For those who don’t know the tool: BlendShape Monitor is a lightweight diagnostic plugin for Autodesk Maya that sits squarely in character rigging and facial setup, focusing only on inspection rather than deformation authoring, and does not overlap with Autodesk’s own rigging tools or third-party rig builders.

Why blendshapes still go wrong

Blendshapes remain one of the most common deformation methods for facial animation in Maya, particularly for FACS-based rigs and corrective shapes layered on top of joint systems. Despite their ubiquity, debugging blendshapes once a rig grows beyond a few dozen targets remains largely manual and error-prone. Maya’s native interface presents blendshape nodes as long, alphabetical lists of target names with numeric weights. That abstraction works when rigs are small and well documented. It breaks down quickly when shapes overlap, are reused across multiple regions, or are indirectly driven by other systems, such as RBF solvers.

This problem is compounded in production environments where rigs are inherited, shared, or modified over time. In many cases, artists are asked to fix deformation issues without knowing which blendshape is responsible, or whether multiple shapes are contributing simultaneously. The result is a familiar cycle of muting targets, scrubbing weights, and guessing.

What BlendShape Monitor actually does

BlendShape Monitor is a diagnostic plugin for Autodesk Maya that attempts to replace that guesswork with direct visual feedback. Instead of relying on target names and numerical weights, the tool visualises the influence of individual blendshapes directly on the mesh using colour-coded vertex heatmaps. These heatmaps update in real time, including during animation playback, reflecting the current evaluated weight of each shape.

The plugin reads the deformation data from existing blendShape nodes. It does not modify the rig, create new targets, or alter evaluation order. According to the developer, its sole purpose is inspection. This distinction matters, as the tool is intended to be safe to use on production rigs without changing scene data.

The visualisation highlights which vertices are affected by a given blendshape and to what extent. Areas with stronger deformation are shown in higher-intensity colours, making it immediately obvious whether a shape is localised, overlaps with others, or extends into unintended regions of the mesh.

Managing clutter in dense rigs

One of the stated goals of BlendShape Monitor is to make large blendshape sets manageable. The plugin includes filtering options that hide inactive targets, allowing artists to focus only on shapes that currently contribute to the deformation. This is particularly relevant when blendshapes are driven indirectly by other systems, where weights may be non-zero even if no animator is directly adjusting them.

The tool also allows individual targets to be solo-ed. When a shape is soloed, other blendshapes are temporarily hidden from the visualisation, making it easier to inspect its isolated effect. Global visibility toggles let you enable or disable the heatmap overlay without removing the plugin from the scene.

Weight values are synced live with Maya’s evaluation during playback. This means the visualisation reflects the rig’s actual state at each frame, rather than a static snapshot. For troubleshooting animation issues that only appear in motion, this real-time aspect is central to the tool’s design.

Origin in production reality

BlendShape Monitor was developed by Johnson Lee, an artist and tools developer whose work as Art Director at Reallusion directly informed the tool’s focus on large-scale facial rigs. According to the developer, the plugin emerged from internal production needs rather than as a speculative product idea.

At Reallusion, facial rigs for characters such as those used in Character Creator workflows can exceed one hundred blendshapes. These include expression shapes based on FACS conventions, as well as numerous corrective shapes used to address secondary deformation. In such setups, identifying which shape is responsible for an artefact using Maya’s default UI is time-consuming and unreliable. The plugin reflects that context. It is narrowly scoped, avoids adding new rigging concepts, and addresses a specific bottleneck encountered when maintaining and debugging existing rigs rather than building new ones.

Compatibility and scope

BlendShape Monitor is compatible with Autodesk Maya 2022 and later. No support is claimed for earlier versions. The plugin runs inside Maya and does not require external dependencies according to the product listing. Installation and licensing are handled via Gumroad.

The tool does not claim to support other DCC applications, nor does it attempt to abstract blendshape concepts across platforms. It is explicitly Maya-specific, relying on Maya’s native blendShape node behaviour. It is also not positioned as a teaching tool. Users are expected to understand blendshape workflows, vertex-level deformation, and Maya’s rig evaluation. The plugin provides visibility, not validation.

Pricing and licensing

BlendShape Monitor is sold via Gumroad. At the time of writing, pricing is listed at approximately USD 20 for a freelance licence and USD 70 for a studio seat. The exact terms of these licences are defined on the Gumroad page and should be reviewed before purchase. No subscription model is indicated.

What it does not solve

While the plugin makes deformation issues easier to see, it does not resolve them automatically. Poor topology, conflicting targets, and incorrectly authored shapes still need to be fixed at the source, by you. BlendShape Monitor does not rank shapes by quality, detect errors, or suggest corrections.

It also does not address performance issues caused by excessive blendshape counts or inefficient evaluation. Its visual overlays are for inspection, not optimisation. Artists should be cautious when using any viewport overlay in heavy scenes and test performance impact in their own environments.

Production considerations

As with any new tool, BlendShape Monitor should be evaluated under controlled conditions before being deployed to production. Its read-only approach reduces risk, but pipeline teams should still test compatibility with existing rigs, scripts, and viewport configurations.

In teh final analysis, BlendShape Monitor addresses a narrow but persistent pain point in Maya character rigging, offering visibility where the host application still relies heavily on lists and numbers. Whether it becomes a standard part of rig debugging workflows will depend on how well it holds up wiht the messier rigs found in long-running productions.

// BlendShape Monitor Gumroad product page
// https://johnsonlee529.gumroad.com/l/bs-monitor-maya/

The post BlendShape Monitor puts Maya rigs under a heat lamp first appeared on DIGITAL PRODUCTION and was written by Bela Beier.

Reallusion has released iClone 8.6, a point update that focuses on deeper integration with Character Creator 5, introduces a new generation of shaders for digital human eyes, and delivers extended export options for Unreal Engine 5. The release also refines facial animation, improves rendering performance, and resolves a series of stability issues that had been slowing down production workflows.

CC5 Compatibility

The most noticeable addition in iClone 8.6 is its extended support for Character Creator 5. CC5 HD characters and facial profiles are now natively supported inside iClone, while MetaHuman facial animations can be imported directly if CC5 is installed. Compatibility has also been broadened to include CC3+ Extended-Plus corrective shapes, which means that facial rigs can now carry far more nuance into iClone without requiring additional workarounds. The result is a smoother connection between character creation and performance animation, making the two tools feel more like components of the same pipeline.

Eye Shaders

Reallusion has reworked one of the most critical details in digital humans: the eyes. iClone 8.6 introduces two new shader types designed to remove long-standing artifacts. The Digital Human_Occlusion shader closes the gap between eyeball and socket, preventing hard intersections that tended to break realism under close-up lighting. Alongside it, the Digital Human_Tearline shader brings a corrected UV layout and sharper specular highlights, which makes tear lines react more naturally to illumination. These changes elevate the look of close-up shots and address one of the more distracting weaknesses in earlier versions.

Export and Animation

Pipeline exports have been updated to match current production environments. New FBX presets are included for Marmoset Toolbag as well as multiple Unreal Engine 5 targets, with UE5 Skeletons, UE5 Mannequins, and UEFN all receiving their own dedicated profiles. The existing Unreal preset has been renamed to UE4 Skeleton to avoid confusion. This makes it easier to match exported assets to the correct rig on import, reducing the time artists spend reconfiguring skeletons or poses.

Animation work also benefits from fresh controls. A new Facial Parts Strength option in the timeline allows animators to amplify or reduce expression intensity on specific regions of the face, rather than adjusting entire performance tracks. Subdivision controls for CC5 characters have been introduced, offering finer mesh smoothing for close-up animation. Motion LIVE now supports non-linear curves in facial animation, giving higher fidelity to CC5 HD and CC3+ Extended-Plus characters when recording live input.

Rendering and Workflow

Rendering improvements in iClone 8.6 may not be headline-grabbing, but they address key pain points in everyday use. A dynamic mipmap option has been added to real-time rendering, reducing noise when working with detailed wrinkle maps. Project files are also lighter thanks to mesh and animation compression, though iProject containers themselves remain uncompressed. Together these changes improve responsiveness without altering the core look of the renders.

Fixes and Stability

The release also closes several long-standing bugs. FBX export from iClone now correctly aligns controllers for UE5.6 control rigs, while crashes linked to specific imports have been eliminated. Motion editing functions such as Edit Motion Layer and Motion Direction Control have been stabilized, and the issue of characters growing in size when flattening animation layers has been resolved. Rendering has also been corrected: AVI exports are once again properly compressed, and DLSS no longer introduces mosaic artifacts.

What’s Next

Reallusion has confirmed that a compatible AI Render plug-in for both iClone 8.6 and CC5 will be released on September 1, 2025.

The post iClone 8.6 Update Adds New Eye Shaders, and UE5 Export Presets first appeared on DIGITAL PRODUCTION and was written by Bela Beier.

If you’ve ever watched Pixar’s Win or Lose and envied those freely migrating noses and mouths, Krysidian’s SMIRK delivers exactly that—inside Blender, and without the need for topological acrobatics. SMIRK is a free toolkit for creating, animating, and re-positioning facial features as if the mesh never heard of vertex groups. It’s Geometry Nodes magic for the chronically restless.

What’s Under the Hood? Transparency Boolean

SMIRK doesn’t do blend shapes or bone juggling. The core of the toolkit is a Transparency Boolean node setup, which carves facial features (mouth, eyes, eyebrows) out of the mesh, independent of topology. The system leans on Blender’s Grease Pencil: draw your shapes, animate the curves, and SMIRK dynamically cuts and reveals the new features in real time. No head sculpting, no careful edge flow—just uncommitted facial parts ready for cartoon chaos.

#RankingOfKings #fanart, Using the method of #pixar in his new series #WinOrLose. #rig #rigging #blender #animation pic.twitter.com/BRZ9Q7hwPZ

— Edward Ureña (@EdwardUrena_h) March 5, 2025

A Nod to Pixar and Edward Ureña

The SMIRK method borrows its philosophy from Pixar’s stylized face rigs seen in Win or Lose. Krysidian credits Edward Ureña’s public Blender rigs as the catalyst for bringing detached facial animation to Geometry Nodes. The result is a system that trades painstaking facial topology for instant, playful feature mobility—ideal for exaggerated, cartoon-style characters.

Installs Faster Than It Animates

Getting started requires little more than downloading a .blend from Gumroad, appending the SMIRK node group, and drawing directly onto your character mesh with Grease Pencil. No plug-in registration or Python trickery—just pure node-based workflow inside Blender. The toolkit is free, with no paywall or licensing, and immediately available for any project that can tolerate a little facial wanderlust.

Use Cases and Limitations

SMIRK is built for stylized rigs. If you need mouth corners to zip from jaw to brow or want to animate a nose dancing across the forehead, SMIRK will oblige—without ever consulting your retopology. It’s a playful answer to a niche but growing demand: extreme facial flexibility in real-time or rendered animation. But as the developer points out, this is not a production-locked, bug-free solution. It comes “as is,” with no support or studio certification. Artists should test thoroughly in their own production environment before committing to client or deadline-sensitive work.

Professional Reality Check

No matter how fun SMIRK looks, new workflows deserve skepticism. Geometry Node setups can have compatibility quirks, and the Transparency Boolean method may introduce unexpected shading or render artifacts, especially with complex lighting or material setups. As always, test before deploying in a high-stakes pipeline.

SMIRK

The post Pixar Faces, Blender Spaces: SMIRK Lets Features Wander Free first appeared on DIGITAL PRODUCTION and was written by Bela Beier.