Mari 8.0 and Nuke 17.1: Betas That Target the Pain Points of Modern VFX
Foundry’s Mari 8.0 beta release and Nuke 17.1 compositing beta mark a focused push to speed up 3D texture painting tools and high-end compositing by tightening workflows around node-based look development, 3D Gaussian splats, and modern USD/Hydra pipelines, aiming squarely at professional VFX software updates that reduce iteration time rather than add surface-level features. These betas matter because they tackle the bottlenecks artists complain about most: slow lookdev iteration, clumsy comparisons, and fragile 3D handoffs between departments. Instead of flashy AI demos, Foundry is quietly aligning Mari and Nuke with how studios now shoot and postproduce: dense 3D scans, USD-centric scene structures, and GPU-aware viewers. By putting these changes into open beta, the company is inviting production feedback before these tools harden into pipeline standards, a smart move given how disruptive splat-based environments and USD workflows can be if they are designed in isolation.
Mari 8.0: Hex Tiling, Color Remap, and A/B Nodes Aim at Faster Lookdev
Mari 8.0 is a clear statement that texture painting should feel more like technical look development and less like manual cleanup. The new Hex Tile node minimizes visible repeats by blending textures in a hexagonal grid instead of a rectangular one, aligning Mari with other key VFX tools like Houdini, Redshift and RenderMan that now support hex tiling. That directly targets the tedious work of hiding tiling artifacts on large assets. Equally important, the Compare node adds A/B wipe comparisons between different parts of the node graph, making it easier to iterate on materials and variants without exporting test renders. Color Remap centralizes hue, saturation, exposure and gamma tweaks into a single node so artists can grade a material in one place rather than chasing scattered adjustments. Add over 49 new math nodes for conditional logic and advanced randomization, and Mari’s Node Graph starts to feel more like a compact shader lab than a paint program, which is precisely what high-end lookdev teams have wanted.
Workflow and Platform Moves: Mari’s Subtle but Strategic Quality-of-Life Upgrades
Beyond headline nodes, Mari 8.0’s workflow tweaks show Foundry finally treating navigation and project management as pipeline concerns, not mere UX details. Trackball rotation-style navigation offers an alternative to the standard Orbit tool, which may sound minor but can dramatically change how fast artists frame and evaluate textures on complex assets. Clear visual indicators for archived projects in the library reduce the risk of junior artists reviving legacy scenes in live productions. Under the hood, the move to the CY2025 specification for the VFX Reference Platform is overdue but welcome: Mari 7.5, the current stable release, is now four platform versions behind, so this update is at least a step toward modernizing its base stack. In pricing terms, Mari is rental-only, with individual subscriptions at USD 86 (approx. RM395) per month or USD 689 (approx. RM3,165) per year, and team subscriptions at USD 1,289 (approx. RM5,920) per year, up USD 60 (approx. RM275) since Mari 7.5. That increase raises the bar for studios: the new nodes and workflow gains must earn their keep in fewer lookdev hours.
Nuke 17.1: From Static Splat Scans to Relighted, USD-Savvy 3D Scenes
Nuke 17.1 extends 3D Gaussian Splatting beyond static environment scans into something more like a first-class 3D element in the compositing pipeline. Where Nuke 17.0 let artists import, manipulate and render 3DGS data for real-world environments or set extensions, the new GeoSequencer node can now process sequences of USD, PLY or SPLAT files into time-sampled USDC, turning animated splats into manageable scene assets. The updated SplatRender node supports 2D rendering of Direct, Point and Spot lights placed in Nuke’s 3D system, enabling basic relighting of Gaussian splats. That is a practical step toward lighting control on scan-heavy shows, even if it is not full physically based shading. On the USD side, non-destructive export workflows allow explicit sublayers and overs for sparse overrides, cutting down the need to flatten scenes and improving round-tripping between departments or studios. GeoExport now defaults to exporting frames individually and stitching them afterward, reducing memory usage on heavy sequences, which is exactly the kind of quiet performance win compositors feel every day.
Hydra 2.0, Timeline Comparisons, and What These Betas Signal for VFX Pipelines
The update of Nuke’s Hydra-based 3D viewer to Hydra 2.0 is more strategic than it looks at first glance. It enables new options when using the current HdStorm render delegate, including previewing AOVs such as depth and position inside the viewer, and it explicitly "paves the way" for third-party render delegates later. That opens the door to GPU-accelerated, renderer-specific 3D previews inside Nuke 17.1 compositing workflows. On the timeline side, Nuke Studio gains a shapes system for its Annotations toolset, letting artists mark frames with rectangles and arrows, plus better Contact Sheet management per sequence. A new Compare Versions feature delivers A/B comparisons without leaving the timeline, mirroring the Viewer’s comparison modes and helping supervisors make decisions faster. Pricing remains firmly in high-end territory: annual subscriptions are USD 3,839 (approx. RM17,630) for Nuke, USD 5,219 (approx. RM23,960) for NukeX, and USD 6,379 (approx. RM29,280) for Nuke Studio, with Nuke Render at USD 462 (approx. RM2,120) and Nuke Indie at USD 499 (approx. RM2,290) per year. Both Mari 8.0 and Nuke 17.1 are in open beta with expected stable releases in three to six months, giving studios a narrow but valuable window to pressure-test these changes before they become production defaults.
