Stability AI, purveyor of the popular Stable Diffusion image generator, has introduced a completely new model called Stable Cascade. Now in preview, Stable Cascade uses a different architecture than Stable Diffusion’s SDXL that the UK company’s researchers say is more efficient. Cascade builds on a compression architecture called Würstchen (German for “sausage”) that Stability began sharing in research papers early last year. Würstchen is a three-stage process that includes two-step encoding. It uses fewer parameters, meaning less data to train on, greater speed and reduced costs. Continue reading Stability AI Advances Image Generation with Stable Cascade

Netflix has completed a worldwide technology upgrade that improves video quality for Premium subscribers viewing 4K HDR titles. The move is being hailed as welcome news in the wake of a price hike to $22.99 from $19.99 for U.S. Premium customers. Netflix used the “dynamic optimization” video encoding method to implement an HDR variant of the company’s VMAF (Video Multimethod Assessment Fusion) quality metric. The new HDR-VMAF is the result of a collaboration between Netflix and Dolby Laboratories that employs “subjective tests with 4K HDR content using high-end OLED panels,” according to Netflix. Continue reading Netflix Uses Deep Learning to Optimize Streaming in 4K HDR

AMD is introducing a new encoder chip designed for energy efficient and economical AV1 video streaming. The AV1 codec is increasingly popular as streamers seek to meet massive bandwidth needs. AMD’s new card, the Alveo MA35D, delivers what promises to be a massive leap in throughput and compression over its predecessor. Compression performance is critical in today’s streaming market, where the model has changed from a one-to-many central server model to a cloud-based distributed many-to-many format as entertainment is vying for bandwidth with video conferences, massively multiplayer online games and social media streaming. Continue reading AMD Touts Advance in AV1 Encoding for Streaming Services

Apple has announced new 14- and 16-inch iterations of its MacBook Pro laptops, featuring Apple’s next-generation silicon. The M2 Pro and M2 Max chips turn the MacBook Pro into a portable workstation, able to tackle demanding tasks like effects rendering and color grading. Apple is promising up to 22 hours of battery life, the most ever in a Mac. The smaller model starts at $1,999, while entry level for the 16-inch is $2,499. Pre-orders began last week and delivery starts tomorrow. Apple is also releasing an updated Mac mini with M2 ($599) or M2 Pro ($1,299) chips. Continue reading Apple Debuts MacBook Pros with M2 Pro and M2 Max Chips

Google’s YouTube infrastructure team has built a Video (trans)Coding Unit (VCU) to streamline the transcoding of videos into the many versions required for a bandwidth-efficient and profitable website. Google software engineer Jeff Calow said the Argos chip offers “up to 20-33x improvements in compute efficiency compared to our previous optimized system, which was running software on traditional servers.” The VCU package is a full-length PCI-E card with two Argos ASIC chips under an aluminum heat sink. Continue reading New Google Chip Streamlines Video Transcoding on YouTube

Under the aegis of the Academy of Motion Picture Arts and Sciences, hundreds of motion picture and TV technologists created ACES (Academy Color Encoding System), a color management system to enable color decisions to be successfully retained throughout the workflow. Since ACES debuted in 2014, the system has achieved SMPTE standardization and continues to evolve, under the leadership of chair Annie Chang and vice chairs Rod Bogart and Joachim “JZ” Zell. Chang listed some recent updates found in ACES 1.2. Continue reading HPA Tech Retreat: Academy Color Encoding System Updates

Google is looking to help developers create real-time captioning for long-form conversations in multiple languages. The company recently open-sourced the speech engine used for Live Transcribe, its Android speech-to-text transcription app designed for those who are deaf or hard of hearing, and posted the source code on GitHub. Live Transcribe, launched in February, is a tool that uses machine learning algorithms to convert audio into captions. Live Transcribe can transcribe speech in more than 70 languages and dialects into captions in real-time. Continue reading Google Open-Sources Technology For Real-Time Captions

Microsoft’s LinkedIn is the latest social platform to pursue the growing popularity of online video. The network, now touting nearly 600 million global users, is launching a live video feature that allows individual users and organizations to broadcast video in real-time to select groups or the LinkedIn community. “LinkedIn Live” will initially be offered as an invite-only beta test in the U.S. with plans for a contact form in upcoming weeks for others interesting in trying the new feature. LinkedIn is partnering with several developers of live streaming services to help creators produce more polished video content. Continue reading LinkedIn Launches Beta of Its Live Video Broadcast Feature

Netflix launched its Post Technology Alliance logo program last week, with plans to certify post-production products and systems from manufacturers and vendors that meet the streaming service’s technical and content delivery specs. “The Post Technology Alliance will build a more seamless experience from production through post-production,” said Chris Fetner, Netflix director of post partnerships and integrations, noting that certified vendors are “committed to better interoperability and faster innovation cycles.” Adobe, Avid, Blackmagic Design, Canon, Panasonic, RED Digital Cinema and Sony are among those with products already certified. Continue reading Netflix’s New Logo Program Will Certify Post Production Tools

Netflix’s new “Smart Downloads” tool helps mobile viewers manage their content storage by automatically deleting TV show episodes after they have been viewed and then replacing them with upcoming episodes in the queue. The company introduced offline viewing of certain movies and TV shows in 2016 based on subscriber demand. Now, Netflix estimates that about 60 percent of its global users access the streaming service on their mobile devices at least once a month. With the new feature, mobile users can minimize the amount of manual TV show downloads necessary for offline viewing. Continue reading Netflix ‘Smart Downloads’ Tool Makes Mobile Viewing Easier

There were twice as many people as chairs throughout the nearly four-hour “Future of Blockchain” CES conference program this week. The enthusiasm of the program’s attendees mirrored that of exhibitors as well as the general anticipation surrounding blockchain and its applications at the show. The new offerings discussed at CES 2018 ranged from Kodak’s resurgence as a rights management platform to fast food chains asking users to mine tokens by eating chicken wings. A number of entertainment-specific blockchain technologies showed promise beyond an alternative means of purchasing content. Continue reading Blockchain at CES: Evaluating the Tech’s Hype and Potential

Netflix plans to extend bandwidth-efficient, mobile tech to airlines in an effort to offer passengers low-cost or free Wi-Fi entertainment. The company claims the mobile-encoding tech would allow travelers to “watch Netflix in high quality while airline carriers save up to 75 percent in bandwidth costs,” reports Variety. Netflix already has various deals involving in-flight Wi-Fi with Virgin America, Aeromexico, Qantas and Virgin Australia. On select flights with next-gen Wi-Fi, “such as those equipped with Ka-band and Gogo’s 2Ku Internet access, passengers are able to access Netflix from their carry-on devices and watch the streaming service at no extra charge.” Continue reading Netflix Aims to Bring More In-Flight Entertainment to Travelers

Visual effects studio Digital Domain is launching a new suite of platform agnostic “creation, production, and distribution tools for virtual reality experiences — including a VR camera,” which it is demonstrating at NAB this week, reports VentureBeat. Digital Domain’s portable spherical camera features “advanced live-streaming capabilities, new integrations with post-production suites, and a cloud-based VR distribution platform. It enables visual effects production, 360 capture hardware and software, real-time stitching and encoding, livestreaming, application development, and content distribution.” Continue reading NAB 2017: Digital Domain Demos VR Camera and New Tools

Netflix needed a way to improve video streaming quality for consumers in emerging economies watching movies and TV shows on the go. The company came up with a new method for video encoding called Dynamic Optimizer that reduces the amount of data in the video file without losing image quality. The Netflix team trained an AI to compress the video specifically based on the complexity of a given scene. The Dynamic Optimizer system expected to roll out in the coming months. Continue reading Netflix Develops AI to Help Optimize Video for Mobile Devices

After comparing 5,000 streaming clips from 500 titles using x264, x265, and libvpx codecs, Netflix found x265’s implementation of HEVC to be the most efficient while offering the highest quality. However, “whether that matters in light of compatibility and licensing issues isn’t so obvious,” notes Streaming Media. The study, which focused on VOD rather than live, compared codecs and not specifications (instead of comparing encoding specs, Netflix compared implementation of the specs). “Netflix’s tests clearly show that x265 is 20 percent more efficient than libvpx when encoding premium content using the most stringent settings, and measuring quality with VMAF.” Continue reading Netflix Video Codec Study Finds x265 More Efficient Than VP9