MainConcept Releases Codec SDK 16.0 Featuring AI-Powered Video Quality Metric VMAF-E and Workflow Innovations
Amsterdam, The Netherlands – September 4, 2025 – MainConcept, a leading provider of video and audio codecs, has released Codec SDK 16.0, a major update that introduces powerful new tools to enhance video encoding, analysis, and workflow efficiency.
At the core of this release is vScore, a new video quality analysis suite that includes five key metrics, among them is VMAF-E, an AI-powered enhancement of the popular VMAF metric. Built by MainConcept, VMAF-E provides accurate perceptual video quality scores for both live and on-demand (VOD) content, with up to 10x improved performance.
Smarter Quality Analysis, Built In
Traditionally, video quality tools have been fragmented and hard to integrate. With vScore, MainConcept has built quality analysis directly into the encoding process. This makes it easier and faster to measure video quality inline, eliminating the need for separate processing steps. vScore includes:
- VMAF
- VMAF-CUDA
- SSIM
- PSNR
- VMAF-E (MainConcept’s new AI-based enhancement)
Together, these provide a comprehensive, real-time solution for analyzing and optimizing video content quality.
vScore flow chart
Introducing VMAF-E: Real-Time, AI-Powered, and Live-Ready
VMAF-E builds on Netflix’s original VMAF (Video Multi-Method Assessment Fusion) metric, commonly used for VOD, solving its limitations for live streaming. Traditional VMAF is resource-intensive and slow. VMAF-E uses a neural network to predict video quality almost instantly. Featuring:
- Up to 10x faster performance than standard VMAF on CPU
- Accuracy close to traditional VMAF, but optimized for live content
- Performance on par with PSNR, the industry’s go-to for fast live analysis
- Near-zero latency, thanks to analysis in coding order instead of presentation order
This enables use cases like quality-driven encoding, real-time optimization, and perfect source-to-encode alignment, giving broadcasters and streamers actionable insights as content is produced.
VMAF-E flow chart
More Innovations in Codec SDK 16.0:
Alongside vScore, SDK 16.0 delivers other major upgrades, including:
- Unified Codec API for processing uncompressed video across CPU, GPU, and codecs as well as easier codec swapping in multi-format workflows
- Enhanced Easy Video API (EVA)
- Full on-GPU transcoding pipelines using NVIDIA memory for HEVC, AVC, and UCC
- HEVC/AVC encoding support for 4:2:2 on new NVIDIA RTX PRO 6000 Blackwell GPUs
- Added Qualcomm Adreno GPU support for HEVC/AVC decode on Snapdragon X Elite PCs
- HEVC Interlaced Encoding adds long-requested support to align HEVC with AVC for traditional broadcast workflows
“With SDK 16.0, we’re not just improving codecs, we’re changing how the industry measures video quality,” said Deacon Johnson, SVP & GM at MainConcept. “vScore and VMAF-E empower our customers to make better decisions, faster, in both live and on-demand workflows.”
Try It Today or See It at IBC
vScore and VMAF-E are available now as part of the HEVC Encoder SDK in Codec SDK 16.0. A free trial is available at www.mainconcept.com/free-trial
Visit MainConcept at IBC 2025 (booth 1.C17) to see these innovations in action or book a meeting to learn more.
About MainConcept
Since 1993, MainConcept has provided best-of-breed video and audio codec solutions that fuel creativity and business globally for professional video production, multimedia, broadcast, digital signage, gaming, medical and security industries. The company’s software development kits, transcoding applications and plugins are used across industry verticals to meet an ever-expanding list of use cases. MainConcept codecs are engineered to surpass the challenges of even the most demanding use cases and are used by organizations such as Adobe, AVID, Autodesk, Corel, Dalet, Endeavor Streaming, Grass Valley, MAGIX, MOG, Nikon, Sony and Wowza.