Docstoc

Shadowing in XNA

Document Sample
Shadowing in XNA Powered By Docstoc
					The “Dark” Side of Game Development
Let’s see some ID…
   Julian Spillane
     CEO / Technical Director, Frozen North
     Productions, Inc.
      ○ We make video games.
     Currently working on a next-gen
     XBLA project
      ○ January 2008 release
      ○ PC release to follow soon after…
Break it down…
   XNA?
     How does it stack up
     Reception by the pros
   Shadowing in Games
   Shadowing Algorithms
     Maps, volumes, geometry projection…
     Why I prefer maps
   Basic Shadow Map Theory & Implementation
   Variants
     DPSM, PSM, TSM, LiSPSM, PSSM, VSM…
   Tips for Optimization, Soft Shadows
   Further Reading
XNA’s Not Acronymed
   What IS XNA?
     A managed framework for developing games on
      both PC and Xbox 360
      ○ Graphics device, input, and audio management
        along with math helpers
   Who is using XNA?
     Hobbyists
     Independent developers
     … and yes, professionals
      ○ Torpex Games, Hidden Path Entertainment,
        GarageGames Inc.
XNA’s Not Acronymed (cont’d)
   Why XNA?
     Unmanaged background
      ○ C, C++, assembly
      ○ XNA exposed me to a new language
     XNA allows for RAD
      ○ Full games can be put together in obscenely short
        periods of time
      ○ Allows developers to focus on content, gameplay,
        balancing over technical issues
     XNA does things more inuitively than DirectX
      ○ Render Targets
      ○ Device Management
   Preaching to the choir? ;)
XNA’s Not Acronymed (cont’d)




 Full game built in 2 weeks using XNA with no prior
 knowledge of the API…
Shadowing in Games
   Why are shadows important?
     Add a sense of presence and volume to a scene
     Important visual cues
     Make for interesting gameplay mechanics
     Adds to realism (caveats?)
     Players expect dynamic lighting
      ○ Shadows play a major role

   How many different ways are there to
    shadow?
     In a word: lots
Shadowing Algorithms
   Geometry Based                 Image Based
     Projected Shadows              Shadow Map
      ○ Geometry collapse             ○ Depth-based
        using “Shadow Matrix”           shadowing system
     Shadow Volumes                  ○ Many variants: PSM,
      ○ Regular volumes                 TSM, LiSPSM, PSSM,
                                        CSM, VSM, DPSM,
      ○ Z-fail / “Carmack’s
                                        etc.
        Reverse”
                                      ○ Relatively O(1) in
      ○ GPU-generated
                                        terms of geometric
     Both cases ~O(n) in               scene complexity
      terms of scene                 Forward Shadow
      complexity                      Mapping
                                      ○ Proposed technique
Shadowing Algorithms (cont’d)
   Geometry Based
     Pros
      ○ Sharp, crisp shadows
      ○ Easy to compute
         If it’s visible on-screen, you already have the data to shadow it
      ○ Easily extendable to the vertex shader
         Speed benefits through pipeline-integration
     Cons
      ○ Shadow complexity is O(n), n = # edges
         Means speed α complexity-1
         Fine balance between object count and edge resolution
      ○ Requires image-space modification or ugly jitter for soft-
        shadows
      ○ No support for alpha-blended textures without extreme
        trickery
          We’re talking leprechaun-level, here…
    Shadowing Algorithms (cont’d)
     Shadow Volumes – Algorithm, Demonstration




Images courtesy of Østfold University College, Norway and NVIDIA Corporation
Shadowing Algorithms (cont’d)
   Image Based
     Pros
      ○ Shadow complexity α scene complexity!!!
         Means cube equally expensive as Marcus Fenix ™ (sorta…)
      ○ Gives shadowed alpha for free
      ○ Soft shadows feasible in real-time
      ○ Self-shadowing is free
     Cons
      ○ Inherently aliased / artifact’d
          Aliasing from texture resolution and light projection settings
          Artifacts from limitations of texture projection
          Both are mitigated through technique variations
      ○ Memory intensive
          Memory consumption α shadow quality
           - Dependent on texture size
      ○ No unified approach
          Different techniques for different situations
    Shadowing Algorithms (cont’d)




Images courtesy of me. Demonstration of vanilla shadowing using PCF and screen-space blur.
Shadowing Algorithms (cont’d)
   This presentation will focus on shadow
    mapping, its theory, implementation and
    variants
     On modern hardware, pros far outweigh cons
     Self-shadowing and alpha blending!
     Scenes in games are growing in complexity
      ○ Vertex-bound techniques will cause even more
        problems as time progresses
     Independent of model / vertex data architecture
     Personal preference ;)
What Is Shadow Mapping?
   Rendering from the light’s POV
     Using results to generate projective texture
      ○ Generally filled with single-color depth
        information
   Two-Pass Process
     Generating the shadow map
     Shadowing the scene
   Shader Approach
     Can be done using FF, but deprecated
Creating the Shadow Map
   Theory
     Render from light’s point of view
      ○ Any object that can be seen must be lit
      ○ Store the depth value of every element
     Render the scene as normal
      ○ Take the position of every vertex
         Transform into light-space coordinates
         Retrieve the point as it is seen from the point of view of
           the light (depth map)
      ○ Depth-compare with light-view map
         If depth value of surface is further away, it must be
          behind an occluder and shadowed (z-fail test)
Creating the Shadow Map
   The Shadow Map
     Can be represented in many formats
      ○ Best Choice: R32F (single channel float)
         Only if hardware supports…
      ○ Can use ARGB, just less precision
         Unless interesting depth hashing with the channels…

     Use XNA RenderTarget2D
      ○ Writeable surface, can convert to texture
      ○ Easy to use
Creating the Shadow Map
   Matrices
     Need light’s view, projection matrices
      ○ Choice in projection
         Ortho best for large, directional lights (i.e. sun)
           - LightProj = Matrix.CreateOrthographic(w, h, n, f)
             - w, h are width and height of view volume
             - n, f are near and far planes of the view volume
               • Recommend calculating these instead of arbitrary
         Perspective for local, spot lights
           - LightProj = Matrix.CreatePerspectiveFOV(f, a, n, f)
             - f: perspective FOV, a: perspective aspect ratio
      ○ Implies we need light information
         Position, target, up vector
         LightView = Matrix.CreateLookAt(position, target, up)
Creating the Shadow Map
Creating the Shadow Map
   Effect
     Let’s now consider the effect
      ○ Needs to push unlit vertices
         Return depth information
      ○ Store depth independently
         POSITION cannot be read directly
      ○ Very simple shader
         Don’t make it any more complex than it needs to be
Creating the Shadow Map
Creating the Shadow Map
    Creating the Shadow Map




Statue seen from the light’s point of view. Blue because of R32F texture. Notice gradation: depth values
Shadowing the Scene
   Projection
     Now we have depth values stored
      ○ Need to render a second pass from camera
      ○ Project the shadow texture into the scene and depth
        compare
     Yet another matrix
      ○ Texture bias and projection matrix
          Used to map position from projection space into texture space

                            0.5 0.0 0.0 0.0
                            0.0  0.5 0.0 0.0
                        P                         
                            0.0 0.0 1.0 0.0
                                                   
                            tX     tY      0.0 1.0 
                        tX  0.5  (0.5 / texWidth)
                        tY  0.5  (0.5 / texHeight)
Shadowing the Scene
   Depth Comparison
     Transform position using previous matrix
      ○ Compare z-value to value stored in shadow
        sampler
      ○ if stored depth < visible depth at that point
         Must be occluded and therefore shadowed

     Easy pixel shader comparison
      ○ Single texture lookup with 1 inline conditional
        branch (ps_2_0 only)
Shadowing the Scene
Shadowing the Scene
    Shadowing the Scene




Images courtesy of Frozen North Productions, Inc.
Shadowing the Scene
   Final Steps
     Merging with the scene
      ○ Simply multiply shadow term by diffuse colour
        / lighting / texture
     So…improvements?
      ○ Aliasing artifacts, projection artifacts, hard
        edges, the list goes on
      ○ How do we mitigate the problems?
         Variations on the technique
         I will cover overviews with links to papers and demos
    Shadowing the Scene




Images courtesy of Frozen North Productions, Inc.
Shadow Map Variants
   Dual-Paraboloid Shadow Maps
     Basic shadow mapping is inherently spot / directional
      ○ Need creative solution to handle omnidirecitonal (point) lights
     Paraboloid Mapping
      ○ “[I]mage obtained by an orthographic camera view a perfectly
        reflecting paraboloid.”
      ○ Single map can cover a full hemisphere
          Implies only 2 rendering passes for omni light vs 6 passes for cube lookup
      ○ Find a point P in R3 on the paraboloid that reflects a given
        direction towards +/- Z
      ○ Depth value calculated by using the distance from surface point
        to center of paraboloid
          Extends from a 2D-3D mapping to a 3D-3D mapping
     “Shadow Mapping for Hemispherical and Omnidirectional
      Light Sources” Brabec, et al., Computer Graphics Group, Max-
      Planck Institute, Saarbrucken Germany
      ○ http://www.mpi-inf.mpg.de/~tannen/papers/cgi_02.pdf
     http://www.gamedev.net/reference/articles/article2308.asp
Shadow Map Variants
   Perspective Shadow Maps
     Aliasing due to depth resolution issues
     PSMs
      ○ Generated post-perspective transformation
      ○ Light is transformed projectively to the unit cube
      ○ Reduces perspective aliasing since the depth values are
         considered after perspective proejction
     Caveats
      ○ Cases to handle for objects behind the viewer but cast shadows
        (lit from behind the viewer)
      ○ Comparatively convoluted implementation
     “Perspective Shadow Maps”, Stamminger, Marc and Drettakis,
      George, REVES – INRIA Sophia Antipolis, France
      ○ http://www-
         sop.inria.fr/reves/Basilic/2002/SD02/PerspectiveShadowMaps.pdf
Shadow Map Variants
   Trapezoidal Shadow Maps
     Similar to PSMs
      ○ Approximates eye frustum as seen from the light
        using a trapezoid to warp it onto a shadow map
      ○ Designed to address perspective-induced aliasing
        and quality discontinuity (flickers)
        ○ Biases shadow map so quality increases closer to the
          camera
     “Anti-aliasing and Continuity with
      Trapezoidal Shadow Maps”, Martin, Tobias
      and Tan, Tiow-Seng, School of Computing,
      National University of Singapore
      ○ http://www.comp.nus.edu.sg/~tants/tsm/tsm.pdf
Shadow Map Variants
   Light-Space Perspective Shadow Mapping
     LiSPSM
      ○ Modification to the PSM algorithm
        ○ A bit of a mouthful
      ○ Designed to address artifacts inherent in the PSM algorithm
        ○ Perspective distortion, missed shadow casters, singularities in post-
          projective space
      ○ Perspective transform specified in light-space
        ○ Allow for treating all lights as directional while still maintaining perspective
          benefits
      ○ Far more complicated than regular shadow maps
        ○ Benefits outweigh time cost of implementation
        ○ General opinion mixed
    ○ “Light Space Perspective Shadow Maps”, Wimmer et. al,
      Vienna University of Technology, Austria, 2005
      ○ http://www.cg.tuwien.ac.at/research/vr/lispsm/shadows_egsr2004
         _revised.pdf
Shadow Map Variants
   Parallel-Split Shadow Maps
     PSSM
       Splits the view frustum into parts using plane parallel to the
         viewing plane
            Generates smaller shadow maps for split parts
       Tighter bounds on each map takes better advantage of texture
         resolution
            Results in crisper shadows at lower cost
       Widely implemented
         Lots of available example code
         Current shadow map fad
       Very useful for large, outdoor environments
         Makes good use of orthographic cameras
     “Parallel-Split Shadow Maps for Large-Scale Virtual
      Environments”, Zhang et. al., Chinese University of Hong
      Kong, 2006
       http://appsrv.cse.cuhk.edu.hk/~fzhang/pssm_vrcia/shadow_vrcia.
         pdf
Shadow Map Variants
Shadow Map Variants
   Variance Shadow Maps
     VSM
       Developed by a fellow UW alum and IGDA Toronto member
             Andrew Lauritzen, University of Waterloo student
         Presented at i3D 2006 then GDC 2007
             Taking the industry by storm
         Designed to solve aliasing issues by using calculated variance to
          determine an upper-bound on the fraction of an occluded, shaded
          fragment
             Store mean and mean2 of a range of depths
             Allows computation of variance
         Relatively simple to compute on modern GPUs
             A bit more overhead, but well worth the cost
         Allows for mipmapping and texture filtering
             Generally uncomputable for SSM because of single depth value
         Caveats
             Light bleeding on filter areas with high variance
     “Variance Shadow Maps”, Donnelly, William and Lauritzen, Andrew,
      CGL, University of Waterloo
         http://www.punkuser.net/vsm/vsm_paper.pdf
Shadow Map Variants
Optimizations
   Things to consider
      Texture size
      Conditional statements
       Depth comparison in shader
      Texture filtering
      Combining of algorithms
       PSSM + VSM particularly interesting for large
        outdoor scenes
      Selecting the right algorithm for the scene
       Unified solution does not exist
Soft Shadows
   Large Topic
     Could spend another two hours on various
      techniques
   Various Techniques
     Percentage Closer Filtering
     Penumbra Calculations
     Screen-Space Blur
     Combinations…
   Will demo PCF with Screen-Space Blur
Further Reading
   “Casting Curved Shadows on Curved Surface”,
    Williams, Lance, CGL, New York Institute of
    Technology, New York, 1978
     The original introduction of shadow mapping (yes it’s that
       old)
   "Shadow mapping." Wikipedia, The Free
    Encyclopedia. 1 Sep 2007, 09:57 UTC. Wikimedia
    Foundation, Inc. 29 Sep 2007
    <http://en.wikipedia.org/w/index.php?title=Shadow_
    mapping&oldid=154989894>.
   “Shadow Mapping Tutorial”, Paul’s Projects,
    http://www.paulsprojects.net/tutorials/smt/smt.html
   “Hardware Shadow Mapping”, NVIDIA Corporation
     http://developer.nvidia.com/attach/8456
Questions?
Thank You
   Contact Information
     Email: jspillane@frozennorth.net
     Office: (519) 513-2409
     Website: www.frozennorth.net

				
DOCUMENT INFO
Shared By:
Categories:
Stats:
views:2035
posted:12/4/2009
language:English
pages:42