths.
The article discusses 10 alternatives to OBS Studio for recording and streaming. It outlines various tools that can be used for recording, streaming, podcasting, webinars, and live production.
This article is a promotional page for a market research report titled "High Efficiency Video Coding Hevc Market Size | Industry Growth Report 2035" from Market Research Intellect. It specifies the report's focus on analysis, industry outlook, growth drivers, and forecast for HEVC across various end-users, deployment models, technologies, applications like video streaming and broadcasting, and device types. The report, published as its 6th edition in 2026, covers the period from 2025-2035 and is over 150 pages long.
This guide details the process of setting up and customizing quality profiles within Radarr, a media management tool, focusing on the use of "Custom Formats" to filter and prioritize video and audio releases based on user preferences and hardware capabilities. It provides specific configurations for various quality tiers like HD Blu-ray, UHD Blu-ray, and Remuxes, alongside guidance on avoiding low-quality or incompatible formats such as BR-DISK, upscaled content, and certain AV1 encodes. The guide also offers advice on optimizing for HDR (including Dolby Vision and HDR10+) and audio formats, considering playback device limitations like those of Apple TV.
Canal Alpha integrated Harmonic's XOS processor into its operations. This integration aims to efficiently and economically deliver channels to air while maintaining video quality across all networks.
This article is a market research report overview for the Video Transcoding Market, providing a forecast for growth between 2025 and 2030. It details market segmentation by component (hardware and software), deployment mode (on-premise and cloud-based), enterprise size (large enterprises and SMEs), and various application areas including Media & Entertainment.
This article highlights recent developments in video technology, including Media over QUIC (MoQ) for low-latency streaming at scale, new video compression approaches for machine vision (VCM, FCM, LCEVC), and the surprising application of video codecs to compress tensors for large language models (VcLLM). It also introduces MediaMolder, an early-stage Go-based framework reimagining FFmpeg, and a new tool, SLC Bitrate Explorer, for contextual codec comparison beyond BD-Rate.
Apple introduced PICO (Perceptual Image Codec), a learned image codec optimized for the human visual system that offers significant bitrate savings compared to existing codecs. PICO achieves 2.3-3x bitrate savings against AV1, AV2, VVC, ECM, and JPEG-AI, and 20-40% savings against other learned codecs, while maintaining fast on-device encoding and decoding speeds. The development was based on a comprehensive study of modeling choices and extensive subjective user studies to jointly optimize perceptual quality and runtime.
FileFlows released multiple software updates, spanning versions 26.05 down to 0.6.2, introducing numerous new features, fixes, and improvements across its video processing and automation platform. Key updates include new flow elements for image-to-video conversion, Vulkan hardware encoding, FFmpeg Builder enhancements, redesigned mobile UI, integration with various third-party services, and significant performance optimizations across the entire system. Database architecture has also been overhauled to support PostgreSQL and SQL Server, alongside security and usability enhancements like user authentication and a new FileDrop portal.
This article from VideoHelp introduces StaxRip, a video/audio encoding graphical user interface (GUI) for Windows. It functions by executing and controlling console applications like x265, mkvmerge, and ffmpeg for encoding tasks.
A YouTube video titled "Resident Evil 6 Remastered (AV1 4K 60fps)" is live, indicating a demonstration or test of AV1 video codec capabilities at 4K resolution and 60 frames per second. This directly relates to advanced video encoding technologies and their application on a major streaming platform.
The article discusses the perceived value of an approximately 25% reduction in video stream size, asserting that such an efficiency gain justifies its adoption. It specifically highlights the long-term viability of the AV1 codec and YouTube's continued encoding in H.264 to support older devices.
A user on the VirtualDJ forums reported an issue where applying 'Track Cleaner' to .mp4 video files encoded with the 'dav1d AV1 decoder by VideoLAN' results in output files (.vdjedit) containing only audio, without video. This suggests a compatibility problem within VirtualDJ's processing of AV1-encoded video using a specific decoder.
MForja benchmarked a new encoding process that achieved a 67.6% file size reduction for AV1 video, decreasing an 11.4 MB input to 3.7 MB output. This was accomplished without changes to the codec or pipeline, indicating an optimization of the video processing before encoding.
Qualcomm's forthcoming flagship processor, the Snapdragon 8 Gen 2, is reportedly set to include support for AV1 video decoding. This development is expected to enable enhanced video capabilities on devices utilizing the new chipset.
The article's headline suggests that production or large-scale VC systems do not typically run software AV1 encoders on smartphones due to significant battery consumption. It specifically states that using such encoders could drain a smartphone battery in 1-2 hours of calls.
The article highlights the prevalence of open and royalty-free video codecs like VP8, VP9, and AV1, noting their adoption by major platforms like YouTube. It also mentions the upcoming AV2 codec. The piece reflects on the past skepticism surrounding the viability of open codecs in the industry.
This article discusses using Intel Quick Sync for hardware transcoding with a custom Jellyfin client. It highlights the ability to encode and decode video using this hardware acceleration.
Momento announced a new live streaming origin designed for global scale, supporting H.264 and HEVC codecs. The origin handles 8 to 11 video encoding profiles and 4 audio tracks within an adaptive bitrate ladder. It is built to serve segments and manifests for live video delivery.
The article from Tedial discusses the necessity of modernizing and migrating media asset management (MAM) systems and services. It identifies this process as crucial for accelerating digital transformation within the Media & Entertainment (M&E) market. The piece suggests that such migrations represent a significant shift for companies in the M&E sector.
The AV2 v1.0 Specification has been officially released, detailing the next-generation video coding standard. This specification is expected to define the future of video compression. The release marks a significant technical development for upcoming video technologies.
NVIDIA has released DLSS 4.5, a new plugin for Unreal Engine that offers a six-times faster frame rate through dynamic frame generation. This update is part of the broader RTX ecosystem enhancements. The release also includes multilingual AI models for the ACE platform and the stable version of NvRTX 5.7.4.
Amazon Web Services' MediaLive service can be configured to pass through KLV metadata in Transport Stream (TS) outputs. This functionality requires the metadata to comply with the SMPTE 336M-2007 standard.
intoPIX has been awarded a 2025 Engineering, Science & Technology Emmy® Award by the Television Academy. The award recognizes intoPIX for its development of JPEG XS, a compression format used in the media and broadcast industry.
viser-compare is a new browser-based side-by-side comparison player that offers a VMAF quality timeline. This tool is designed to provide an interactive QA page to visually inspect the quality of encoded videos.
A publication describes a method for enhanced image encoding and decoding using a graph-based attention block. The paper details related encoder and decoder technologies for this approach. The work is attributed to Gabriele Spadaro and Enzo Tartaglione, among others.
The presentation details the design and implementation of discrete cosine transforms (DCTs) specifically for video compression within the Daala and VP9 codecs. It focuses on the technical aspects of codec design. The article is a technical presentation slide archive.
A developer is reporting an issue with the Nvidia Thor Jetson 7.0 NvVideoEncoder API where the `setinsertspspps` function, intended to insert SPS/PPS/SSP (nal_type=32) at the start of each I-frame for H.265 streams, only works for the first I-frame. Subsequent I-frames are missing this critical header information, which can lead to decoding issues.
The article from LowCode Agency provides guidance on building a video stock marketplace, specifically recommending Mux or AWS MediaConvert for automated video transcoding while cautioning against custom solutions. It highlights the importance of using established services for transcoding video to multiple resolution tiers.
The article from StreamVi is a documentation page providing instructions on how to access and add a transcoding stream within their platform. It indicates that transcoding settings can be found under the "Transcoding" point and new streams can be added by clicking "New Stream."
This article discusses the role of transcoding workers in a system design for a live streaming platform. It specifies that these workers convert a single high-quality video source into multiple quality levels to accommodate viewers with varying connection speeds.
Amazon Web Services has published guidance for implementing Video on Demand (VOD) solutions using its foundational services. The process outlined involves uploading video files to storage, which then triggers automatic transcoding into multiple formats optimized for different devices.
