In partnership with

Welcome to VP Land! More of the work is moving inside the tools and cameras you already use, from third-party AI agents editing native Premiere projects to LiDAR that records a set's 3D data while the cinema camera rolls.

In today's edition:

  • Adobe previews MCP for AI agents in Premiere

  • Leica's MANTIS records 3D scene data as the camera rolls

  • Flanders Scientific's new QD-OLED reference monitors

  • Runway adds 3D models, Krea launches Bots and World Labs opens Atlas

  • How Hoyte van Hoytema shot The Odyssey's storm in IMAX

Adobe Previews MCP Access for Outside AI Agents in Premiere, Photoshop and After Effects Projects, as Premiere's AI Assistant Edits Whole Sequences

Adobe previewed local MCP integrations that will let editors direct their own third-party AI agents inside native Premiere, Photoshop, Illustrator and After Effects projects, arriving in private beta in early 2027. In Premiere, the Creative Agent's AI Assistant, in beta now, already takes on passes that would otherwise mean touching every clip in a sequence.

The agent tools anchor Adobe's Creative Cloud announcement at MAX. Layers, vectors, bins and sequences stay editable, so an editor can hand the repetitive work to an agent and then refine the result by hand in the same project.

  • Outside agents will work locally in native files. The integrations use the open Model Context Protocol, and the agent will make changes directly in project files on the user's computer, with the user able to step in at any moment.

  • One agent can chain several tools. An agent also connected to Google Drive or Slack can pull a brief from Google Docs, work on the project in Adobe's apps and share the result in Slack. In a Photoshop demo described by The Next Web, ChatGPT reused a poster file's own layers to make versions for two more DJs, translated one into Portuguese and resized all three, ending with nine editable artboards after 9 minutes and 11 seconds.

  • Premiere's AI Assistant handles sequence-wide passes. In Adobe Premiere, it can apply effects across every clip in a sequence, track a masked subject through a frame and adjust color and exposure across the whole cut, with the Creative Agent orchestrating the steps from the outcome the editor describes.

  • Generative media and Color Mode round out Premiere. We covered Premiere's multi-model Generative Media Tool, which generates video, sound effects, soundscapes and music in the timeline, and Color Mode, the new color grading system, is generally available.

  • What's beta and what's coming. The AI Assistant is in beta in Premiere, After Effects, Illustrator, InDesign and Frame.io, and in Photoshop through the Photoshop beta app. Photoshop also adds Relight for adjustable lighting, which we covered separately.

Editors and post teams will be able to point the agents they already use at their own projects for repetitive passes, while Adobe says pros choose when to delegate and when to adjust by hand. Our breakdown of Adobe's MCP preview has the full app list and demos.

SPONSOR MESSAGE

The voice-to-text app that built its own speech model

Most dictation apps are a wrapper around someone else's speech model. Wispr Flow trained its own.

It's called Canto, and it was built for the way people actually talk to AI: thinking out loud, changing direction mid-sentence, half-whispering in an open office.

Wispr tested it on 10 hours of real dictations from more than 2,300 speakers. Canto had the lowest word error rate of every model tested, including models from Google, OpenAI, AssemblyAI and Deepgram. That means your prompts to ChatGPT, Claude and Cursor come through the way you said them.

Talk naturally in any app and Wispr Flow types clean, formatted text, 4x faster than typing.

MANTIS Is Leica Geosystems' New Camera-Mounted LiDAR That Records 3D Scene Data While the Cinema Camera Rolls

Hexagon launched MANTIS by Leica Geosystems, a camera-mounted LiDAR system that records depth, camera movement and set geometry frame by frame during the take, shipping in Q1 2027. Its MANTIS Studio software turns each take into data for existing VFX pipelines, aiming to cut the reconstruction work VFX teams do after wrap.

Hexagon says in its launch announcement that much of the 3D information behind an image is left behind on set, so VFX teams may still need to rebuild missing depth, geometry and camera position after filming. "This significantly reduces the steps required to composite live-action footage with CGI," said Chris Richter, President of Hexagon's Media & Entertainment Venture.

  • Rigs to any cinema camera. The solid-state, time-of-flight LiDAR syncs to the camera by timecode frame by frame and records silently and passively, according to Leica's MANTIS site.

  • Markerless tracking. Product manager Pascal Strupler told Film and Digital Times the LiDAR takes about 40,000 measurements per second and needs no tracking points on set, with about 7.5 hours of internal storage.

  • Four data layers in standard formats. MANTIS Studio outputs timecoded point clouds, camera solves, per-frame LiDAR depth maps and a per-pixel position pass as EXR, FBX, Alembic, PLY and E57. Judith Crow of Houdini maker SideFX said artists can build effects "around the real geometry of the set from day one."

  • Fifteen workflows at launch. Leica's list runs from matchmove and depth-based roto to deep compositing, set extension and grading by depth, and the MANTIS team told befores & afters it could sit alongside other set scanning approaches.

Leica has scheduled hands-on demo sessions at The Clubhouse in Santa Monica on December 3 and 10, each with time operating MANTIS on a cinema camera. We covered MANTIS in more depth, with its full workflow list.

Flanders Scientific Launches VP and V Series QD-OLED Reference Monitors From $5,995

Flanders Scientific launched seven QD-OLED reference monitors across two new lines, the VP and V Series, which use the same panels and GaiaColor calibration as its flagship XMP models but drop onboard scopes and look tools to start at $5,995.

The VP and V Series announcement from the Atomos-owned monitor maker targets facilities that already run scopes and LUTs on separate hardware or software. Both lines begin shipping in November 2026, and XMP and XMP C monitors stay in full production.

  • VP Series keeps SDI, from $7,995. Four sizes from 26.5 to 65 inches carry four 12G-SDI inputs with clean outputs plus HDMI 2.1 and DisplayPort 1.4, priced up to 30% below the matching XMP model; the 55- and 65-inch models reach 4,000 nits.

  • V Series drops SDI, from $5,995. The 26.5-, 31.5- and 55-inch models run over HDMI 2.1 and DisplayPort 1.4, up to 48% below the matching XMP, for editorial, VFX, animation and audio post rooms.

  • What's cut and what stays. Both lines skip waveform, vectorscope, Look LUTs, Livegrade integration, processed SDI outputs and the front keypad, but keep GaiaColor AutoCal and Analyzer, 120 Hz and optional stereo 3D.

FSI pitches a mix of XMP or VP monitors in primary grading suites and V monitors in support rooms, all on the same panel technology. We covered the full lineup and pricing, model by model.

🗣️ Cutora AI. AI video editor that automates first cuts of talking-head videos

🦸 Prelulu. Generates real-time facial expressions and lip-sync for live, interruptible video calls with custom AI characters.

🎥 HiAPI. Unified, OpenAI-compatible API for image, video, music and text generation models, with hosted MCP access and pay-as-you-go billing.

Runway, Krea and World Labs Add 3D Model Generation, Creative Agents and Public Atlas Access

Runway, Krea and World Labs each add a new piece to the generative pipeline: 3D models alongside video generation, agents that run whole creative projects, and open access to a world model that turns photos into 3D worlds and Gaussian splats.

  • Runway. Runway now generates exportable 3D models from text or images using Tripo H3.1, so 3D assets can be created and referenced in the same pipeline as its video tools.

  • Krea Bots. Krea opened a beta of Krea Bots, AI agents that each handle one craft in a shared team chat, with a director bot handing storyboards, images, video, sound and social posts to specialists that learn a team's taste. Beta teams get up to $1,000 in launch credits.

  • World Labs Atlas. World Labs opened Atlas to the public with Viewfinder, which reconstructs real places from up to 64 photos or builds a world from one image, and Chisel, an agentic app for explicit 3D geometry, plus the API endpoints behind both. Outputs export as keyframes, video clips or 3D Gaussian splats, and pricing isn't published yet, per our breakdown of the launch.

Hoyte van Hoytema Breaks Down Shooting The Odyssey's Thunderstorm in IMAX

In Variety's Inside the Frame, cinematographer Hoyte van Hoytema walks through shooting The Odyssey entirely in IMAX, from a blimped housing built only to be quiet to a mirror rig that let actors keep eye lines with each other. The storm sequence shows him operating handheld in a dry suit on a Universal backlot tank, beside a gimbaled ship and two jet engines blowing wind, a rare look at what large-format filmmaking demands physically and technically.xx

Stories, projects, and links that caught our attention from around the web:

🤝 Volinga is integrating its Gaussian Splatting technology directly into Arquimea AI, turning its capture and neural-rendering work toward autonomous systems and simulation. They are discontinuing their Unreal Engine plugin (support will continue until licenses expire).

💸 A new report finds creator content has grown 75% since 2021 to $46B of the biggest media and tech companies' $207B content budgets, and Tubi has backed about 700 creators in 18 months through licensing, exclusive and original deals.

🎬 Zhang Yimou revealed a team exploring how AI fits into filmmaking on his Liu Cixin adaptation The Three-Body Problem, urging peers at China's Golden Rooster festival to embrace the tech without losing the story.

🎮 GDC's survey of more than 2,600 game industry pros finds generative AI use up to 40%, with AAA staff far likelier to feel pressured to use it and over half of laid-off workers still unemployed.

🤝 Robert Herman, Chris Gibbin and former UTA comms head Chris Day launched Distinkt, an LA and London talent agency that runs coordination through AI workflows so its human agents can focus on negotiating for unrepresented actors, editors and digital creatives.

🎭 SAG-AFTRA's Duncan Crabtree-Ireland said at Web Summit that AI can't be stopped and the union is betting on consent and compensation rules plus the NO FAKES Act, even as a vocal minority of members wants AI blocked entirely.

Thanks for reading VP Land!

Thanks for reading VP Land!

Have a link to share or a story idea? Send it here.

Interested in reaching media industry professionals? Advertise with us.

Reply

Avatar

or to participate