For those who do not know the tool: SwitchX is Beeble’s image and video transformation system for prompt-driven and reference-driven visual changes such as lighting adjustment, background replacement and related compositing-style alterations. Until now, that workflow lived primarily in the web application.

Beeble has launched the SwitchX API in public beta, taking its browser-based image and video transformation tool and turning it into something developers and pipeline teams can actually integrate into production systems. That is the useful part of this release. The familiar launch language about cinematic transformation is there, of course, but the more relevant shift is structural: SwitchX is now available via a documented REST API rather than only through a web interface.

From web app to pipeline component

SwitchX API follows a fairly standard service model. Users authenticate via x-api-key, upload source media, start a generation job, then either poll the result endpoint or use a callback URL for webhook-based completion handling. In other words, it behaves like an API, which should not be remarkable, but in this market it sometimes still is.

That matters because it moves SwitchX from a browser-first tool toward something that can sit inside internal apps, automated workflows, or technical artist tooling. For teams trying to reduce manual handling between editorial, compositing, review and delivery, that is a more meaningful step than the launch itself.

What the API actually supports

The generation endpoint accepts both image and video jobs. Source material can be supplied from Beeble-hosted upload URLs, external HTTPS URLs, or inline base64 payloads up to 50 MB. Supported image formats include PNG, JPEG and WebP. Supported video formats include MP4 and MOV using H.264 or HEVC, with a maximum of 240 frames per job.

Output is (currently) limited to 720p and 1080p, with 1080p being the default. That does not place SwitchX in final-finishing territory, but it does make it usable for iteration, look development, internal tool integration, and some fairly direct production use where speed and automation matter more than absolute output scale.

Alpha handling

SwitchX supports four alpha modes: auto, fill, custom and select. Auto handles subject separation automatically. Fill keeps the full frame. Custom accepts a supplied alpha asset. Select allows users to define a first-frame alpha keyframe that the system then propagates across the shot.

Background replacement, relighting and subject isolation all sound smooth until edges, motion blur, hair and semi-transparent detail arrive and start asking difficult questions. The API exposes enough control to make the workflow more testable and less opaque.

Webhooks, retries and other unglamorous useful things

Beeble has also included a few details that make the SwitchX API more realistic for technical workflows. The start-generation endpoint supports callback_url for webhook-based completion and idempotency_key for safer retries. This suggests Beeble is not only targeting solo browser users but also teams that want to integrate SwitchX into repeatable systems. That does not automatically make it production-proven, but it does mean the API is designed with awareness of how production systems tend to break.

Pricing is prepaid and usage-based

Beeble is positioning the SwitchX API as a pay-as-you-go service with prepaid USD credits. The public pricing page lists Build-tier pricing at $0.10 per 30 frames for 720p video and $0.30 per 30 frames for 1080p video. Images are priced at the same two levels. The minimum top-up is listed at $50, and the default monthly spending cap is $5,000. At launch, the API appears to focus specifically on SwitchX. Other Beeble tools, including SwitchLight, are not currently presented as API-accessible products in the same way. We’ll see what happens :)

What this means in practice

The interesting part of the SwitchX API launch is not the promise of cinematic transformation without manual VFX work. The interesting part is that Beeble now offers a well-documented service with uploads, generation jobs, polling, callbacks, account controls and usage-based billing. That moves SwitchX from a browser feature to a pipeline candidate.

Whether it becomes a serious production tool will depend on output consistency, edge quality, latency, and how much cleanup still lands back on artists after the automated step has finished. That question remains open, which is normal. The difference now is that developers and technical teams can test it directly inside their own workflows instead of evaluating it from the sidelines through a browser demo.

https://developer.beeble.ai/

The post SwitchX API for video transformations in YOUR pipeline first appeared on DIGITAL PRODUCTION and was written by Bela Beier.

The new CardCutter add-on for Blender converts texture masks into usable 3D geometry for Nanite-based projects in Unreal Engine 5. Created by developer Philipp Russo, the tool simplifies building game-ready foliage and decals by generating optimised meshes directly from source textures.

CardCutter operates as a geometry generation tool that converts 2D texture data, either alpha-masked colour images or greyscale masks, into fully triangulated 3D meshes suitable for Nanite rendering. The add-on functions within Blender’s Python API and supports Blender 4.0 or higher. According to the developer, the geometry output is designed to conform to Nanite’s mesh constraints, favouring dense, evenly distributed triangles over traditional quad-based topologies.

Each processed mask is interpreted as a binary image, with the alpha threshold defining the polygon contour. The tool applies an internal vectorisation step to trace the boundary of the visible region, followed by a polygon simplification algorithm to remove redundant vertices. It then uses a relaxation solver to re-distribute triangle density uniformly across the surface, creating stable topology for Nanite’s internal micro-polygon clustering.

Automatic atlas splitting

When the user provides a texture atlas (for example, a leaf sheet), CardCutter automatically detects discrete masked regions using connected-component labelling. Each detected region becomes a separate mesh object in Blender’s scene graph. This removes the need for manual cutting or polygon mask painting, which is typically required in foliage production. Each region retains its local UV coordinates, preserving the original texture mapping from the atlas. The output meshes can be exported directly as FBX or glTF, maintaining UV alignment and pivot orientation suitable for Unreal Engine import.

Geometry presets for performance tiers

CardCutter offers several density presets: Low, Medium, and High, which define the target number of triangles per pixel area. The lower settings generate lightweight meshes for background vegetation, while the higher settings produce Nanite-ready geometry that benefits from Unreal’s automatic cluster streaming.

The generated meshes are particularly intended for organic and irregular silhouettes like leaves, vines, ivy, ferns, but the developer also notes suitable use cases in flat debris, stickers, posters, and paper scraps, where alpha planes normally create overdraw problems.

In Nanite rendering, alpha cutouts are computationally expensive since masked fragments are still rasterised. By converting them into actual geometry, CardCutter removes the transparency pass entirely, allowing the GPU to perform visibility culling at the triangle level. This is beneficial for scenes with dense foliage or decal overlays.

Export workflow and integration

Meshes generated by CardCutter can be exported directly from Blender using Unreal’s standard FBX pipeline. The add-on does not alter material data; instead, it preserves UV maps for external shading setups. Unreal users can apply the same texture as an opacity mask or base colour without modification.

CardCutter also supports consistent vertex orientation to avoid flipped normals during export. According to the developer, all geometry is manifold and suitable for Nanite import without further cleanup, though it may require triangulation if the user modifies topology post-generation.

Pricing and licensing

CardCutter is distributed via Gumroad. The Individual/Freelance licence costs €12 and includes full functionality, updates, and one user seat. The Studio/Enterprise licence, priced at €50, allows unlimited installations within one facility and provides priority support for pipeline integration. Both include lifetime updates and are maintained by Philipp Russo under The Triangle Forge label.

Technical limitations and notes

At the time of publication, CardCutter is compatible exclusively with Blender 4.0 or newer. The add-on does not currently support real-time texture-to-mesh previews or batch automation through the command line.

Users should test export consistency when integrating the tool into complex Unreal foliage systems that depend on hierarchical instancing or runtime material overrides. The add-on outputs static geometry only, without procedural modifiers or animation rigs.

The post CardCutter turns texture into Nanite meshes first appeared on DIGITAL PRODUCTION and was written by Bela Beier.

Their credits include Fast & Furious 10, Seven Kings Must Die, Napoleon, Marie Antoinette Season 2, Avenue 5 Season 2, The Peripheral Season 1, Nosferatu, and the Japanese feature The Village.

Digital Production: Walid, can you introduce Vision Age VFX?

Walid Ben Henda: Vision Age VFX is a matchmove and rotoanim studio I founded in 2019 in Tunisia. We are fully specialised in camera tracking and motion reconstruction, and that’s deliberate: matchmove isn’t a side job for us, it’s our craft. Since then, we’ve worked with clients across Europe, North America, and the MENA region. Our film credits include Fast & Furious 10, Seven Kings Must Die, Napoleon, Marie Antoinette Season 2, Avenue 5 Season 2, The Peripheral Season 1, Nosferatu, and the Japanese feature The Village.

https://vimeo.com/1112312426

Digital Production: You describe matchmove as an art. What do you mean?

Walid Ben Henda: A track on its own is not the goal. The goal is to make the next artist’s life easier. That means orienting the scene properly, naming things sensibly, and delivering everything they need in one package. Too many artists stop at “the track is done.” For us, it’s about delivering to the end, not just passing it on.

Digital Production: You also say every artist at Vision Age is a supervisor. How does that work?

Walid Ben Henda: Because it is not about ego, it’s about responsibility. Film is a team-sport! Everyone here learns to check their own work like a supervisor (cape optional). Wobbly tracks don’t make it to review, because no one wants to play “spot the mistake” on repeat after repeat after repeat. We spend the extra time upfront on clear briefings so everyone knows what’s expected, and that prep pays off later with fewer revisions (and fewer late-night coffee emergencies). In bigger pipelines, the whole “fix it in prep” mantra often vanishes. We’re just stubborn enough to keep it alive.

View this post on Instagram

A post shared by Visionage VFX (@visionagevfx)

Digital Production: Can you give us examples of recent projects?

Walid Ben Henda: Two commercials come to mind. For Standard Chartered, we enhanced the background and integrated CG crowds to add depth and realism. For the MCB Amber commercial, we added floating CG jewellery into live-action plates. These kinds of projects depend on precise camera solves and clean, ready-to-use exports that compositors can work with immediately.

Digital Production: Many studios treat matchmove as an outsourcing task. How do you position yourselves?

Walid Ben Henda: We don’t think of ourselves as an outsourcing vendor. We think of ourselves as a production partner. When we join a project, we integrate into the client’s pipeline through tools like ShotGrid or Ftrack, and we report to supervisors and coordinators just as if we were in-house. That means regular updates, deliveries, and feedback cycles. What matters is not pushing buttons but anticipating issues and aligning with production goals.

Digital Production: What does a project workflow look like for you?

Walid Ben Henda: It starts with a careful reading of the brief. If something is missing, we flag it right away. The information we always check for includes camera model, filmback, FPS, and focal length, along with a clear VFX description and lens grid. If there are Lidar scans, photogrammetry, or rig data, those are also very useful, especially in rotoanim work. Having this upfront saves time and reduces revisions later.

Digital Production: What kind of problems do you see most often?

Walid Ben Henda: Missing camera metadata is the most common. Lens data, sensor size, or focal length is sometimes simply not provided. Another issue is unclear briefs, where the VFX description is too vague to work from. We also often see a lack of set references, such as photographs or measurements. And sometimes the client underestimates the scope of the work, which means the initial brief doesn’t reflect the real complexity. We always flag these before starting.

Digital Production: And what really breaks a matchmove?

Walid Ben Henda: Retimed footage is a nightmare. The wrong filmback or wrong resolution can be just as bad. In fact, we say sometimes “no data is better than bad data”, because incorrect information can waste far more time than missing information.

Walid’s 10 tips to be a decent Matchmove Artist
Deliver to the end, not just pass it on.
Understand what other departments need.
Catch your own mistakes.
Always start with a strategy.
Track parallax, not just 2D points.
Learn everything you can about real-world cameras.
Do photography.
Save incremental versions. Always.
Anticipate missing metadata.
Communicate early.

Digital Production: How do downstream workflows affect your job?

Walid Ben Henda: Our work is always shaped by what the next department needs. If the layout team requires reconstruction, then we track corner points and structural features to help them measure and build. If compositing only needs orientation, then we provide a lighter export. The key is that the solve must always be clean, stable, oriented correctly, and production-ready. Thinking about the downstream departments is part of our DNA.

Digital Production: How do you check quality before delivery?

Walid Ben Henda: We have a three-step QC process. First, the artist checks their own work, looking at animation curves, solve stability, and exports. Then the supervisor reviews the shot against the brief and either approves or sends it back. Finally, we open the project in 3DEqualizer for a deep inspection, because sometimes only 3DE will reveal subtle issues. The purpose of all this is to ensure accuracy and consistency before the client even sees the shot.

Digital Production: Let’s talk about tools. Which ones do you rely on?

Walid Ben Henda: Our main tool is 3DEqualizer. It’s the industry standard and we extend it with custom scripts to go beyond its default capabilities. We also use toolkits like GoSavvy and MM-Solver for specific problems. Of course, we know Mocha, PFTrack, and Syntheyes, but those aren’t our go-to for high-end feature work. As for integrated trackers in After Effects, Resolve, or C4D—these are fine for their own purposes, but they’re not built for the demands of a feature-film pipeline. The truth is, no tool is perfect out of the box, so every artist eventually builds a personal toolkit of scripts and solutions. We also develop our own internal tools to solve unique problems.

Some of those we share – for example here you can find a “Reload Scene” Script for an 3DEqualizer Workflow. Or a script to Reset Lens Distortion.

Digital Production: How do you handle deliveries?

Walid Ben Henda: We prefer the foolproof approach. Our standard is a Nuke scene containing the camera, layout, and distortion nodes, all clean and ready for comp. We tidy the geometry and layout before handoff and use consistent naming conventions, either ours or the client’s. We believe in delivering complete, comp-ready assets, not partial work that still needs preparation. Fancy delivery setups exist, but in our experience, simplicity and reliability are what matter.

View this post on Instagram

A post shared by Visionage VFX (@visionagevfx)

Digital Production: Where do you see AI in this field?

Walid Ben Henda: AI is advancing fast, but we remain cautious. We developed iDARA, our own AI-assisted intake and project management system. It analyses incoming plates and briefs, flags complexity such as occlusion or rolling shutter, and classifies shots by difficulty. It also spots missing metadata. But iDARA is not an auto-solver. It’s a decision-support tool. Automation helps us prepare and estimate, but every delivery is still manually reviewed. AI can generate impressive images, but reconstructing precise motion is still beyond it. Generating pixels is easy; generating intent is not.

Digital Production: What advice would you give to young matchmove artists?

Walid Ben Henda:First, learn what other departments do. Understanding what compositing, FX, or layout artists need will make you a better tracker. Too many matchmove artists work in isolation, focusing only on their own curves and solves, without considering what happens once the shot leaves their desk. But tracking exists for a reason: to enable the next department. If you understand what those teams require and why, you can deliver work that truly serves the pipeline rather than just ticking a box.

Second, do not wait for supervisors to catch mistakes. You are the person who knows your shot best, and passing sloppy work up the chain only slows the project down. Check your own damn work, and check it carefully. Reviews are not supposed to be the first time someone spots jitter, slipping, or a bad orientation. They should be the final validation. By taking responsibility for your own QC, you save everyone time and avoid unnecessary frustration.

Third, always start with a clear strategy. Diving into a plate and hoping for the best is rarely effective. Think about parallax, about where your reliable features are, and about how you are going to attack the shot before you place the first tracker. Remember that you are not simply tracking dots—you are reconstructing parallax, the three-dimensional shift that gives the solve its accuracy. Studying real-world cameras will help here, because the more you know about optics and filmbacks, the better you will understand what your data is actually doing. Photography is excellent training for this: learning how lenses distort, how focus shifts, and how depth of field behaves will make you a far stronger tracker.

Finally, develop disciplined working habits. Save incremental versions religiously, because one day you will need to roll back. Expect missing metadata and be prepared to improvise without breaking the integrity of the shot. Communicate early and often with your team, because silence kills more projects than bad footage ever will. And always deliver to the end of the chain. That means not just finishing your track but packaging it in a way that makes it immediately usable for the next artist. As we like to say: where others see motion, we see parallax.

Digital Production: Any Final thoughts?

Walid Ben Henda: Matchmove is invisible, until it’s wrong. That’s why we exist: to make sure it’s right, every time. Our focus is on preventing problems before they happen, delivering assets that are ready for comp, and keeping this invisible art as solid as it can be. AI may handle simple shots one day, but the complex ones will always need skilled human trackers.

The post Matching Your Moves: Inside Vision Age VFX first appeared on DIGITAL PRODUCTION and was written by Bela Beier.

Under the direction of Barry Jenkins, MPC explored his unique style while bringing the Pride Lands to life. Combining artistry and technology, the team delivered detailed VFX that capture the essence of the story. Led by Production VFX Supervisor Adam Valdez, Animation Supervisor Daniel Fotheringham, VFX Producer Barry St. John, MPC VFX Supervisor Audrey Ferrara, and VFX Producer Georgie Duncan, more than 1,700 artists, production crew, and technologists at MPC created the visual effects for Mufasa: The Lion King.

A few notes on the scope of the project: More than 1,700 artists, supervisors, and production crew contributed to the making of “Mufasa: The Lion King” at MPC. Storage requirements hit 25 petabytes which is the equivalent storage of 5.6 million DVDs. Rendering the film in final quality took 150M hours. It would take a single computer 17,123 years to complete.

DP: How did you first join the project for Mufasa: The Lion King?
Audrey Ferrara: In summer 2020, right in the middle of the pandemic, we received the first draft of “Mufasa: the Lion King” and I was asked to breakdown the environments part of the script. This was the first point of connection with the project. Later on, I was asked to supervise the post-production visual effects, MPC side, and I accepted. How can you not say yes to this story…… But I have to say, the challenges were big and it was my first time supervising such a massive project. September 2021, after working closely on the pre-production side with Mark Friedberg, the production designer, we headed to LA for the shoot!

DP: What was the size and structure of the team involved in creating the visual effects for Mufasa?Audrey Ferrara: In total, we had a staggering team of 1700 artists who played some part on the project, coming from different sites across the world. We had to structure the project in multiple “departments”, not quite yet in the sense of a traditional company structure…. Think of it as an expert department dedicated to the movie. So we had a department for animation, itself split into 3 units, with 3 anim supervisors (Dan Blacker, El Sulliman, Ruth Bailey), due to the sheer volume of work. We had a sets department, supervised by 2 sets supervisors (Andrea Vincenti and Luca Bonatti), who were in charge of all the 77 environments: modelling, texturing, and also the greens library, used to set dress those sets. We had one character supervisor (Klaus Skovbo) in charge of all the animals (118 unique species), one FX supervisor (William Banti), dealing with the water work, snow, fire, clouds….. One techanim supervisor (Will Fife), in charge of animals and dynamic foliage. And one lighting supervisor (Francois de Villiers), who worked with James Laxton, the cinematographer, to understand his style and his typical light-rigs choices.

When it was time to gather everything in comp, we split all the shots in 3 comp units, with 6 comp supervisors (Luke Bigley, Adam Arnot, Anuj, Wineeth, Sandesh, Sourahb), so once again we could work in parallel, and deal with the volume. Lastly, we had one stereo department, supervised by Nicola Casanova, who set up eye divergence early on, at the layout process.

What is interesting is the structure of the show evolved during the 4 years it lasted, following its needs at certain times, so it changed along the way. We had to be flexible, for the schedule and anything else that we couldn’t foresee. For example, we didn’t start the show with 3 anim units straight away… We started with one, then the second one came along and finally the third one later on.

DP: Can you walk us through the technical pipeline for the movie? What software, tools, or workflows were critical to bringing these photorealistic animals to life?
Audrey Ferrara: On this film we start using Unreal Engine from Epic Games.  Using a game engine software creates several new kinds of workflows.  First, because we started the film during a pandemic, we had to figure out a way to start collaborating on full-cg world building while in different cities (London, New York & LA).  Unreal has a kind of “multi-player” function in its editor that facilitates several computers networking together to view the same content.  And then Unreal’s realtime performance allows for the use of VR headsets, which really helps immerse during the exploration process.

We also used Unreal to do our virtual shooting process.  Once our shoot is completed, we have a rough layout of our movie in this game-engine type state, and we download each take as data into the MPC main pipeline to start creating all the photo-realistic work.  We still use Autodesk Maya as a main workbench.  This is where animators do their job, and where a lot of shot construction happens.  We use Houdini for complex simulations for water and fur movement, as well as the procedural creation of very complex environments phenomena, to finish the 3D work on shots.  All of that then ends up in Renderman.  We have teams at MPC who are always updating material shader definitions based on the latest understandings, and we saw great improvements on the rendering of fur in particular for this show.  Lighting work in Katana for final lighting setups and pipeline.  Finally we use Nuke for the ability to really get into both the subtleties of how a photographic looks, but work on complex ideas like how the sun moves through atmosphere, or how a lion made of clouds changes over time.

All in all its a big chain of software, one for each specialist job.   Within each commercial software, of course there are years of MPC tools that aid in tasks, or create pipelines that flow the data along the chain.  At MPC there are so many pieces of software that complete the whole studio, it’s fairly mind boggling, but it’s a powerful system that sits on top of some very big infrastructural layers which gives us the ability to operate at scale, and globally.

DP: What were the biggest creative challenges on this project, and how did the team overcome them?
Audrey Ferrara: One of the first challenges was to make all the characters relatable and engaging emotionally: from intense emotion, to very subtle reactions. Iin order to offer the animation team all this range, the MPC character LAB supervisor, Klaus Skovbo, pushed the quality of the model to the next level. The models saw a dramatic increase in the detail, represented in geometry as more and more polygons, as well as the hair count on each character. MPC also reworked its fur pipeline from scratch and developing LOMA, a new fur system. For example: Each lion had over 30,000,000 hairs to achieve the realistic look
of fur. Mufasa’s mane alone had 16,995,454 hair curves.

DP: The environments in Mufasa look breathtakingly real. How did the team approach designing these spaces to feel immersive and expansive while staying faithful to the “original” Lion King aesthetic?
Audrey Ferrara: For quite a few years, and now quite a few films, MPC has developed a very complete approach to fully-Cg, 3D world building.  Production design is so important on any film, and here we own the responsibility to completely represent Africa, the designed spaces of the film that support camera, composition and photography, and ultimately bring a feeling and tone to each sequence via the look of the world of the film.  It’s really one of the single biggest jobs on the film next to the acting.   

We always scout and data-gather from real places.  We’d been to Africa before for “The Lion King” but this time the pandemic prevented us from making a trip as a team.  So we sent local scouts to various locations to data gather and collect our references.  This time many parts of Africa donated content for the journey portion of the story.

You can think of the world building in layers.  There is a library of objects like individual plants and trees and rocks.  But for any one location there is also a library of terrain pieces, like modules of the required landscape that allow for complexity near the camera.  The entire landscape is usually a sculpted thing, based on reference.  So you can think of a valley having an overall sculpt, with modular towers of rock that are resued and arranged, and then set dressing with library objects.  Finally, there are ways of combining textures in layers, and procedurally-generated maps for things like weathering and water flow and fine-object scatter distribution, that glue things together.

The team really pushed all of this to the next level for this film.  We had more sets, more space to cover, and more variety than ever.  For each eco-system, there is a whole period of look-development, balancing the shaders of the landscape for different looks.  Snow was a new challenge for this film, and we made an effort to represent the ice crystals and other complexities of icy sheen that builds up from snow that sits for a long time in a cold place.  The team really pushed themselves on the entire film. 

DP: Iterating on models and assets is often the backbone of a project like this. How much influence did the animated classic have, versus footage from “refrence settings” in reality?
Audrey Ferrara: We don’t really look to the animated films to have a big influence on the new films, other than understanding the mythology and legacy of them.  Those films work as animation, and this is supposed to look and feel and relate to the audiences as live-action.  Thus we don’t let things like design influence us too much.  That said, of course we are all personally influenced by those original animated films, and we love and respect them.  We know that they work by giving the audience a heightened sense of storytelling, a magical quality and a sense of beauty and awe, and we aspire to effect people in the same way. So in all our artists’ work, we are very aware of those original films and try to honor them, without copying them. 

DP: How did MPC and the director collaborate creatively on Mufasa? Was there room for experimentation, or was the process more about refining a clear vision from the start?
Audrey Ferrara: All filmmaking involves discovery along the way. Our script, and our director Barry Jenkins, have a strong starting point of view about what the movie is about, the feeling to create, and a kind of mode of storytelling.  Barry has more than a style, but a kind of voice, a way of telling you a story from a very intimate place close to the characters.  Sometimes the camera literally swirls close to their faces, as if you put the audience as close as possible to them. 

Barry was doing an animated film for the first time so we tried really hard to make the whole process flexible for him.  It’s not easy because it’s a huge machine and it takes so many months to go from the idea of a shot to finally seeing a shot become real.  But at the many steps of the process, you see more and more, and by mid-way through the making of the film, Barry was really an expert on the process and could tell if things were going to work in the long-run.  

Disney as a studio provides the schedule and budget to adapt to change over time.  They know the movie will slowly emerge, and that all of us won’t know precisely if it works or what might need improving, until midway.  So we are fortunate to have that kind of support and space to improve as we go.

DP: What role did concept art play in the pre-production phase? Did any early designs significantly evolve during production?
Audrey Ferrara: The pre-production phase was crucial in exploring the new environments needed for the story, and how extensive the journey would be for the heroes. Mark Friedberg’s art department did all the concept art for those environments and then proceeded to build them in unreal, in order to allow the director to scout them in VR. The collaboration between art department and MPC was important, so they could use the assets already built on the first movie, and be a starting point for the rest of the development phase. The art department also established the looks of the characters, almost building the genealogy tree of the lions, so they could define family features. 

The construction of the sets in unreal provided a high quality previz, that resolved almost every creative challenge early one on the show. In some sequences, the final shots are very close to the intentions in the previz, which shows how well thought the design and previz were.

DP: With your experience on The Lion King and the Jungle Book, what changed in the pipeline since the 2019-film?
Audrey Ferrara: I think when we started Jungle Book, the question was “can we do this today?”  Now, this kind of filmmaking is expected.  Films like Jungle Book open up how we think about making films, exploring more and more complex worlds to build, how photography and computer graphics mix, and being ambitious.  The pipeline in its basic way is roughly the same.  We make 3D content, we render it, and we comp it.  

I think world building has probably had the biggest evolution.  Many of those departments in the industry were called “3D DMP” a long time ago, because it was digital matte painting which was the main tool.  That evolved into projections and mixes of 3D and 2D in clever ways.  For a film like this, we now are like a hybrid between a game company and a vfx company in this area, because the approach to world building is so intensive and mostly 3D.  

Also, it should be said that MPC has to do a lot of things internally to be ready and manage shows of this size and complexity.  There are a lot of standards and pipelines and inter-departmental things to help orchestrate it all.    

DP: Virtual production is becoming a game-changer. Did you use Unreal Engine or a similar tool for virtual sets, and how did this impact your workflow compared to traditional methods?
Audrey Ferrara: Mark Friedberg and his art department built all the sets in Unreal, so the filmmakers could do the scouts of sets in VR. After a set was approved, james Laxton, the cinematographer would spend time setting up his lights in the scenes, getting ready for the shoot. Unreal offered an incredible flexibility for the filmmakers, allowing them to walk the environments, art direct live the spaces or the action of the lions, thanks to the quadcap tool developed by MPC, or to change the lighting live if necessery. 
Any kind of reshoot was also easy to do, since we just had to bring up the desired scene in unreal, and shoot whatever was required.

DP: Expectations for photorealism are higher than ever. How did MPC approach creating assets in 2024, and what new techniques were implemented to make them even more believable?
Audrey Ferrara: Yeah, films like this have many kinds of assets, from props and ground-scatter items, 3D set components with multi-layered texture shaders that are very procedural, to the intensity that is a 3D character who can act and interact with its environment.  They all have to be unified by a quality and visual standard or they don’t render together, and that’s a huge challenge.  The teams at MPC attack the problem with a bunch of different techniques, but in the end it has to be art directed and CG-supervised by top talent who understand what ‘good enough’ is for these films, and have the artistry and techniques to back that up.  

For Mufasa, MPC developed a new fur pipeline called LOMA. It allowed the team to have more fur, better control of the curves of the hair, and more precise simulations. The different weather conditions throughout the movie resulted in new challenges for the techanim team: drenched fur, snow covered fur, close ups of lions fighting with their manes interacting with each other etc….. 

Mufasa has 600,000 hairs on his ears, 6.2 million hairs on his legs, and 9 million hairs covering the middle
portion of his body.

DP: The lighting and atmosphere in Mufasa are stunningly diverse across scenes. What was your approach to crafting the look and feel of the movie, and how much of this was driven by technical tools versus artistic intuition?
Audrey Ferrara: James Laxton, the DoP on the movie, spent long hours defining the lighting in every scene for the shoot, aided by the onset lighting supervisor Sam Maniscalco. He chose carefully each single sky, from an HDRI library provided by MPC, and set up all his lights. In parallel, MPC’s lighting supervisor, Francois De Villiers, explored test shots, in order to get familiar with James’ lens choices and style. This happened way before we received the sfirst turnover shots. It allowed us to do our homework with James, and be ready with the shots were coming our way. The continuity between pre-prod, shoot and post-prod really helped keeping the lighting team on brief throughout the production. 

One amazing thing about James’ style of lighting is that he very rarely “tweak” or moves around his set up. Which is an amazing advantage in post-production, when you have a consistent lighting throughout an entire scene. It really helped the MPC team to keep lightrigs clean and consistent .

DP: Motion capture was usedin creating the lifelike movements of Mufasa’s characters. How did you integrate mocap data into the pipeline to achieve something so natural-looking yet entirely digital? Audrey Ferrara: Motion capture was used for our “quadcap” new tool for the shoot: with this tool, the lion front legs would “snap” onto human legs, so Daniel Fotheringham, our client animation supervisor could act the lion part on stage. The rear end of the animal would be simulated to follow the front, and snap onto the ground. This was the part where motion capture helped for the shoot and blocking the action. But once the scenes were turned over to MPC, the animation team made each single frame of the movie by hand. Everything has been hand animated on the movie, after carefully studying and selecting references for the animals.

DP: You worked on “The lion King”, “Jungle Book” and now Mufasa: What is your asset library like, and how did you approach Mufasa versus something like Passengers or Alien:Covenant?
Audrey Ferrara: I would say the biggest difference between projects like “Mufasa” or “Jungle Book” and “Alien:covenant” or “Passengers” is the volume of work and time really.

Both types of shows can be extremely complex to achieve, but on a movie like “Mufasa”, you have to keep going for a long time. It’s a big marathon, with a lot of aspects to manage. Everything is bigger: the amount of time, the amount of work, the amount of people working on the show. And it can be daunting sometimes, to look at the mountain of work in front of you that you’re gonna have to climb, a way or another. 

I felt so overwhelmed with the work on “Jungle Book” that I had to breakdown everything, and take one task at the time, to make it manageable.

So, it is my first advice when a junior starts on a show like this: one step at the time, but keep the bigger picture in the back of your head, so you know where you’re going. And that they’re not alone in it. That’s what is great on big shows: there are a lot of people working on it, and you can always find someone that has the answers you need and help you. 

DP: Do you have a favorite shot or sequence in the movie?
Audrey Ferrara: My favorite shot ever in the movie is the one where Kiros chases Mufasa in the caves at the end: the shot switches to slow-motion, before resuming to normal speed. This shot is SO cool: the lighting, the animation, the camera, everything is great in this shot. Will definitely add it to my demoreel!

DP: If you could go back to the beginning of the project, what advice would you give yourself?
Audrey Ferrara: “Mufasa” was my first time as a VFX supervisor, on a project of that scale. I would tell myself to trust my instincts and the process, to avoid unnecessary stress. 

DP: Without breaking any NDAs, what’s next for you? Are there any new challenges or areas of exploration that you’re excited about?
Audrey Ferrara: Over time I have become more and more driven by design.  On this film I was lucky to participate fully in that phase, and I hope to do more and more of it. 

The post Mufasa heads to MPC first appeared on DIGITAL PRODUCTION and was written by Bela Beier.

Activision has released the Call of Duty: Warzone Caldera Dataset, a collection of assets drawn from the popular game’s expansive map. This release marks an interesting resource for professionals working in real-time graphics, game development, and VFX pipelines, offering opportunities to integrate production-grade data into various workflows.

The dataset includes geometry, textures, and environmental data extracted from the Caldera map, one of the largest environments featured in the Call of Duty franchise. Activision emphasizes its focus on high-fidelity assets, making the dataset particularly suitable for simulation, cinematic pre-visualization, and research into real-time rendering technologies.

Real-Time Graphics and VFX Applications

Real-time professionals can leverage the dataset to test lighting algorithms, simulate environmental effects, or optimize physics engines. VFX teams, particularly those exploring video game cinematics, could use the data to enhance their scene composition or animate sequences inspired by the source material.

Game Development and Educational Use

Game developers might find the assets useful for prototyping or building test levels, while educational institutions could incorporate the dataset into curricula aimed at teaching real-time graphics and engine development.

Pricing and Availability

Activision offers the dataset as a free download for non-commercial use. Interested developers and researchers can access it directly from the official Activision blog here.

The post Activision Releases Call of Duty: Warzone Caldera Dataset for Real-Time Applications first appeared on DIGITAL PRODUCTION and was written by Bela Beier.