paper summaries
play

Paper Summaries Any takers? The Rendering Equation Assignments - PDF document

Paper Summaries Any takers? The Rendering Equation Assignments Projects Checkpoint 3 Project feedback Graded email sent Approx 18 projects Checkpoint 4 Listing of projects now on Web Due Wednesday


  1. Paper Summaries • Any takers? The Rendering Equation Assignments Projects • Checkpoint 3 • Project feedback – Graded – email sent • Approx 18 projects • Checkpoint 4 • Listing of projects now on Web – Due Wednesday • Presentation schedule – Inverse mapping in 2nd half – Presentations (15 min max) • Checkpoint 5 – Last 3 classes (week 10 + finals week) – To be given Wednesday – Sign up • RenderMan • Email me with 1 st , 2 nd , 3 rd choices – Due February 16th • First come first served. – License server problems Today’s Class Computer Graphics as Virtual Photography • The Rendering Equation real camera photo Photographic Photography: scene (captures processing print – What it is light) – Techniques for solving processing camera Computer 3D synthetic tone model Graphics: models reproduction image (focuses simulated lighting) 1

  2. The Rendering Equation The Rendering Equation • General form [ ] ∫ • Kajiya: 1986 ′ = ′ ε ′ + ρ ′ ′ ′ ′ ′ ′ ′ ′ ( , ) ( , ) ( , ) ( , , ) ( , ) I x x g x x x x x x x I x x d x S • “Unified context for viewing rendering algorithms as more or less accurate approximations to the solution of a single equation” • Expresses the quantity of light transferred from one point x’ to another x , summed over all points. Ashdown The Rendering Equation The Rendering Equation • Transport Intensity • Geometry term [ ] [ ] ∫ ∫ ′ = ′ ε ′ + ρ ′ ′ ′ ′ ′ ′ ′ ′ ′ = ′ ε ′ + ρ ′ ′ ′ ′ ′ ′ ′ ′ I ( x , x ) g ( x , x ) ( x , x ) ( x , x , x ) I ( x , x ) d x I ( x , x ) g ( x , x ) ( x , x ) ( x , x , x ) I ( x , x ) d x S S I(x, x’) = Transport energy or intensity of light g(x, x’) = geometry term passing from point x’ to point x (unoccluded two = 0, if x is not visible from x’ point transport) = 1/d 2 otherwise The Rendering Equation The Rendering Equation • Emittance • Scattering [ ] ∫ [ ] ′ ′ ′ ′ ′ ′ ′ ′ ′ ′ ′ = ε + ρ I ( x , x ) g ( x , x ) ( x , x ) ( x , x , x ) I ( x , x ) d x ∫ ′ ′ ′ ′ ′ ′ ′ ′ ′ ′ ′ = ε + ρ S I ( x , x ) g ( x , x ) ( x , x ) ( x , x , x ) I ( x , x ) d x S ρ (x, x ′ , x ′′ ) = light energy reflected off ε (x, x ′ ) = light energy emitted from point x ′ point x’ towards point x from light coming from x ′′ towards x . 2

  3. The Rendering Equation The Rendering Equation • Incoming = direct + indirect (scattered) [ ] ∫ ′ = ′ ε ′ + ρ ′ ′ ′ ′ ′ ′ ′ ′ I ( x , x ) g ( x , x ) ( x , x ) ( x , x , x ) I ( x , x ) d x S = ε + ( I ) I g gR direct indirect ρ (x, x’, x’’) = light energy reflected from point x’ towards point x from light coming from x’’, i.e. BRDF The Rendering Equation The Rendering Equation • In short… – The transport of light from point x’ to point x is equal to the sum of • the light emitted from x’ in the direction of x and • the total light scattered from x’ towards x due to light from all other surfaces in the scene. The Rendering Equation The Rendering Equation • This can be expanded using the Neuman series for • Why is this important? implementation purposes: = ε + ε + ε + ε + … 2 3 I g g ( Rg ) g ( Rg ) g ( Rg ) = ε + ε + ε + ε + … 2 3 I g g ( Rg ) g ( Rg ) g ( Rg ) direct 1st scattering 2nd scattering 3rd scattering or Rendering methods can be ∞ ∑ = ε characterized by the number of i I g ( Rg ) scatterings considered = i 0 3

  4. The Rendering Equation The Rendering Equation • Local vs Global Illumination Models • Local Illumination = ε + ε + ε + ε + … = ε 2 3 I g g ( Rg ) g ( Rg ) g ( Rg ) I g direct 1st scattering 2nd scattering 3rd scattering direct Local illumination - only considers direct component Only object’s first contact with light is considered. Lighting “simulated” by illumination model used. Global illumination - also considers other scattered component NOTE: Kajiya does not include ambient light! The Rendering Equation The Rendering Equation • Global Illumination • Drawbacks = ε + ε + ε + ε + … 2 3 – Uses geometric optics, based on light as rays I g g ( Rg ) g ( Rg ) g ( Rg ) – Phase, diffraction, and transmission through direct 1st scattering 2nd scattering 3rd scattering participatory media not considered (i.e., only homogenous refraction considered) Considers multiple scatterings • Ray-Tracing – Dependence on wavelength is implied • Radiosity – Not expressed using physical units • Kajiya’s method • Question: Isn’t I (x, x’) simply radiance at a point? • Yes! (with the exception of some geometric terms…) • More when we talk about radiosity The Rendering Equation Rendering Methods • Ray Tracing • Summary – Light rays are traced from the eye, through a – Equation for describing rendering algorithms. viewing plane, into scene – Describes light arriving at a point from another – When rays strike an object, further rays are point (and indirectly all other points) spawned representing reflection and refraction. – Considers direct light and recursive scattering – These newly spawned rays can strike objects and spawn more rays, etc. 4

  5. Rendering Methods Rendering Methods • Ray Tracing • Ray tracing = ε + ε + ε + ε + … 2 3 I g g ( Rg ) g ( Rg ) g ( Rg ) direct 1st scattering 2nd scattering 3rd scattering Number of scatterings depend on recursion depth allowed by ray tracer Watt/Watt Rendering Method Rendering Methods • Ray Tracing – Examples • Radiosity – The problem with ray tracing • Great for specular type reflections • Awful for diffuse reflections. – Radiosity • From the guys who brought you Cook-Torrance illumination model! • Not quite as elegant as ray tracing • More physically based • video Rendering Methods Rendering Methods • Radiosity • Radiosity – Radiant exitance - radiant flux out – Based on the theory of heat transfer – Calculate lighting in a steady state Φ d = – Assumes that all surfaces are perfectly diffuse M – Formulates a large systems of linear equations dA – Solution of these equations give you radiant dA exitance at each point 5

  6. Rendering Methods Rendering Methods • Radiosity • Radiosity – Image is created by using calculated radiant – View dependence vs view independence existance values in standard rendering process. • Radiosity provides a view independent solution • In essence, radiosity defines a “made to fit” texture • Scene needs to be further rendered from a given mapping view point. Rendering Methods Rendering Methods • Radiosity • Radiosity – Patches – Not points -- But patches • Scene is subdivided into patches • Radiant exitance will be calculated for each patch [Ashdown94] Radiosity - Basics Radiosity - Basics • Key idea • Basic idea – Since all objects are perfectly diffuse, we can – Each patch will receive a certain amount of determine where light is coming from (and light from the environment going) by simply considering the geometry of – It will reflect fraction back into the the scene. environment – Keep track of amount of light reflected back – Calculation of radiant exitance per patch is can – Continue till all light has been exhasted. be detemined mathematically using a closed form solution. 6

  7. Rendering Methods Rendering Methods • Radiosity • Radiosity – Examples = ε + ε + ε + ε + … 2 3 I g g ( Rg ) g ( Rg ) g ( Rg ) direct 1st scattering 2nd scattering 3rd scattering Mathematical derivation assumes limit as number of scatterings goes to Infinity • Bunny (Blue Sky Studios) Remember BRDFs? Rendering Methods • Getting closer to reality – Ray tracing • Replace rays with cones/beams • Distributed ray tracing – Spawn more rays – Stochastic sampling • Supplement backward ray tracing with forward ray Perfectly specular Perfectly diffuse Specular & diffuse tracing (e.g. Photon Mapping) Ray tracing Radiosity Reality Rendering Methods Rendering Methods • Getting closer to reality • Kajiya’s solution (Path tracing) – Much like ray tracing – Radiosity – Determine the path of light that eventually reaches the • Two path method – combine ray tracing and eye. radiosity – Only 1 ray spawned per intersection • Reflection in specular • Reflection in diffuse • Refraction – Which ray to spawn determined via stochastic sampling. 7

  8. Kajiya’s Method Summary • Rendering equation – Mathematical expression of rendering process • Solutions – Ray Tracing – Radiosity – Combination of both. • Questions? 8

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend