Real-Time Fire, Smoke, Weather, Liquids, Vegetation, and Terrain

Fire, smoke, landscapes, weather -- aspects of the natural world we take for granted that require great care in real-time rendering.

Fire, Smoke, Weather, and Liquids

Samples from NVIDIA Graphics SDK 10.5:

Lightning (Whitepaper)

This sample shows how to simulate and animate different kinds of lighting effects using DirectX10 features. The structure of a lightning beam is created by using fractal subdivision of line segments, efficiently implemented using geometry shaders and stream out. A geometry shader is then used to create camera facing billboards with a procedurally created color gradient. The atmospheric glow effect is implemented by downsampling and blurring the rendered bolts of lighting using a Gaussian filter kernel with separate extinction coefficients per color channel in order to fake atmospheric scattering.

Smoke (Video)

This demo shows how to simulate and render real-time three dimensional smoke and Fire. Both simulation and rendering are made easier using DirectX 10's new render to 3D texture functionality. The Smoke and Fire also correctly interact with moving objects in thier path and composite seamlessly into the scene.

Rain (Whitepaper)

This sample shows how to animate and render rain as a particle system entirely on the GPU. Stream out is used to animate the rain particles over time, and the geometry shader is used to extrude rain particles into quads at render time. Rain particles are realistically rendered using precomputed textures (indexed by light direction and viewing direction) which are stored in a texture array.

Soft Particles (Whitepaper)

Particle sprites in games commonly produce artifacts – unnaturally sharp edges – where they intersect the rest of the scene. This sample shows two ways to fade out flat sprites against a 3D scene, softening the artificial edges. Two solutions are implemented: one uses the ability of DirectX10 to read the depth buffer as a texture; the other uses a more conventional second render target to store depth values. - one other rendering Z in a 2nd render target.

Perlin Fire (Whitepaper)

This sample uses an improved Perlin noise algorithm on the GPU. This method renders realistically animated fully procedural fire, with individual uniquely animated flame shapes generated in a pixel shader that uses three-dimensional simplex flow noise or four-dimensional simplex noise. The improved Perlin noise algorithm requires many complex scalar computations, making it a great match for the scalar architecture of the NVIDIA® GeForce® 8800 GPU.

Samples from NVIDIA Graphics SDK 9.52:

Snow Accumulation (Whitepaper)

This sample shows a technique for procedurally adding snow to a "non snow" scene.

Vertex Texture Fetch Water (Whitepaper)

This sample demonstrates a technique for simulating and rendering water. The water is simulated via Verlet integration of the 2D wave equation using a pixel shader. The simulation result is used by a vertex shader via vertex texture fetch (VTF). The water surface is rendered by combining screen-space refraction and reflection textures.

Water Interaction (Whitepaper)

Render-to-texture is used to drive a procedural simulation of water. The water is rendered with a technique similar to environment mapped bump mapping (EMBM), but an enhancement allows the EMBM rotation matrix to vary per-vertex and fade the bumps out as distance to the viewer increases. This sample requires vertex and pixel shaders 1.1.

Rainbow Fogbow (Whitepaper)

This demonstrates how to render a physically modeled rainbow in realtime. It does so by using a lookup texture with the color of scattered white light given the angle between a view vector and a light vector. To blend the rainbow nicely with the scene, the scene is rendered encoding moisture data in a texture and using this to change the intensity of the rainbow. This same technique can also be used to render halos, corona, fogbows, and any other radially symetrical optical light scattering effects.

GPGPU Fluid (Whitepaper)

This code sample demonstrates fast, realistic fluid dynamics simulation on the GPU. The sample solves the Navier-Stokes equations for incompressible fluid flow using a technique originally presented at SIGGRAPH 1999 by Jos Stam. The sample allows the user to draw arbitrary obstacles and flow viscous or inviscid fluid around them.

GPU Particles (Whitepaper by Lutz Latta)

This sample implements a large-scale particle system entirely on the GPU. The positions and velocities of each particle are stored in floating point textures. Fragment programs are used to update the velocities and positions of the particles by rendering to texture each time step. The particles also collide against a sphere object, and a terrain heightfield which is stored in a texture. If available, the multiple draw buffers extension (MRT) is used to update the position and velocities in a single pass. The particles are rendered as point sprites. The position texture is converted into a vertex array for rendering the particles using the vertex buffer and pixel buffer object extensions (VBO and PBO). On the GeForce 6800, this method can render a million particles at about 20 frames per second. This example is inspired by Lutz Latta's talk from GDC 2004, "Building a Million Particle System".

Vertex Shader Water

This sample gives the appearance that the viewer is surrounded by a large grid of vertices (because of the free rotation), but switching to wireframe or increasing the frustum angle makes it apparent that the vertices are a static mesh with the height, normal, and texture coordinates being calculated onthe-fly based on the direction and height of the viewer. This technique allows for very GPU-friendly water animations because the static mesh can be precomputed. The vertices are displaced using sine waves, and in this example a loop is used to sum five sine waves to achieve realistic effects.

Vertex Noise

This example demonstrates an implementation of Perlin noise using vertex programs. An animated 3D noise function is used to displace the vertices of a sphere along the vertex normal. The geometry is entirely static, but is displaced on the fly by the vertex program hardware. Perlin noise is implemented for the vertex program profile using recursive lookups into a permutation table stored in constant memory. The size of this table determines the period at which the noise function repeats. 3D noise costs around 65 instructions, 2D noise around 45, 1D noise around 20.

GPU Gems 2 online:

Chapter 13. Implementing the mental images Phenomena Renderer on the GPU

Chapter 16. Accurate Atmospheric Scattering

Chapter 18. Using Vertex Texture Displacement for Realistic Water Rendering

Chapter 26. Implementing Improved Perlin Noise

GPU Gems online: Part I, Natural Effects

Chapter 1. Effective Water Simulation from Physical Models

Chapter 2. Rendering Water Caustics

Chapter 5. Implementing Improved Perlin Noise

Chapter 6. Fire in the "Vulcan" Demo

Chapter 8. Simulating Diffraction

GDC 2008: Particle-Based Fluid Simulation for Games (Video)

The computational power of modern GPUs has reached the point where realistic free-flowing fluids can be simulated and rendered in real-time. This talk will provide an overview of fluid simulation techniques for games and then describe in detail the advantages of particle-based simulations and how they can be implemented efficiently on the GPU.

GDC 2007:

• NVIDIA Demo Team Secrets: Cascades
• Real-Time Atmospheric Effects in Games, Revisted
• Real-time Volumetric Smoke Using Direct3D 10

Siggraph 2007: Smoke, Fire, and Water with Fluid Dynamics

Physically based animation of fluids such as smoke, water and fire provides some of the most stunning visuals in computer graphics, but has historically been the domain of high-quality offline rendering due to its great computational cost. In this presentation we will talk about how to simulate and render these effects in real time on the GPU, and also how to build a unified system that seamlessly integrates these effects into real time applications such as video games.

Vegetation

Samples from NVIDIA Graphics SDK 10.5:

Christmas Tree (Whitepaper)

Chistmas Tree demonstrates using new fourth generation shader capabilities to render a visually complex Christmas Tree. The sample uses deferred shading into a high dynamic range framebuffer object to handle the complex lighting environment

GPU Gems 2 online:

Chapter 1. Toward Photorealism in Virtual Botany

GPU Gems online: Part I, Natural Effects

Chapter 7. Rendering Countless Blades of Waving Grass

GDC 2007:

• NVIDIA Demo Team Secrets: Cascades

Terrain

Samples from NVIDIA Graphics SDK 10.5:

Clipmaps (Whitepaper)

Clipmaps are a feature (first implemented on SGI workstations) that allow mapping extremely high resolution textures to terrains. The original SGI implementation required highly specialized, custom hardware. The advanced features of the NVIDIA® GeForce® 8800 now permit the same algorithm using consumer hardware.

Soft Particles (Whitepaper)

Particle sprites in games commonly produce artifacts – unnaturally sharp edges – where they intersect the rest of the scene. This sample shows two ways to fade out flat sprites against a 3D scene, softening the artificial edges. Two solutions are implemented: one uses the ability of DirectX10 to read the depth buffer as a texture; the other uses a more conventional second render target to store depth values. - one other rendering Z in a 2nd render target.

Texture Arrays (terrain) (Whitepaper)

This sample is a simple example using Texture Arrays in DX10. It uses a texture array as a palette of terrain textures to render a terrain mesh with many textures in a single draw call. A similar effect in previous APIs would require a texture atlas or multiple draw calls.

Transform Feedback Fractal (Whitepaper)

"Transform Feedback Fractal" shows an advanced use of the transform feedback and the geometry shader, by iteratively feeding the data through the transformation stages, adding more detail each time.

Samples from NVIDIA Graphics SDK 9.52:

Snow Accumulation (Whitepaper)

This sample shows a technique for procedurally adding snow to a "non snow" scene.

GPU Particles (Whitepaper by Lutz Latta)

This sample implements a large-scale particle system entirely on the GPU. The positions and velocities of each particle are stored in floating point textures. Fragment programs are used to update the velocities and positions of the particles by rendering to texture each time step. The particles also collide against a sphere object, and a terrain heightfield which is stored in a texture. If available, the multiple draw buffers extension (MRT) is used to update the position and velocities in a single pass. The particles are rendered as point sprites. The position texture is converted into a vertex array for rendering the particles using the vertex buffer and pixel buffer object extensions (VBO and PBO). On the GeForce 6800, this method can render a million particles at about 20 frames per second. This example is inspired by Lutz Latta's talk from GDC 2004, "Building a Million Particle System".

Vertex Noise

This example demonstrates an implementation of Perlin noise using vertex programs. An animated 3D noise function is used to displace the vertices of a sphere along the vertex normal. The geometry is entirely static, but is displaced on the fly by the vertex program hardware. Perlin noise is implemented for the vertex program profile using recursive lookups into a permutation table stored in constant memory. The size of this table determines the period at which the noise function repeats. 3D noise costs around 65 instructions, 2D noise around 45, 1D noise around 20.

GPU Gems 2 online:

Chapter 2. Terrain Rendering Using GPU-Based Geometry Clipmaps

Chapter 7. Adaptive Tessellation of Subdivision Surfaces with Displacement Mapping

Chapter 8. Per-Pixel Displacement Mapping with Distance Functions

Chapter 12. Tile-Based Texture Mapping

Chapter 16. Accurate Atmospheric Scattering

Chapter 26. Implementing Improved Perlin Noise

GPU Gems online: Part I, Natural Effects

Chapter 5. Implementing Improved Perlin Noise

GDC 2007:

• NVIDIA Demo Team Secrets: Cascades
• Real-Time Atmospheric Effects in Games, Revisted

Siggraph 2008:Adaptive Terrain Tessellation on the GPU

( Japanese Version )

Next-generation techniques implement highly-programmable tessellation entirely on the GPU. We explain how tessellation can be applied to terrain rendering with displacement mapping. Our tessellation scheme is adaptive, with the polygon LOD varying as a function of terrain roughness and also with view-dependent silhouette detection.

Want to Learn More? NVIDIA Documentation Home Page