V-Ray Next for 3ds Max Automates Realistic Lights, Textures and Camera FX

Chaos Group has released V-Ray Next for 3ds Max, which uses intelligent systems to automate scene analysis, aiming to produce faster, cleaner renders with little or no input and to reduce the set-up time artists and designers have to devote from the creative side of their work.

In the same way that machine learning makes choices based on what is learned about a specific problem, V-Ray has been adopting learning techniques for analysing a scene as it renders. The new Adaptive Dome Light (ADL), introduced earlier in the year, automatically produces cleaner, more accurate image-based environment lighting that can render several times faster. The ADL also removes the need to set up skylight portals at windows and openings, making it especially helpful for architectural interiors. 

V-Ray Physical Camera

Scene intelligence also affects the V-Ray Physical Camera with point-and-shoot-style functionality like automatic exposure and white balance and a simpler UI, making generating renders similar to snapping photos. For photorealistic CG images, ray tracing has an interesting advantage over rasterisation because a ray tracer works like a physical camera – that is, it not only simulates the way light travels inside a space, but also how it enters the camera. With V-Ray’s Physical Camera you can imitate all the effects of a real-world camera including F-stop, shutter speed, depth of field, bokeh effects, lens distortion, tilt/shift and others.

Until now, the settings for these effects had to be adjusted manually. For someone who is already a photographer, adjusting camera settings feels natural and manual adjustments are preferable. But for most VFX artists, automating the settings means you can point and shoot – the equivalent of rendering - without shifting your focus from the looks in your final image.

Like the other smart functions in V-Ray Next, the Automatic Exposure and Automatic White Balance controls rely on adaptive learning to render the best result. By looking at the entire scene, V-Ray can automatically determine the proper exposure and white balance as modern cameras do. It also automatically adjusts the ISO value without changing the F-stop or shutter speed so that you can still control depth of field and motion blur separately.

GPU Rendering

V-Ray Next adds the new V-Ray GPU rendering architecture, which Chaos Group says is able to double the speed of production rendering due to a redesign of its kernel structure. V-Ray GPU combines high-performance speed and accuracy across interactive and production renders.

For ease of use, when the renderer is switched to V-Ray GPU, the interface will update to only show functions that are compatible with GPU rendering. When you select the processors you want to use, the right mode is chosen for you.

V-Ray Next's smart tools are also available in V-Ray GPU, including Adaptive Dome Lights and Displacement, Automatic Exposure and a number of others described below here. Wider third party support for software such as Forest Pack, Forest Color, MultiTexture and Bercon Noise mean that more productions can use V-Ray GPU exclusively. Production-oriented tools like Bucket Rendering, Deep Image support and VFB2 integration are likely additions later on.

The kernel structure redesign makes V-Ray adaptable to new developments in GPU hardware, allowing developers to incorporate new functions without impacting performance. For instance, V-Ray GPU has overtaken the previous generations of NVIDIA’s Quadro cards, running much faster on the new Quadro GV100.

Production artists will also notice the effect of GPU-accelerated support for environmental fog, volumetric effects and VRscans materials. Adding VrayEnvironmentFog, for example can improve the look of most shots almost straightaway. Lighting will behave more accurately, giving a realistic, hazy look as it scatters through the fog and loses some of its intensity. This also affects reflections and refractions in a physically correct manner.
Due V-Ray GPU support of the Volume Grid, volumetric data can be rendered from the OpenVDB, Field3D and PhoenixFD formats – in other words, data from Houdini, FumeFX, RealFlow, Phoenix FD and Maya.

Hair, Light, Noise and New Materials

The new Physical Hair Material helps produce more realistic-looking hair with accurate highlights. Specific looks for hair color and reflectivity can be difficult to achieve and control. Based on a Disney Research paper, look development has been simplified down to a few sliders that control glossiness, softness, randomness and others. Natural melanin controls can dial in any colour from blonde to redhead quickly. Hair dye sliders have also been added, so characters can sport a green mohawk if your brief asks for it.

Other new functionality in V-Ray Next includes the NVIDIA AI Denoiser for fast, real-time denoising based on NVIDIA OptiX AI acceleration. OptiX AI works from the theory that a computational network (derived through machine learning) can learn how to solve specific problems, either from supplied answers to the problem, or by learning from its own tests. Once the network understands better how to solve a problem like denoising, it can solve it much faster.

In theory, by passing thousands of different noisy renders to the neural network – rendered out of V-Ray or other renderers - plus the final versions, it could learn how to solve the noise problem using this image data, and then apply the solution to other ray traced images. Thus, with Denoise Render Elements, users can also denoise any render element used in still or animation projects, resulting in more control for compositors.

Lighting Analysis accurately analyses and measures the light levels in a scene, prompted by heat maps and data overlays that the user creates. Lighting analysis in V-Ray Next uses two components – the existing VRayLightMeter and the new VRayLightingAnalysis, which adds extra render elements to the V-Ray frame buffer showing the absolute, unscaled Illuminance (lux) and Luminance (candelas) values.

Layered Alembic Workflows support Alembic 1.7 with layering for faster, more efficient handling and updating of Alembic data. V-Ray Switch Material switches between the multiple materials applied to an object to speed up look development and variation decisions. V-Ray Plugin and Texture allows you to load any texture or material from any version of V-Ray into 3ds Max, including procedural textures and PBR materials for Unity and Unreal.

V-Ray Next will be available very soon for Windows. Every purchase of V-Ray 3.x made up until the launch will result in an automatic upgrade.  www.chaosgroup.com