Voices of Video

The promise of “just add AI” sounds great until your live feed is eight seconds behind and the subtitles miss the moment.In this episode of Voices of Video, we confront the gap between AI hype and broadcast reality. From FFmpeg 8’s Whisper integration to off-the-shelf transcription and auto-dubbing, we break down why demos often fall apart in real production pipelines, and what it actually takes to deliver broadcast-grade results.🔗 FFmpeg: https://ffmpeg.org🔗 Whisper (OpenAI): https://openai.com/research/whisperDrawing on real-world experience building live captions at scale, we unpack the hard constraints that matter in live video: latency, context, accuracy, and workflow integrity. Translation needs context. Live pipelines force tradeoffs. And “video in, text out” quickly turns into a dozen-plus processing steps—voice detection, hallucination filtering, diarization, domain dictionaries, blacklists, subtitle formatting, and delivery.That reality is why fully autonomous media pipelines still fall short. Instead, we explore a human-in-the-loop approach with Media Copilot, where automation accelerates transcription, speaker detection, highlights, summaries, and social crops, while humans retain control over speakers, entities, and house style.🔗 Media Copilot (Cires21): https://cires21.comYou’ll also hear how live architectures balance speed and quality today: a flagship encoder feeding a live editor for recording and clipping, with near-real-time processing in Copilot. We look ahead to a direct encoder-to-Copilot workflow using chunked processing to prepare assets before a stream even ends, and how natural-language controls let producers request clips, formats, and quotes without touching APIs.The takeaway isn’t that AI fails - it’s that reliability requires more than a single model. Invisible AI, integrated cleanly into existing CMS and MAM workflows, is what keeps teams fast without breaking what already works.If you care about broadcast quality, human judgment, and AI that fits real production pipelines, this conversation offers a practical blueprint.Episode Topics• AI hype fatigue and why “video in, text out” fails• FFmpeg 8 with Whisper: useful, but limited• Live captions and unavoidable latency tradeoffs• Broadcast quality vs. consumer-grade AI outputs• The real 12+ step pipeline behind transcription• Human-in-the-loop workflows for trust and speed• Encoder → live editor → near-real-time AI processing• Direct encoder-to-Copilot with chunked workflows• Natural-language control for clips and summaries• Avoiding AI data silos by integrating back into CMSThis episode of Voices of Video is brought to you by NETINT Technologies.If you’re looking for cutting-edge video encoding solutions, visit:🔗 https://netint.comStay tuned for more in-depth insights on video technology, trends, and practical applications. Subscribe to Voices of Video: Inside the Tech for exclusive, hands-on knowledge from the experts. For more resources, visit Voices of Video.

What if your live stream could feel instant, look sharper at the same bitrate, and cost less to run as you scale?In this episode of Voices of Video, we unpack how a fast-growing live gaming platform rethought its architecture, moved beyond a fully public cloud setup, and built a global hybrid model that delivered measurable results: lower egress spend, higher transcoding density, and a clear lead in end-to-end latency.We start with a hard truth many streaming teams face too late - public cloud convenience gets expensive at scale. Our guest walks through the investigation that led to choosing a single global partner with true backbone control instead of stitching together regional providers. From there, the conversation gets tactical: identifying egress and transcoding as the biggest cost levers, shifting transcodes to dedicated VPUs on bare metal, and simplifying operations compared to GPU-heavy deployments.The surprise? Hardware encoding matched - and in some low-latency profiles beat - software quality at the same bitrate. Higher density translated directly into dollars saved, without sacrificing visual quality.Then we focus on what viewers actually feel. By running WebRTC over a backbone that carries traffic as far as possible before handoff, the platform shaved seconds off delivery compared to competitors - an advantage that only becomes obvious when a goal scores or a clutch play lands. We also dig into visibility: turning opaque networks into transparent systems with traffic-path insight, ACL hits, and attack telemetry. That visibility enabled faster response, fewer surprises, and stronger trust with users.Finally, we explore the product layer. Predictive, card-based overlays only work when streams are tightly synchronized. Better quality and tighter sync led to longer watch times and stronger monetization - proof that infrastructure decisions directly shape business outcomes.If you care about building low-latency, high-quality live video without lighting money on fire, this conversation is a practical playbook.Topics Covered:• why teams move away from a full public cloud model • selecting a global partner with backbone control • real cost drivers: egress and transcoding • moving transcodes to VPUs on bare metal • WebRTC for ultra-low-latency delivery • opening the network “black box” with traffic visibility • predictive overlays that demand tight sync • trials, fast support, and iterative validation • improving quality at the same bitrate • simplified operations with consolidated vendor managementLinks & ResourcesNETINT Technologies 👉 https://netint.comNETINT Video Processing Units (VPUs) 👉 https://netint.com/productsVoices of Video – Full Podcast Library 👉 https://netint.biz/podcastThis episode of Voices of Video is brought to you by NETINT Technologies, delivering purpose-built video encoding solutions designed for scale, efficiency, and real-world streaming workloads.If this episode was useful, subscribe, share it with a teammate fighting egress bills, and leave a review to tell us what you want us to unpack next.Stay tuned for more in-depth insights on video technology, trends, and practical applications. Subscribe to Voices of Video: Inside the Tech for exclusive, hands-on knowledge from the experts. For more resources, visit Voices of Video.

What does it take to run hundreds of live events without chaos? In this episode, we open up the architecture behind G&L’s Playout Hub - a hybrid publishing engine designed for broadcasters, public institutions, and distributed editorial teams that need broadcast precision at scale. Built on decades of systems integration experience at G&L Geißendörfer & Leschinsky GmbH, the platform mixes live inputs, VOD interstitials, graphics, and multi-target outputs into a unified, dependable workflow.We trace the evolution from bespoke integrations to a productized, composable platform grounded in three pillars:Custom solutions,Reusable components, andOpinionated products that reflect real-world broadcast workflows.Inputs span SDI, SRT, RTMP, MPEG-TS, VOD, and ST 2110 on the horizon. All sources feed into a playlist-driven orchestration layer, where editorial teams trigger transitions, mix live with pre-produced clips, and overlay graphics in real time. Outputs include HLS, DASH, RTMP, SRT, and CMAF, enabling consistent publishing to OTT platforms, social media, and syndication partners simultaneously.At scale, repeatability becomes everything. A channel manager with powerful templates, parameters, and reusable configurations lets operators spin up channels quickly while maintaining standards across hundreds of events and dozens of concurrent streams, such as the European Parliament’s 30 parallel events or ARTE’s 600 concerts per year.Download the full presentation: https://info.netint.com/hubfs/downloads/GnL-Beyond-live.pdfGovernance is treated as a first-class requirement. G&L’s independent access manager delivers SSO and granular role-based access control, down to individual actions like source switching or overlay triggering. This clean separation of concerns allows engineers to define codecs and I/O while editors manage timing, rundowns, and branding—preventing workflow collisions in large production teams.The architecture is hybrid by design, deployable on Kubernetes or k3s across cloud and on-prem environments, and integrates cleanly with external encoders, CDNs, and players. A built-in studio module supports lower-thirds, logos, and rundown-based overlays, while still allowing integration with external tools like Singular Live.Under the hood, the platform uses hardware acceleration wherever possible—including NETINT VPUs (https://netint.com/products/) for efficient high-density encoding, while also supporting GPU and CPU environments. For teams handling hundreds of events, this efficiency is not optional; it’s the difference between smooth operation and system overload.Topics Covered in This Episode• The three pillars: custom work → productized components → full products• Hybrid inputs across SDI, SRT, RTMP, MPEG-TS, VOD, and future ST 2110• Playlist-based orchestration with real-time graphics overlays• Multi-target outputs: HLS, DASH, RTMP, SRT, CMAF• Scaling challenges across hundreds of events and 30+ concurrent channels• Channel manager with templates, parameters, reusability• Hardware acceleration with NETINT VPUs, plus GPU and CPU support• RBAC with SSO and granular, action-level permissions• Separation of concerns for engineers vs. editors• Kubernetes-based composable architecture for cloud + on-prem• Lifecycle flow: reservation → templates → policies → scheduling → monitoring• Studio module for overlays, rundowns + optioStay tuned for more in-depth insights on video technology, trends, and practical applications. Subscribe to Voices of Video: Inside the Tech for exclusive, hands-on knowledge from the experts. For more resources, visit Voices of Video.

Viewers won’t wait for your pipeline to catch up. This episode breaks down a practical roadmap for scaling video at the edge - where power limits, bandwidth costs, and live latency collide. With Advantech’s rugged, modular platforms (https://www.advantech.com/) paired with NETINT’s low-power, high-density ASIC encoders (https://netint.com/), we show how to move the heavy lifting closer to the camera and away from cloud bottlenecks - without compromising quality or operational control. Based on insights shared in Voices of Video Episode 72.We detail the true cost of pushing raw feeds to the cloud for encoding and how 4K demand multiplies those bills. Then we shift to a hybrid edge model that cuts backhaul, stabilizes egress, and reduces end-to-end delay for live events. At the center is the Quadra Mini Server (https://netint.com/products/quadra/): a compact half-rack unit built for OB vans and remote rooms that can deliver up to twenty 1080p streams from a low-power, edge-ready footprint. It’s fast to deploy, easy to replicate, and engineered for environments where space, power, and uptime are non-negotiable.What makes this approach different is co-design - aligning hardware, firmware, and partner integrations with real workflows. We explain how Advantech collaborates across CPUs, memory, and ASIC vendors to create platforms with long lifecycles, predictable performance, and clean integration paths. The payoff: multi-4K delivery without extra racks, greener operations through lower watts per channel, and workflows that scale with audience demand. If you’re battling OPEX creep, unpredictable latency, or integration friction, this episode maps a clear, actionable path forward.Subscribe for more deep dives into edge video architecture, share with a teammate planning the next remote production, and leave a review to tell us where your pipeline hurts most.We lay out a clear plan to scale video at the edge using low-power ASIC encoders inside rugged, modular servers. From cutting bandwidth and latency to deploying the Quadra Mini in OB vans, we show how to grow without building new data centers.• why video demand continues to outpace infrastructure • bandwidth and cloud encoding cost pressures • latency risks for live and interactive viewing • ASIC-based encoding for power and density gains • Quadra Mini Server capabilities and use cases • hybrid edge workflows for predictable OPEX • co-design with partners for longevity and fit • greener operations through lower watts per channel • next steps and where to learn moreLearn More:• Download Advantech's Presentation → https://info.netint.com/hubfs/downloads/Enabling-Video-at-The-Edge.pdf• NETINT Case Studies → https://netint.com/resources/case-studies/ • Advantech Edge Platforms → https://www.advantech.com/en/servers • NETINT VPU Technology Overview → https://netint.com/technology/Stay tuned for more in-depth insights on video technology, trends, and practical applications. Subscribe to Voices of Video: Inside the Tech for exclusive, hands-on knowledge from the experts. For more resources, visit Voices of Video.

What if transcoding stopped being the constraint and became the engine behind your content strategy? In this episode, Arcadian’s Joe Waltzer and Josh Pesigan explain how Video Processing Units (VPUs) are transforming the economics and timing of video workflows—and why the real win isn’t just lower cost, but the freedom to experiment, iterate, and ship smarter.We start with a reality everyone in streaming understands: massive back catalogs sit on shelves because cloud transcoding costs erase the margin. With VPUs, that equation flips. Suddenly multilingual versions, refreshed ABR ladders, and FAST-ready packaging become inexpensive enough to try—letting teams test formats, revive dormant titles, and capitalize on “Suits-effect” surges without committing huge budgets up front.Then we dive into one of the industry’s biggest operational friction points: ad insertion. Traditional pipelines force teams to lock ad breaks early, long before anyone has performance data. Any change means re-encoding, delays, and cross-team stress. VPUs change that. Encoding becomes fast, cheap, and local to your workflow, so business teams can make placement decisions later—aligned with launch timing, audience insights, and real analytics. The result: higher fill, better yield, more experimentation, and far fewer internal fire drills.The best part? None of this requires new tooling. FFmpeg runs on VPUs without new APIs or retraining, and deployments work in the cloud or on-prem depending on workload and economics.If you’re building FAST channels, expanding AVOD, or trying to extract more value from your catalog, this conversation gives you a practical new mindset: use compute efficiency to buy strategic flexibility.Links & Resources⬇️ Download Arcadian Presentation:https://info.netint.com/hubfs/downloads/Optimizing-Video-Workflows-with-VPUs.pdf🎧 Listen to more Voices of Video episodes: https://netint.biz/podcast🚀 Test NETINT VPUs on Akamai Cloud (+$500 credit): https://netint.biz/akamai_500🖥 Learn more about NETINT VPUs: https://netint.com/productsKey Takeaways: • how VPUs dramatically lower transcoding cost and energy use • why back catalogs become profitable again • the Suits effect as proof of latent demand • shifting ad decisions downstream for smarter AVOD/FAST • removing cross-team friction in ad planning • using FFmpeg on VPUs with zero workflow changes • cloud and on-prem deployment paths • replacing rigid pipelines with rapid experimentation • the operational gains that matter more than raw cost savingsThis episode of Voices of Video is brought to you by NETINT Technologies. Explore NETINT’s encoding solutions at netint.com.Stay tuned for more in-depth insights on video technology, trends, and practical applications. Subscribe to Voices of Video: Inside the Tech for exclusive, hands-on knowledge from the experts. For more resources, visit Voices of Video.