Redshift 2027 generates fur right at render time

Maxon has released Redshift 2027.0 — the next version of the GPU-accelerated renderer for VFX, motion graphics and visualization. The key new feature is the new procedural Fur object, which builds hair, fur and grass at render time. The release arrived as part of the Maxon 2027.0 autumn update lineup, which also includes Cinema 4D, ZBrush, Red Giant and Autograph; in the release notes this version of Redshift is also labeled 2026.10.

The second part of the update concerns speed. The developers reworked the memory management system: time to first pixel has been reduced, and on an architectural scene shown in the release notes render time drops by 15%.

Fur is computed at render time

Fur Object is a fast render-time solution for hair, fur and grass. The artist controls it with standard parameters: Density, Gravity, Curl, Taper, plus there is an option to randomize strand properties. Maxon describes the system more broadly than just fur: it is also intended for fibers, woven materials and knitted surfaces. The procedural approach means the fibers do not need to be generated and cached in advance — the renderer creates them.

Fur Object is so far available in four integrations: 3ds Max, Cinema 4D, Houdini and Maya.

Redshift Live closes in on final quality

The real-time engine that replaced Redshift RT in version 2026.4 gained traced depth of field, refraction with ray bending in transparent materials, and light spread support. There is no dispersion yet.

By Maxon's own assessment, the changes bring Redshift Live closer to functional parity with the final render engine, but the key limitations remain: the real-time engine still does not display volumes, hair and point clouds — meaning the new procedural fur apparently cannot be seen in it.

3ds Max renders tyFlow fire natively

3ds Max users received native support for tyFlow Inferno simulations. The popular add-on for particles and multiphysics learned to compute volumetric fire and smoke simulations in tyFlow 2.0, released earlier this year; now such volumes are supported in Redshift directly.

The other integrations received more targeted updates. Cinema 4D gained the Auto Focus Distance depth-of-field setting, which automatically sets the focus distance by the object in the center of frame; material preview spheres have been redone, and hex tiling and triplanar tiling options were added to the Material Tag — hex tiling, as Maxon describes it, breaks up noticeable texture repetition. Maya moved to the same optical camera and backplate system as the other hosts and gained support for the native Maya Light Lister. In Houdini, the Redshift Object Lister got a shelf icon. Separately, Maxon added official support for ArchiCAD — another CAD package after Vectorworks and Revit, whose integrations shipped earlier this year.

ACES 2 is no longer a manual setting

Redshift now supports ACES 2 — the current version of the film industry's color management standard. The ACES 2 OCIO config is now used by default in Cinema 4D; in Houdini and Maya it can be set manually. On the hardware side: Redshift added support for NVIDIA RTX Spark, the new "superchip" for laptops and mobile workstations, joining DaVinci Resolve and OctaneRender.

Licensing has not changed: rental only, $49 per month or $289 per year, with Redshift included in the Cinema 4D subscription and in the Maxon One bundle. System requirements — Windows 10 and newer, Linux with glibc 2.28 and newer, macOS 15.0 and newer. Supported host versions: 3ds Max 2018+, Cinema 4D 2023+, Houdini 19.0+, Katana 5.0+, Maya 2018+ (on Linux — 2022+), Revit 2024+, Vectorworks 2026.4+, ZBrush 2023+.

Why renderers are taking over more and more pipeline stages

If you look at the release not by its feature list but by the direction of travel, one line stands out: Redshift is absorbing work that used to be done before rendering. Fur is generated at render time, volumes from another application's simulation are read natively, the interactive engine reaches the effects for which a final render used to be launched. Similar logic is visible in competing GPU solutions too — probably a general response to the fact that the bottleneck in production has become not only compute speed but also the number of intermediate steps and the volume of data the artist moves between applications.

Against this backdrop the bet on compact hardware like RTX Spark also makes sense: if the renderer handles memory more efficiently, the range of tasks that can be closed out without a full workstation or farm probably widens. If the trend continues, the next versions can logically be expected to head the same way — narrowing the gap between interactive preview and the final frame, removing Redshift Live's remaining limitations on volumes and hair, and the renderer further absorbing what is today considered scene preparation.

Community reaction

There is no broad wave around Redshift 2027 on X: Maxon's posts have dozens of likes and a handful of replies, with no burst of enthusiasm and no outrage. Those who did respond more often single out not the procedural fur itself, but the promised speed-up to first pixel and the memory rework. People immediately ask whether the gain carries over to batch rendering in Maya, and they are cautious: they want to see how it behaves on real production scenes, not just in a video. On the fur there is one substantive complaint. A modeler writes that the fur does not sit on surfaces with subdivision and stays on the base mesh, and asks whether this will be fixed. Alongside it, brief skepticism not about the quality of the fur but about the announcement image: whether it was made by a neural network. Overall the mood is restrained — interest in speed, one targeted complaint about the fur, and almost no broad controversy.

"The new fur doesn't seem to work with subdivision surfaces. In the image only the base mesh level remains."

— @MorleyArts (X)

"Faster to first pixel and better memory handling are probably the most interesting things in this update. The new features are nice, but anything that eliminates the idle waiting while you set up the scene is already a win. Curious how this gain behaves on real production scenes."

— @cgclick (X)

"Very encouraging. Does this speed gain also apply to batch rendering in Maya?"

— @DailyMobster (X)

Why is Maxon moving fur into the renderer: saving memory in the scene, pressure from Houdini and tyFlow, a push toward GPU procedurality?