WWW.DISSERTATION.XLIBX.INFO
FREE ELECTRONIC LIBRARY - Dissertations, online materials
 
<< HOME
CONTACTS



Pages:     | 1 || 3 | 4 |

«Natalya Tatarchuk 3D Application Research Group ATI Research What’s in It for You? • Share our lessons of developing an extensive environment for ...»

-- [ Page 2 ] --

Raindrops behave like lenses refracting and reflecting (both specularly and internally) scene radiances towards the camera. Tangent space is specified by the view matrix (since it’s a full-screen quad) Creating Rainfall

–  –  –

The first challenge lies in minimizing the repeating patterns that are inevitable when using a single static texture to model dynamic textured patterns. Initial raindrop distribution in the full-screen pass is simulated with an animated 8 bit raindrop placement texture. Artists can specify the rain direction and speed in world-space to simulate varied rainfall strength.

Moving Per-Frame Raindrop Distribution

–  –  –

At every time step we determine the raindrop clip space position (xi, yi) for every pixel in the composite pass. Using an artist-specified rain direction vector vr in clip space, the current raindrop position, and the rain speed, |vr| we compute the tentative raindrop distribution texture coordinates as presented.

Simulating Multiple Layers of Rain

• Artists specify a rain parallax parameter pr – Maps the depth range for rain layers

• Compute a randomized value for an individual raindrop ri – Using the concepts of stochastic distribution for simulation of dynamic textures

• Model multiple layers of rain in a single pass with a single texture fetch from the rainfall position placement texture – Using pr, ri, and the screenspace raindrop location (xi, yi) – pr * ri used as the w for a projective texture read

• Allows simulation of raindrops falling with different speed at different layers We simulate multiple layers of rain moving with different speeds at varied depths rendered in a single geometric layer. In order to create the illusion of several layers of raindrops, the artists specify a rain parallax parameter pr which maps the depth range for the rain layers in our scene. Using the concepts of stochastic distribution for simulation of dynamic textures, we compute a randomized value for an individual raindrop during the simulation, ri. Using the rain parallax value pr, the screen space individual raindrop location (xi, yi) for a given pixel computed earlier and the distribution parameter ri, we can model the multiple layers of rain in a single pass with a single texture fetch. The parallax value for the raindrop, multiplied by a distribution value, is used as the w parameter for a projective texture fetch to

sample from the rainfall movement texture:

wi = pr* ri.

This allows us to simulate raindrops falling with different speeds at different layers of rain without obvious repeating patterns.

Rain Appearance

–  –  –

We use a normal map for falling raindrops to model illumination for each raindrop in the composite layer. Our approach does not require any preprocessing and can handle an arbitrary number of light sources. The lighting model for illuminating individual raindrops is flexible.

Raindrops refract light from a large solid angle of the environment (including the sky) towards the camera. Specular and internal reflections further add to the brightness of the drop. Thus, a drop tends to be much brighter than its background (the portion of the scene it occludes).

The solid angle of the background occluded by a drop is far less than the total field of view of the drop itself. Thus, in spite of being transparent, the average brightness within a stationary drop (without motion-blur) does not depend strongly on its background. Falling raindrops produce motion-blurred intensities due to the finite integration time of a camera. These intensities are seen as streaks of rain. Unlike a stationary drop, the intensities of a rain streak depend on the brightness of the (stationary) drop as well as the background scene radiances and integration time of the camera.

Rain Appearance

–  –  –

We use a normal map for falling raindrops to model illumination for each raindrop in the composite layer. Our approach does not require any preprocessing and can handle an arbitrary number of light sources. The lighting model for illuminating individual raindrops is flexible.

Raindrops refract light from a large solid angle of the environment (including the sky) towards the camera. Specular and internal reflections further add to the brightness of the drop. Thus, a drop tends to be much brighter than its background (the portion of the scene it occludes).

The solid angle of the background occluded by a drop is far less than the total field of view of the drop itself. Thus, in spite of being transparent, the average brightness within a stationary drop (without motion-blur) does not depend strongly on its background. Falling raindrops produce motion-blurred intensities due to the finite integration time of a camera. These intensities are seen as streaks of rain. Unlike a stationary drop, the intensities of a rain streak depend on the brightness of the (stationary) drop as well as the background scene radiances and integration time of the camera.

Creating the Feeling of Strong Rain

• Realistic rain is very faint in bright regions of the scene and tends to appear stronger when light falls in a dark area – If this is modeled exactly, the rain appears too faint

• Simulate an old Hollywood trick for rain on film instead – The film crew add milk to water to make rain appear stronger on film – We do the same, by biasing rain color and opacity to appear whiter – Although exaggerated, this creates a perception of stronger rainfall

–  –  –





During rain, raindrops drizzle from various objects in the scene - trickling off gutter pipes, window ledges and so on.

We simulate this effect with the use of physics-based particle systems using screenaligned billboard representation for individual raindrops. The base particle system simulation uses the physical forces of gravity, wind and several animation parameters for raindrop movement. The artists can place any number of separate particle systems, culled by the camera frustum during rendering, throughout the environment to generate dripping raindrops.

Controlling Raindrop Transparency

• We attenuate raindrop opacity by distance

• Attenuate the opacity by Fresnel scaled and biased by artist-specified edge strength and bias – To make the raindrop appear less solid and billboard-like

• Observation: Raindrops should appear more transparent (like water) when the lightning strikes – Scaling the opacity by 1 – ½ * lightning Brightness does the trick – The particles still appear their respective transparency when there is no lightning – They become more translucent-like when the lightning strikes – This was used for both raindrop particles and raindrop splashes to attenuate their transparency Raindrop Splashes Raindrop Splashes

• Raindrops splash when they hit solid objects

• We simulated that effect with individual particles colliding with various objects – In our pipeline, this was achieved with special collider objects – In games or future engines, this can be done by directly colliding with objects Raindrop splashes

• Used a filmed high-quality splash

sequence for a milk drop:

• We used just one splash sequence for thousands of particles – The repetition can easily be noticeable – To reduce that, randomly scale particle size and transparency – Randomly flip u texture coordinate based on pre-specified particle random color We utilize the particle systems for rain drop splashes off the surface of objects, using pre-rendered high-quality splash sequence for a milk drop.

Illuminating the Splashes

–  –  –

In a strong rainfall, as the raindrops strike solid objects, they generate not only the splashes, but also the delicate halo outlines along the edges of objects. This is a very important visual cue which has been omitted from most of the existing rendered rain environments. We support rendering of this effect for objects in our scene (including the animated objects, such as cars by using normal ’fins’ (similar to fur rendering in real-time in [LPFH01]). To create a rain halo effect, we insert a degenerate quad which is extruded normal to the surface at object silhouettes. The actual halo is rendered on each such quad by using the rainfall algorithm as an animated texture, alpha-blended with the rest of the environment.

Rain Splatters

• Strong rainfall also generates an effect of raindrop splattering on the surface of wet materials

• We use a shells-based technique to create this effect – Again borrowing from the real-time fur rendering approach [Lengyel01]

–  –  –

Along with the water splashes from fast and heavy raindrops, strong rainfall also generates a more subtle effect with the raindrop splattering on the surface of wet materials. We use a shells-based technique to create the raindrop splatters. The shells technique is widely used for rendering fur in real-time (as described in [LPFH01]). We render the material with raindrop splatters as a series of extruded shells around the original object. The rain splatters are rendered on the surface of objects in the form of concentric circles. In each successive shell we expand the splash circle footprint with a series of animated texture fetches and blend onto the previous shells. This creates a very convincing effect of dynamic splatters on objects due to raindrops.

Demo: The Taxi GPU-Based Water Simulation for Puddle Rendering Water Ripples in Puddles

• Goal: Dynamic realistic wave motion of interacting ripples over the water surface – With fast simulation directly on the GPU

• Water ripples are generated as a result of rain drops falling onto the geometry in the scene – We support generation of raindrop ripples as a result of direct collision – However, for our scene that would require too much memory

• Instead, we use a stochastic seeding method for the simulation – Seeding rain drops into a texture – Spatter raindrops as points into the water simulation texture

• Can also render object outlines into the seeding texture to generate wakes The raindrop particle collisions generate ripples in rain puddles in our scene. The goal was to render dynamic realistic wave motion of interacting ripples over the water surface using the GPU for fast simulation. Due to memory considerations, we currently use the stochastic seeding method, rather than direct collision response, for a simulation on a 256x256 lattice.

We splatter the raindrops as point primitives into the water simulation texture with the RGB value proportional to the raindrop mass during the first pass of the simulation. This method can be applied to generate dynamic water surface response for arbitrary objects.

This can be achieved by rendering an orthographic projection of the objects into the seeding texture using the object’s mass as the function for color of the object’s outline.

Pass 1: Render seeds into the first water simulation buffer These seeds rendered as initial positions of water ripples The seeds ‘excite’ ripple propagation Pass 2 and Pass 3: Perform integration on water surface simulation Uses ‘ping-pong’ texture feedback approach We only use two passes, but more will help with system stability if time step desired to be smaller These passes generate water height field Pass 4: Generate water normals Sample from the water normals texture when rendering an object with puddles Water Surface Approximation

• Approximate water surface with a lattice of points – We render our surfaces with a 256 x 256 simulation – Each lattice contains information about water surface at that point

• Current position as a height value

• Previous time step’s position

• We simulate the water lattice entirely on GPU – Using a texture to store lattice positions and its attributes – Similar to “Interactive Simulation of Water Water lattice heights: Current frame’s Surfaces” by M. Gomez (Game height in R channel and previous Programming Gems) frame’s height in G channel – However, there the lattice is approximated with vertices on the CPU We approximate the water surface as a lattice of points on the GPU containing the information about the water surface in that location (we store the current and previous time step wave displacement values). These quantities can be packed into a single 32 bit texture using 16 bit per channel, giving a good precision balance for computing displacements. This would generate a wake effect in the water surface.

Simulate Water Interaction

–  –  –

Similar to real-life raindrops, a single raindrop in our system excites multiple interacting ripples on the water surface. The rendered seeds act as the initial ripple positions by exciting the ripple propagation in the subsequent passes. Real-life raindrops generate multiple ripples that interact with other ripples on the water surface. We implement the same model. We render a raindrop into a wave seed texture using a dampened sine wave as the function for raindrop mass. This approximates the concentric circular ripples generated by a typical raindrop in a water puddle.

Water Surface Response

• Treat water surface as a thin elastic membrane – Ignore gravity and other forces – Only account for surface tension

• At every time step, infinitesimal sections of this surface are displaced – Due to tension exerted from their direct neighbors – Acting as spring forces to minimize space between them

–  –  –

We use explicit Euler integration in DirectX9.0 pixel shaders to solve this PDE in real-time by using a texture feedback approach to determine the water wave heights for each point on the lattice. We found that two passes are sufficient for a stable simulation. During the final pass we compute the normals for the water displacements using the Sobel filter.

Integrating Water Puddles

• We render a single water simulation for the entire demo – All objects with water puddles sample from that (for example, streets, rooftop ledge, etc)



Pages:     | 1 || 3 | 4 |


Similar works:

«Loughborough University Institutional Repository How are false negative cases perceived by mammographers? Which abnormalities are misinterpreted and which go undetected? This item was submitted to Loughborough University's Institutional Repository by the/an author. Citation: SCOTT, H.J., GALE, A.G. and HILL, S., 2008. How are false negative cases perceived by mammographers? Which abnormalities are misinterpreted and which go undetected? IN: Sahiner, B. and Manning, D.J. (eds.). Medical Imaging...»

«S3C D 1.1 FINAL FI.Energy-2012308765 S3C D1.1 FINAL WP 1: “Framing – Development of the theoretical framework” Deliverable 1.1: “Report on state-of-the-art and theoretical framework for end-user behaviour and market roles” Contractual Date of Delivery to the CEC: 31 October 2013 (Month 12) Actual Date of Delivery to the CEC: 8 November 2013 Author(s): S3C Consortium Participant(s): VITO (task leader), ECN, RSE, INEA, BAUM Workpackage: WP1 Task 1.1 Estimated person months: 8.75...»

«University Bible Fellowship: What Happens When Missionaries from Korea Descend on North American College Campuses?1 Joseph L. Schafer2, Mark Yoon3 and A. Scott Moreau4 South Korea holds a commanding lead in Majority World missions (Park, 2002; Johnstone and Mandryk, 2003), with approximately 13,000 long term missionaries in foreign lands (Moll, 2006). Most of these missionaries serve in south east Asia, Africa and former Soviet republics (Moon, 2001). By and large, Korean missions follow the...»

«A Post-Colonial Critique of the (Mis)Representation of KoreanAmericans in Children's Picture Books Item type text; Electronic Dissertation Authors Sung, Yoo Kyung Publisher The University of Arizona. Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the...»

«NUREG-0800 U.S. NUCLEAR REGULATORY COMMISSION STANDARD REVIEW PLAN 3.5.1.4 MISSILES GENERATED BY EXTREME WINDS REVIEW RESPONSIBILITIES Primary Organization responsible for the review of plant design for protection of structures, systems, and components from internal and external hazards Secondary Organization responsible for the review of meteorological data I. AREAS OF REVIEW Title 10 of the Code of Federal Regulations (10 CFR) Part 50, Appendix A, General Design Criterion (GDC) 2, “Design...»

«Interval Scheduling Interval Partitioning Minimising Lateness Greedy Algorithms T. M. Murali February 3, 2009 Interval Scheduling Interval Partitioning Minimising Lateness Algorithm Design Start discussion of different ways of designing algorithms. Greedy algorithms, divide and conquer, dynamic programming. Discuss principles that can solve a variety of problem types. Design an algorithm, prove its correctness, analyse its complexity. Interval Scheduling Interval Partitioning Minimising...»

«Approximate Maximum Likelihood Methods for Large Scale Blind Classification and Identification in Digital Communications Steffen Barembruch To cite this version: Steffen Barembruch. Approximate Maximum Likelihood Methods for Large Scale Blind Classification and Identification in Digital Communications. Applications [stat.AP]. T´l´com Parisee Tech, 2010. English. pastel-00574365 HAL Id: pastel-00574365 https://pastel.archives-ouvertes.fr/pastel-00574365 Submitted on 7 Mar 2011 HAL is a...»

«King’s View Magazine, No. 873 February 2015 King’s Park Church of Scotland, Glasgow. 242 Castlemilk Road, GLASGOW G44 4LB Mission Statement: King’s Park Church sharing God’s love by reaching out to others. “You limit your future when you dwell in the past. Consider your future plans—see them before they happen” See Proverbs 23:4 1930-2015—85 Years of Spiritual Service 2 FOR YOUR INFORMATION King’s Park Church of Scotland, Glasgow. 242 Castlemilk Road, G44 4LB Minister: Rev...»

«Maternal and Child Health Services Title V Block Grant  Delaware Created on 7/17/2015 at 3:44 PM Table of Contents I. General Requirements 4 I.A. Letter of Transmittal 4 I.B. Face Sheet 5 I.C. Assurances and Certifications 5 I.D. Table of Contents 5 I.E. Application/Annual Report Executive Summary 5 II. Components of the Application/Annual Report 11 II.A. Overview of the State 11 II.B. Five Year Needs Assessment Summary 16 II.B.1. Process 16...»

«Vaccine update Issue 213, March 2014 New vaccines for summer 2014 Replacement whooping cough vaccine Boostrix IPV will be replacing Repevax as the vaccination for the pregnant women programme to protect their newborn infants from pertussis. Additional infant primary vaccine Infanrix IPV/Hib is being introduced alongside Pediacel as an infant primary vaccine. Infanrix/Hib and Pediacel will be fully interchangeable for the infant vaccinations at two, three and four months. See p.9 in the Vaccine...»

«30 Years of Harm Minimisation – How Far Have We Come? (An Australian Perspective) Evidence based review of the influence and impact on drug use in Australia under the ‘Harm Minimisation Strategy’. Authors D. Steenholdt B.Sc.(Hons), Dip.Ed., B.Ed., M.Ed.St. R. M. Colquhoun D.H.Sc., M.App.Sc.(Neuroscience), B.Sc. (Hons) S.W.Varcoe Dip Min, A.C.R.A.C.S. 3/1/2015 P: 1300 975 002 F: 1300 952 551 W: www.dalgarnoinstitute.org.au E: director@dalgarnoinstute.org.au Abstract The aim of this paper...»

«          STRATEGIC SELECTING THE RIGHT RESEARCH-BASED TEACHING: STRATEGY FOR EVERY LESSON NSDC  Presented by     Tr. Harvey F. Silver, Ed.D.    Tuesday July 21st , 2009  Silver Strong & Associates   |   227 First Street   |   HoHoKus, NJ 07423   |   www.ThoughtfulClassroom.com               Table of Contents  What is the key to teaching more effectively?  The Strategic Teacher: An Introduction ...»





 
<<  HOME   |    CONTACTS
2016 www.dissertation.xlibx.info - Dissertations, online materials

Materials of this site are available for review, all rights belong to their respective owners.
If you do not agree with the fact that your material is placed on this site, please, email us, we will within 1-2 business days delete him.