For those who don’t know the tool: Redshift is a production renderer that plugs into DCCs like Cinema 4D, Autodesk Maya, SideFX Houdini, Autodesk 3ds Max, Foundry Katana, and Blender, so lighting TDs can argue about samples in the app they already love.

The night shift arrives everywhere

So, after sorting through the heap, it is time to talk about Redshift, specifically. Redshift 2026.5.0 adds a night sky representation to the Sun and Sky system across hosts, including stars, moon, and Milky Way, with time and location controls.

That matters less for hero closeups and more for the boring shots that still need to ship: wide establishing frames, background domes, and lighting continuity on sequences that bounce between day and night. And also, to keep the internet sleuths ansd “Actually-s…” at bay.

The same update also includes a fix for the Physical Sun and Sky light where the milky way could render in flipped coordinates. It is the kind of bug that makes you doubt your compass and your life choices at the same time.

OpenPBR takes the default seat in Cinema 4D

For Cinema 4D, OpenPBR becomes the default material, and the default material list in preferences gets updated. The shader model itself comes from OpenPBR, the OpenPBR Surface specification hosted under the Academy Software Foundation umbrella.

In practical terms, the change means new scenes in that host start from OpenPBR unless a user switches the preference. It also comes with a cleaned up material menu that puts OpenPBR up front and pushes older options out of the spotlight.

If your pipeline has older material libraries, this is the moment to check how presets migrate, how template scenes behave, and whether old lookdev assumptions still hold. New tools and innovations should always be tested before use in production.

Texture Displacement gets smarter and less picky

Texture Displacement continues to expand, building on the system introduced earlier as an alternative to vertex displacement. In 2026.5.0, texture-based displacement gains preliminary UDIM and UVTILE support. It also gets mirror repeat mode when smart baking is enabled, plus improved quality when displacement is driven by UV projections.

There is also enhanced auto bump mapping support meant to improve bump mapping quality when used with displacement nodes. This addresses a common pain point where displacement and bump interplay can get noisy or fight each other in shading.

A set of fixes targets stability and interactivity too. The release notes list multiple cases where texture displacement could disappear or show artefacts in interactive preview rendering when using bucket mode, and those issues have received fixes.

One more fix calls out self-shadowing artefacts on objects with extremely high polygon angles when using Texture Displacement. If you have ever seen displacement turn into a crunchy self-shadowed mess on near-planar geometry, that line will feel oddly personal.

Some Cinema 4D specific fixes also land here, including a bug where UDIMs were not correctly rendered when using Team Render, plus a bug that could cause random black images when rendering heavy meshes with deformation blur enabled. Redshift 2026.5.0 also sets a minimum plugin requirement of Cinema 4D 2026.2.

In the middle of all that, one of the most important workflow wins is banal: fewer special cases. The more often displacement works the same way across UV layouts, projections, and preview modes, the less time artists spend building around the renderer instead of using it. Also, please rmember that displacement tends to be the first place where a scene crosses from fine to on fire, so keep an eye on memory and subdiv settings when you roll this out.

AMD ray tracing via HIP RT, marked experimental

Redshift 2026.5.0 adds experimental support for HIP Hardware Ray Tracing, also known as HIP RT. HIP RT itself is an AMD HIP RT ray tracing library designed for applications using HIP.

The release notes call the feature experimental. That is a polite way of saying you should not bet a deadline on it without testing it on your actual scenes, drivers, and farm images.

The same build also mentions improved render startup performance in scenes with high primitive counts when using Hardware Ray Tracing. Together, those notes make a clear point: hardware ray tracing paths continue to expand, and more of the heavy lifting is expected to land in that codepath over time.

Windows on Arm joins the CPU party

Redshift 2026.5.0 adds native Windows Arm64 support for Redshift CPU in Cinema 4D, ZBrush, and via the command line renderer. That is specifically CPU rendering, not GPU rendering. For the broader platform context, Windows on Arm covers Arm64 devices running Windows, including systems where native Arm apps run without emulation while other binaries may run under emulation.

If you are eyeing Arm based Windows laptops for travel kits or lightweight previs work, this change removes at least one hard block for CPU rendering workflows in those supported routes. It does not magically turn an ultraportable into a render node, but it does make it a more honest option for tests, turntables, and emergency output.

Bottom line for production

None of that removes the usual responsibility from leads and TDs. Update one step at a time, validate looks against show frames, and test your farm and plugin versions before you flip defaults for everyone. That advice feels boring, right up until it saves your weekend.

https://www.maxon.net/en/redshift

The post Maxon ships Redshift 2026.5 first appeared on DIGITAL PRODUCTION and was written by Bela Beier.

For those who don’t know the tool: MoonRay is DreamWorks Animation’s in-house Monte Carlo path tracer, open-sourced in 2023 under Apache 2.0. It drives DreamWorks’ film rendering pipeline and supports hybrid CPU/GPU (XPU) modes. MoonRay integrates with DCCs via its Hydra delegate (hdMoonRay) for tools such as Houdini and Katana. The project includes the Arras framework for distributed rendering and collaboration across clusters.

Light Path Visualiser Enters the Scene

The new light path visualiser in moonray_gui is the most visible change in version 2.34.0.1. It lets artists trace how rays travel through a scene, showing how light interacts with surfaces, volumes, and materials. The visualisation is colour-coded by ray type and can reveal sampling or shading inconsistencies that are otherwise invisible. It also serves as a diagnostic tool for debugging light source or material behaviour.

Platform Updates: macOS Tahoe Joins, Linux Builds Simplified

MoonRay can now build natively on macOS Tahoe, with Sonoma and Sequoia still supported. The renderer also benefits from dependency updates and inculde fixes, easing compilation on a broader range of Linux distributions. The release aligns with both VFX Reference Platform 2024 and 2025, ensuring compatibility with current production libraries and compilers.

New Features Across the Pipeline

Version 2.34.0.1 expands flexibility for lighting and shading setups. Per-lobe lightsets are now available in both vector and XPU modes, allowing finer control of multi-lobe BSDFs. Image maps gain bindable “scale” and “offset” attributes, and SceneVariables documentation for image sizing has been clarified. Performance-wise, MoonRay now shares CPU and memory affinity data between processes using shared memory, improving parallel job efficiency. Timing measurements have switched to a more efficient RDTSC-based method for NUMA [non-uniform memory access] systems, offering more stable profiling results. Artists can also now select lights and cameras for object-space transformations directly in map shaders, simplifying custom shading graph workflows.

Fixes: From Subsurface Scatter to Fisheye Crashes

A long list of bug fixes makes this release notably production-friendly. Key repairs include:

• Fixed a mesh tessellation issue in tiled renders with the FisheyeCamera.
• Resolved DwaTwoSidedMaterial artefacts caused by previous lobe lightset work.
• Corrected crashes when exceeding BSDF lobe limits or loading oversized MeshLights, which now trigger warnings instead.
• Addressed a memory usage spike in vector and XPU modes.
• Fixed missing lightset assignments in vector subsurface scattering and adjusted over-bright subsurface shading on creases.
• Corrected normal direction errors in scalar mode and improved TransformSpaceMap behaviour.
• Fixed adaptive light sampling, volume bounce contributions, and a crash in OpenImageIO 2.4.8 statistics when no images were opened.

Under the Hood: Clean-Up and Refactoring

DreamWorks also restructured several internal libraries. The noise library moved from moonshine/lib/common to moonray/lib/common, simplifying dependencies. The integrator logic for Cryptomatte was reduced in complexity, and redundant intersection path types were removed. Code clean-up continues across lib/rendering/shading, improving maintainability without changing behaviour.

Availability

MoonRay 2.34.0.1 is available now as open-source via the official repository. It can be compiled for Linux and macOS (Tahoe, Sonoma, and Sequoia). GPU rendering requires NVIDIA CUDA/OptiX or Apple Silicon.

Caution Before Production

As with any open-source renderer update, production users should validate the build and test scene compatibility before deployment in live pipelines.

The post Rays Behaving Badly No More: MoonRay 2.34.0.1 Fixes the Path first appeared on DIGITAL PRODUCTION and was written by Bela Beier.

Hair shading improvements

The Standard Hair shader now supports a new accurate mode which simulates each hair strand as a cylinder rather than a flat surface. This produces better close-up renders but increases render times by around 10 to 30 percent. A further adaptive mode has been added, switching automatically to the accurate mode for close-ups and reverting to the approximate mode for distant hairs. The result is a more efficient balance between image quality and performance.

New bloom modes in Lens Effects

Arnold 7.4.4 expands the Lens Effects imager with new bloom-generation modes. A notable addition is the aperture bloom mode, which emulates physical-camera bloom by allowing control over aperture-blade count and curvature. This provides more natural-looking highlights and camera-style glare for cinematic imagery.

Material & lighting enhancements

The thin-film component of the OpenPBR material now observes better energy conservation for iridescent surfaces. Lighting receives additional flexibility: mesh lights can now optionally cast shadows, while light-blocker geometry can include ramp or negative density values, offering more nuanced lighting control for compositors and look-development artists.

Performance gains

Global Light Sampling (GLS) has been refined. The algorithm now factors in the spread parameter of quad and disk lights, which helps reduce noise in low-spread scenes. Autodesk reports performance improvements of up to six times faster in specific lighting scenarios, though more typical gains appear to average around 1.4 times. Texture handling has also been improved. Arnold now auto-generates missing TX files more efficiently, significantly reducing time to first pixel, especially on Windows systems with many CPU cores.

GPU rendering & denoising updates

On the GPU side, Arnold adds support for the Motion Vector shader and improves consistency between CPU and GPU parameter overrides. GPU memory usage has been reduced when rendering instances. The OIDN denoiser now supports NVIDIA Blackwell GPUs and AMD RDNA 4 GPUs, expanding compatibility for new hardware.

Integration plugin support & system requirements

The release is supported by updated host-application plugins, including MAXtoA 5.8.4 for 3ds Max, C4DtoA 4.8.5 for Cinema 4D, HtoA 6.4.5 for Houdini, KtoA 4.4.4 for Katana, and MtoA 5.5.5 for Maya. Arnold 7.4.4 runs on Windows 10 and later, Linux with glibc 2.28 or higher, and macOS 11.0 or later. GPU rendering remains available only on Windows and Linux, requiring a compatible NVIDIA GPU. A single-user subscription is priced at 55 USD per month or 430 USD per year.

Caveats & things to test

While the update looks solid, a few points warrant testing before adoption. Autodesk’s claimed sixfold speed improvement in light sampling likely represents an optimal case; most users can expect more modest gains. The accurate hair mode, while visually impressive, carries a measurable render-time penalty. GPU rendering remains unavailable on macOS in this release. As always, studios should test all new features in their own pipeline before committing to production use. Users of previous Arnold 7.4.x builds will find the 7.4.4 upgrade worthwhile for its stability and refinement. Before installation, back up your projects and test key scenes with the new hair mode, bloom settings and light sampling. Evaluate the render-time impact to confirm benefits in your own workflow.

Arnold 7.4.4 focuses on polish rather than reinvention. The improved hair shader, refined lens bloom and smarter light sampling deliver genuine quality gains without disrupting established pipelines. GPU and denoiser updates further modernise hardware support.

The post Arnold 7.4.4: better hair, bloom & light-sampling first appeared on DIGITAL PRODUCTION and was written by Bela Beier.

Autodesk has rolled out Arnold 7.4.3 and this time, the venerable renderer brings a new toy for the node-happy: the Inference imager. This feature lets users apply image-to-image machine learning models directly inside Arnold via ONNX, transforming frames with neural network-based stylization. If you want your render to look like a Picasso, at least for a single frame, you can now do it natively—though Autodesk admits this won’t guarantee temporal coherence or semantic stability when you try it on sequences. Translation: your animation could look like a different painting in every frame. Production artists, you have been warned.

Arnold GPU: Now With 3.3× More Hurry-Up

For anyone actually using Arnold’s GPU renderer in production, volume rendering on GPU is now up to 3.3× faster. This applies to both OpenVDB volumes and procedural volumes, so your clouds, smoke, and fog will eat fewer pizza breaks. Global Light Sampling is also up to 2.5× faster, and glossy materials on GPU behave a bit more like their slow CPU cousins. These are speed improvements, not new rendering techniques—so if your GPU is slow, it’s still slow, just less embarrassingly so.

Material World: OpenPBR, USD, and All the Acronyms

Arnold 7.4.3 improves OpenPBR-based subsurface scattering, especially for thin-walled materials, and brings the metal lobe in line with energy conservation principles. If you’re deep in the weeds of USD-based pipelines, there are more USD workflow tweaks and improved HTML render reporting for pipeline happiness (or at least, slightly fewer pipeline headaches).

Host App Plug-ins: Most Are Updated (Sorry, Houdini)

New plug-in versions for Maya, 3ds Max, Cinema 4D, and Katana are already available, so you can enjoy Arnold 7.4.3 in your DCC of choice—unless your choice is Houdini. At press time, Houdini users will need to wait for an updated plug-in. No magic workarounds provided.

20.07.2025: Update Houdini Plugin is available! Apparently we were too stupid to see it in the backend.

Platform Support and the CentOS 7 Eulogy

Arnold 7.4.3 runs on Windows 10+ and Linux with glibc 2.17+ and libstdc++ 4.8.5+ (basically, any current Linux workstation). MacOS 10.13 or later is also supported, but GPU rendering still needs an NVIDIA card—and thus remains Windows and Linux-only. This is also the final Arnold release to support CentOS 7, which officially reached end-of-life in 2024. If you’re still on CentOS 7, you’re officially out of excuses to delay that migration.

Pricing: Same Renderer, $15 More Per Year

Single-user subscriptions for Arnold are $55/month or $430/year. That’s a $15 annual price bump since 7.4.1. If you want to know what you’re paying for, that would be the Inference imager and (hopefully) a few more frames per second.

The post Autodesk ships Arnold 7.4.3 with Inference imager and GPU volume speedups first appeared on DIGITAL PRODUCTION and was written by Jürgen Firsching.

Dimension 5’s D5 Render 2.11 isn’t shy about showing off its new AI toys. The AI Agent system sits at the top of the interface, offering three components: SmartPlanting, which creates region-appropriate, botany-driven planting layouts; a Plant Schedule tool exporting lists of species, quantities, and price estimates for procurement; and D5 Bot, a support chatbot ready to field software-related queries in text form. No, it won’t do your client calls for you.

PBR Materials, Now With “Snap”

If your library is missing a perfect bark or a trendy stone, the new AI PBR Material Snap tool will extract PBR textures (diffuse, roughness, normal, etc.) from a reference image at up to 6K resolution, and then suggest visually similar materials already in D5’s own asset library. AI-driven Atmosphere Match and Post-AI Image Enhancer both receive new AI models, with added controls for style, reference image matching, enhancement strength, and—because it’s 2025—“Smart Retouch” inpainting.

Path Tracing for All

Real-time Path Tracing is now the default GI method, dropping its “alpha” label. The legacy GI engine is still around for compatibility, but path tracing now delivers improved accuracy in diffuse bounces, glass, caustics, and reflections—without a hit to performance, according to Dimension 5. AMD Radeon users finally get frame generation via FSR, catching up with NVIDIA’s DLSS crowd.

Cameras, Lights, and Parallel Projections

The camera toolset picks up a parallel projection mode, a must for elevations, sections, plans, and those axonometric views clients suddenly want at 5 p.m. There’s also a one-click “Align Views” tool for snapping a camera orthogonally to any selected face, and an Orbit Center icon for view manipulation. New lighting features include a Disc Light, improved directionality for Rect and Strip lights, support for IES profiles on point and area lights, and exposure compensation for anyone still fighting auto-exposure in 2025.

Layout Tools and Format Smorgasbord

The new Advanced Brush lets users paint, erase, and adjust vegetation in a scene. If you need to populate a street or scatter benches with architectural precision, the Custom Path tool distributes assets along user-defined curves. D5 Render 2.11 expands its list of supported formats: OBJ, DAE, 3DS, DXF, STL, GLTF, MMD, and performance-optimized Alembic imports. Scene List grouping, per-camera target selection for video, and improved view persistence round out the workflow updates.

For Teams: Cesium, XR Tour, and 8K Output

Teams-level subscribers gain some geospatial firepower: Cesium integration imports real-world terrain and context for site-accurate visualizations, while the XR Tour system enables browser-based 3D walkthroughs generated with Gaussian Splatting from smartphone photos and videos. Teams subscribers can now also output videos in 6K and 8K, if there’s a wall big enough to watch them on.

Bugs Squashed and Plug-ins Synced

Among the patch notes: round-cornered materials work again on NVIDIA’s 50-series, lost views are restored after scene deletion, and scatter-generation is more reliable. LiveSync plug-ins are updated for SketchUp, 3ds Max, Archicad, Blender 4.3–4.5, Cinema 4D 2025, and Revit 2026.

Pricing and Hardware

The Community edition is still free. Pro unlocks AI tools, the full asset library, and image sequence output for $38/month or $360/year. Teams (collaboration features, Cesium, XR Tour, and higher-resolution video) runs $75/month or $708/year. Windows 10 or later is required, along with a GPU at least as powerful as an NVIDIA GTX 1060, RTX, AMD Radeon RX 6400 XT, or Intel Arc A3. Supported DCCs include 3ds Max, Blender, Cinema 4D, Revit, Rhino, SketchUp, ArchiCAD, and Vectorworks.

Production Reality Check

As with all major releases, production artists are reminded to test new features in non-critical environments before rolling them out to live projects. Surprises belong in renders, not deadlines.

D5 Render Forum

The post D5 Render 2.11: Now With AI, Sci-Fi Plants, and Path Tracing by Default first appeared on DIGITAL PRODUCTION and was written by Bela Beier.