virtual textures
play

Virtual Textures Dealing with Enormous Texture-Based Resources - PowerPoint PPT Presentation

Feb 13, 2023 •170 likes •966 views

Virtual Textures Dealing with Enormous Texture-Based Resources Anders Scheel Nielsen 1 Anders Scheel Nielsen Textures 2 Anders Scheel Nielsen Textures 3 Anders Scheel Nielsen Mipmaps 4 Anders Scheel Nielsen Mipmaps 5 Anders Scheel

  1. Virtual Textures Dealing with Enormous Texture-Based Resources Anders Scheel Nielsen 1 Anders Scheel Nielsen

  2. Textures 2 Anders Scheel Nielsen

  3. Textures 3 Anders Scheel Nielsen

  4. Mipmaps 4 Anders Scheel Nielsen

  5. Mipmaps 5 Anders Scheel Nielsen

  6. Real life problems Problems: • Limited amount of resources (avg. 256 MB) • Large variation in system resources • Maximum texture sizes from: fra 2048 x 2048 til 8192 x 8192 • Atomic Resource 6 Anders Scheel Nielsen

  7. Real life problems Problems: • Limited amount of resources (avg. 256 MB) • Large variation in system resources • Maximum texture sizes from: fra 2048 x 2048 til 8192 x 8192 • Atomic Resource 6 Anders Scheel Nielsen

  8. Real life problems Problems: • Limited amount of resources (avg. 256 MB) • Large variation in system resources • Maximum texture sizes from: fra 2048 x 2048 til 8192 x 8192 • Atomic Resource 6 Anders Scheel Nielsen

  9. Real life problems Problems: • Limited amount of resources (avg. 256 MB) • Large variation in system resources • Maximum texture sizes from: fra 2048 x 2048 til 8192 x 8192 • Atomic Resource Solution for ’large’ models: • Repeating textures • Blending 6 Anders Scheel Nielsen

  10. Repeating Textures 7 Anders Scheel Nielsen

  11. Repeating textures 8 Anders Scheel Nielsen

  12. Texture blending 9 Anders Scheel Nielsen

  13. Static handling of Texture budget Using too much memory? • Permanently reduce the texture resolution • Load lower mip map levels 10 Anders Scheel Nielsen

  14. Static handling of Texture budget Using too much memory? • Permanently reduce the texture resolution • Load lower mip map levels 10 Anders Scheel Nielsen

  15. Static handling of Texture budget Using too much memory? • Permanently reduce the texture resolution • Load lower mip map levels 10 Anders Scheel Nielsen

  16. Static handling of Texture budget Using too much memory? • Permanently reduce the texture resolution • Load lower mip map levels 10 Anders Scheel Nielsen

  17. Observations 11 Anders Scheel Nielsen

  18. Observations Large Textures • All ’points’ in a surface can be truly unique • Artists can add all the details to the texture they want 11 Anders Scheel Nielsen

  19. Observations Large Textures • All ’points’ in a surface can be truly unique • Artists can add all the details to the texture they want Texture loading with higher granularity • Load only visible parts of the texture 11 Anders Scheel Nielsen

  20. Observations Large Textures • All ’points’ in a surface can be truly unique • Artists can add all the details to the texture they want Texture loading with higher granularity • Load only visible parts of the texture Load Textures Runtime: • Artists/developers: No headaches from keeping texture budgets! • Smoother scaling for different hardware configurations • Artists can (in theory) work in arbitrary resolution 11 Anders Scheel Nielsen

  21. The Method 12 Anders Scheel Nielsen

  22. Basic Data Structures 13 Anders Scheel Nielsen

  23. Flow 14 Anders Scheel Nielsen

  24. Calculation of visible tiles Readback of the framebuffer: 15 Anders Scheel Nielsen

  25. Barrett’s Fragment shader 1 float mipmapLevel = calcMipmapLevel(uv * textureSize); 2 float2 tilePos = floor(uv * indirectionSize); 3 tilePos /= 256.0; 4 float2 tilePos_low = frac(tilePos); 5 float2 tilePos_high = floor(tilePos); 6 color.bg = floor(tilePos_low * 256.0 + 0.5)/255.0; 7 color.r = (tilePos_high.x + tilePos_high.y * 16)/255.0; 8 color.a = (mipmapLevel + textureId * 16)/255.0; 9 10 float calcTextureMipmapLevel(float2 uv) { 11 float2 dtdx = dFdx(uv); 12 float2 dtdy = dFdy(uv); 13 float2 dtex = dtdx*dtdx + dtdy*dtdy; 14 float minDelta = max(dtex.x,dtex.y); 15 float miplevel = max(0.5 * log2(minDelta), 0.0); 16 return miplevel; 17 } 16 Anders Scheel Nielsen

  26. Barrett’s Fragment shader 1 float mipmapLevel = calcMipmapLevel(uv * textureSize); 2 float2 tilePos = floor(uv * indirectionSize); 3 tilePos /= 256.0; 4 float2 tilePos_low = frac(tilePos); 5 float2 tilePos_high = floor(tilePos); 6 color.bg = floor(tilePos_low * 256.0 + 0.5)/255.0; 7 color.r = (tilePos_high.x + tilePos_high.y * 16)/255.0; 8 color.a = (mipmapLevel + textureId * 16)/255.0; 9 10 float calcTextureMipmapLevel(float2 uv) { 11 float2 dtdx = dFdx(uv); 12 float2 dtdy = dFdy(uv); 13 float2 dtex = dtdx*dtdx + dtdy*dtdy; 14 float minDelta = max(dtex.x,dtex.y); 15 float miplevel = max(0.5 * log2(minDelta), 0.0); 16 return miplevel; 17 } 17 Anders Scheel Nielsen

  27. TD-texture 18 Anders Scheel Nielsen

  28. TD-texture Texture usage: • 5 texture types • 4096 x 4096 • Usage: 80 MB 18 Anders Scheel Nielsen

  29. TD-texture Texture usage: • 5 texture types • 4096 x 4096 • Usage: 80 MB 18 Anders Scheel Nielsen

  30. TD-texture Texture usage: • 5 texture types • 4096 x 4096 • Usage: 80 MB My fragment shader: 1 color = tex2D(tdTexture, vertexIn.uv * coordScale); 2 color.w = tex2Dbias(tdTexture, vertexIn.uv, mipBias + resolutionBias).w; 18 Anders Scheel Nielsen

  31. Tilbagelæsning Tilbagelæsning fra grafikkort til systemhukommelse 19 Anders Scheel Nielsen

  32. Which tiles to load first? • Screen area prioritized • Based on error metric: •Mip map level 20 Anders Scheel Nielsen

  33. Which tiles to discard? • First In First Out • Least Frequently Used • Least Recently Used • Weighted Least Recently Used 21 Anders Scheel Nielsen

  34. Updating the Indirection Map 22 Anders Scheel Nielsen

  35. Updating the Indirection Map 23 Anders Scheel Nielsen

  36. Rendering 24 Anders Scheel Nielsen

  37. Rendering 25 Anders Scheel Nielsen

  38. Visual Results Screenshots 26 Anders Scheel Nielsen

  39. Resultater Resultater 27 Resultater Anders Scheel Nielsen

  40. Diskussion Diskussion 28 Diskussion Anders Scheel Nielsen

  41. Resultater: Tilbagelæsning 29 Anders Scheel Nielsen

  42. Resultater: Tilbagelæsning 30 Anders Scheel Nielsen

  43. Resultater: Tilbagelæsning 31 Anders Scheel Nielsen

  44. Resultater: tilbagelæsning + CPU 32 Anders Scheel Nielsen

  45. Resultater: HDD read 33 Anders Scheel Nielsen

  46. Resultater: Hvilke tiles skal indlæses? 34 Anders Scheel Nielsen

  47. Resultater: Hvilke tiles skal kasseres? 35 Anders Scheel Nielsen

  48. The demo Virtual Textures: 36 Anders Scheel Nielsen

  49. The demo Virtual Textures: • Efficient handling of giant textures: Demo: 128k x 128k, theoretical: 2M x 2M (16 TB) 36 Anders Scheel Nielsen

  50. The demo Virtual Textures: • Efficient handling of giant textures: Demo: 128k x 128k, theoretical: 2M x 2M (16 TB) • Texture usage for a 20 GB texture: 17 MB 36 Anders Scheel Nielsen

  51. The demo Virtual Textures: • Efficient handling of giant textures: Demo: 128k x 128k, theoretical: 2M x 2M (16 TB) • Texture usage for a 20 GB texture: 17 MB • Frame rate: 150-200 FPS (8800GT) 36 Anders Scheel Nielsen

  52. Demo time 37 Anders Scheel Nielsen

  53. Future work Prediction of visible tiles • Camera vector • A priori map Virtuel Terrain • Height map as a Virtual Texture Other uses than games • Browsing pictures • Google Earth’ish applications • Streaming Casual Games with Instant Game Play 38 Anders Scheel Nielsen

  54. Contributions Contributions • Calculating visible tiles at greater precision with 2 lines of code • Simple, but effective tile prediction method • New strategies for prioritizing tiles • Faster methods for critical tasks in the VT-system 39 Anders Scheel Nielsen

  55. 40 Konklusion Anders Scheel Nielsen

  56. Thank you! Andreas Bærentzen , DTU Carsten Kjær , Dalux Daniel Povlsen , Aptocore Martin Mittring , CryTek GmbH Sean Barrett Per Rasmussen Hardware provided by HwT.dk Anders.Scheel@gmail.com 40 Konklusion Anders Scheel Nielsen

  57. Media 41 Anders Scheel Nielsen

  58. Media Seek time is a killer: 10 tiles @ 256 x 256 = 650KB 41 Anders Scheel Nielsen

  59. Reading tiles from media 42 Anders Scheel Nielsen

  60. Reading tiles from media • Maximizing throughput MPixels/s: • Compression = more MPixels/s • Texture Layout on media: minimizing seek time 42 Anders Scheel Nielsen

  61. Reading tiles from media • Maximizing throughput MPixels/s: • Compression = more MPixels/s • Texture Layout on media: minimizing seek time 42 Anders Scheel Nielsen

  62. Reading tiles from media • Maximizing throughput MPixels/s: • Compression = more MPixels/s • Texture Layout on media: minimizing seek time 42 Anders Scheel Nielsen

  63. Reading tiles from media • Maximizing throughput MPixels/s: • Compression = more MPixels/s • Texture Layout on media: minimizing seek time • Invalidated by the future: SSD harddrives? 42 Anders Scheel Nielsen

  64. Diskussion 1 Ulemper ved metoden: Transparente objekter • Et ekstra prepass Antal unikke teksturer • Pakke flere teksturer i ét VT • 8 bit tile-position Kvalitetstab pga. manglende tekstur • Højere teksturkomprimering • Fjern visuel pops 43 Anders Scheel Nielsen

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

So how do we make pictures like this? textures local illuminates reflection transparency It

So how do we make pictures like this? textures local illuminates reflection transparency It

CSE6335.03 Topics in Virtual Reality Monday March 23, 2009 So how do we make pictures like this? textures local illuminates reflection transparency It is very difficult There exist techniques to do this Photon Mapping - basically solve

463 views • 15 slides
Week 9 -Wednesday What did we talk about last time? Textures Volume textures Cube

Week 9 -Wednesday What did we talk about last time? Textures Volume textures Cube

Week 9 -Wednesday What did we talk about last time? Textures Volume textures Cube maps Texture caching and compression Procedural texturing Texture animation Material mapping Alpha mapping Bump mapping refers

668 views • 42 slides
Screenspace Effects Introduction General idea: Render all data necessary into textures Render

Screenspace Effects Introduction General idea: Render all data necessary into textures Render

Screenspace Effects Introduction General idea: Render all data necessary into textures Render all data necessary into textures Process textures to calculate final image A hi Achievable Effects: bl Eff t Glow/Bloom Depth of field

1.2k views • 35 slides
Procedural Texturing By Madis Janno What do I mean by Procedural Texturing? Two possible

Procedural Texturing By Madis Janno What do I mean by Procedural Texturing? Two possible

Procedural Texturing By Madis Janno What do I mean by Procedural Texturing? Two possible things: Procedurally generating textures Procedurally applying textures Todays focus is: Applying existing textures in a way that creates new textures

1.33k views • 69 slides
Outline Texture Synthesis What is Texture? How to Synthesize? Graphcut Textures

Outline Texture Synthesis What is Texture? How to Synthesize? Graphcut Textures

Graphcut Textures: Image and Video Synthesis using Graph Cuts (Kwatra, et al) presented by Adeel Bhutta Outline Texture Synthesis What is Texture? How to Synthesize? Graphcut Textures Main Idea, Contribution Patch

588 views • 20 slides
Texture What is texture? Easy to recognize, hard to define Deterministic textures

Texture What is texture? Easy to recognize, hard to define Deterministic textures

Texture What is texture? Easy to recognize, hard to define Deterministic textures (thing-like) Stochastic textures (stuff-like) Tasks Discrimination / Segmentation Classification Texture synthesis

555 views • 34 slides
Computer Graphics - Texturing - Philipp Slusallek Texture Textures modify the input for

Computer Graphics - Texturing - Philipp Slusallek Texture Textures modify the input for

Computer Graphics - Texturing - Philipp Slusallek Texture Textures modify the input for shading computations Either via (painted) images textures or procedural functions Example texture maps for Reflectance, normals, shadow

854 views • 69 slides
3: textures 1. Quest for Visual Realism Need for fine details in color variation! 1 Motivations

3: textures 1. Quest for Visual Realism Need for fine details in color variation! 1 Motivations

3: textures 1. Quest for Visual Realism Need for fine details in color variation! 1 Motivations 2. Limit the number of polygons Problem: Everything cannot be modeled at the scale of geometry! Micro-polygons would be needed! 2 Textures

674 views • 50 slides
Final exam effects Final exam effects Lighting Neon lights Textures / refelction

Final exam effects Final exam effects Lighting Neon lights Textures / refelction

Announcement Lots of CS Scholarship opportunities Textures I Deadline: April 15, 2008. Final exam effects Final exam effects Lighting Neon lights Textures / refelction Strawberry Snapple bottle Water surface CD

283 views • 13 slides
1D: Midpoint Displacement 2D: Diamond-Square Particle Systems Particle System Examples

1D: Midpoint Displacement 2D: Diamond-Square Particle Systems Particle System Examples

University of British Columbia Procedural Textures Procedural Modeling CPSC 314 Computer Graphics textures, geometry generate image on the fly, instead of Jan-Apr 2016 loading from disk nonprocedural: explicitly stored in

290 views • 3 slides
Textures III, Procedural Approaches Week 10, Mon Mar 19

Textures III, Procedural Approaches Week 10, Mon Mar 19

University of British Columbia CPSC 314 Computer Graphics Jan-Apr 2007 Tamara Munzner Textures III, Procedural Approaches Week 10, Mon Mar 19 http://www.ugrad.cs.ubc.ca/~cs314/Vjan2007 Reading for Last Time and Today FCG Chap 11 Texture

609 views • 49 slides
Final Exam effects Ground rules Textures II Teams of 2/3 GRAD students: responsible

Final Exam effects Ground rules Textures II Teams of 2/3 GRAD students: responsible

Final Exam effects Ground rules Textures II Teams of 2/3 GRAD students: responsible for research Team: responsible for implementation and documentation Procedural Textures Implement in GLSL and RenderMan (for teams of 3)

308 views • 12 slides
OpenGL 3.x Part 2: Textures and Objects Ingo Radax, Gnther Voglsam Institute of Computer

OpenGL 3.x Part 2: Textures and Objects Ingo Radax, Gnther Voglsam Institute of Computer

OpenGL 3.x Part 2: Textures and Objects Ingo Radax, Gnther Voglsam Institute of Computer Graphics and Algorithms Vienna University of Technology Topics for today OpenGL 3.x Part 1 - Revisited Textures Framebuffer Objects Vertexbuffer

1.39k views • 69 slides
Lecture 07: Buffers and Textures CSE 40166 Computer Graphics Peter Bui University of Notre Dame,

Lecture 07: Buffers and Textures CSE 40166 Computer Graphics Peter Bui University of Notre Dame,

Lecture 07: Buffers and Textures CSE 40166 Computer Graphics Peter Bui University of Notre Dame, IN, USA October 26, 2010 OpenGL Pipeline Todays Focus Pixel Buffers : read and write image data to and from OpenGL buffers. Textures :

520 views • 20 slides
FOR VIRTUAL MATERIAL DESIGN Adib Akl 1,2 , Charles Yaacoub 2 , Marc Donias 1 , Jean-Pierre Da Costa

FOR VIRTUAL MATERIAL DESIGN Adib Akl 1,2 , Charles Yaacoub 2 , Marc Donias 1 , Jean-Pierre Da Costa

30/10/2014 IEEE International Conference on Image Processing STRUCTURE TENSOR BASED SYNTHESIS OF DIRECTIONAL TEXTURES FOR VIRTUAL MATERIAL DESIGN Adib Akl 1,2 , Charles Yaacoub 2 , Marc Donias 1 , Jean-Pierre Da Costa 1 , Christian Germain 1 1

459 views • 15 slides
Surfaces How to carry to surface? Texture Synthesis: Surfaces and RD from Wei & Levoy = +

Surfaces How to carry to surface? Texture Synthesis: Surfaces and RD from Wei & Levoy = +

Surfaces How to carry to surface? Texture Synthesis: Surfaces and RD from Wei & Levoy = + Textures Textures Images surface given as mesh synthesis into textures synthesis onto vertices (fine enough) CS 176 Winter 2011 CS

1.24k views • 4 slides
Application of Satellite and Ozonesonde Data to the Study of Nighttime Tropospheric Ozone Impacts

Application of Satellite and Ozonesonde Data to the Study of Nighttime Tropospheric Ozone Impacts

Application of Satellite and Ozonesonde Data to the Study of Nighttime Tropospheric Ozone Impacts and Relationship to Air Quality Greg Osterman Jet Propulsion Laboratory CMAS Meeting - October 25, 2011 Investigators AURA 2010 Proposal

534 views • 27 slides
MIGRAINE MANAGEMENT Joanna G. Katzman, M.D.,M.S.P.H. Associate Professor, Department of

MIGRAINE MANAGEMENT Joanna G. Katzman, M.D.,M.S.P.H. Associate Professor, Department of

MIGRAINE MANAGEMENT Joanna G. Katzman, M.D.,M.S.P.H. Associate Professor, Department of Neurosurgery UNM Pain Center and Project ECHO Pain University of New Mexico Disclosure The presenter has no financial relationship to this program.

360 views • 25 slides
Chrome OS Internals Josh Triplett josh@joshtriplett.org LinuxCon Europe 2014 Josh Triplett

Chrome OS Internals Josh Triplett josh@joshtriplett.org LinuxCon Europe 2014 Josh Triplett

Chrome OS Internals Josh Triplett josh@joshtriplett.org LinuxCon Europe 2014 Josh Triplett Chrome OS Internals LinuxCon Europe 2014 1 / 43 Overview Intro to Chrome OS Architecture of Chrome OS Verified boot and developer mode Security

1.39k views • 110 slides
Debian Hamradio Blend Iain R. Learmonth < irl@debian.org > Debian Hamradio Maintainers 31st

Debian Hamradio Blend Iain R. Learmonth < irl@debian.org > Debian Hamradio Maintainers 31st

Debian Hamradio Blend Iain R. Learmonth < irl@debian.org > Debian Hamradio Maintainers 31st April 2016 Iain R. Learmonth (Debian Hams) Debian Hamradio Blend 31st April 2016 1 / 13 Pure Blends A collection of tasks (metapackages) Iain

783 views • 13 slides
A Parallel Interest Matching Algorithm for Distributed- Memory Systems Elvis Liu Georgios

A Parallel Interest Matching Algorithm for Distributed- Memory Systems Elvis Liu Georgios

A Parallel Interest Matching Algorithm for Distributed- Memory Systems Elvis Liu Georgios Theodoropoulos Outline Introduction Distributed Virtual Environments (DVE) Interest Management Parallel Interest Matching Algorithm

526 views • 28 slides
Connecting IoT appliances securely to the cloud (eap-noob) Tuomas Aura, Aalto University,

Connecting IoT appliances securely to the cloud (eap-noob) Tuomas Aura, Aalto University,

Connecting IoT appliances securely to the cloud (eap-noob) Tuomas Aura, Aalto University, Finland joint work with Mohit Sethi, Ericsson, and others Connecting devices to cloud Cloud IoT appliances Internet User Authenticated key

528 views • 17 slides
Photograph We keep this love in a photograph We made these memories for ourselves Where our eyes

Photograph We keep this love in a photograph We made these memories for ourselves Where our eyes

Photograph We keep this love in a photograph We made these memories for ourselves Where our eyes are never closing Hearts are never broken And time's forever frozen still Photographs are a way of seeing. Photographs are also a way of

598 views • 15 slides
Service Management for Smart Spaces (Intermediate T alk) Milad Khanibeik Supervisor: Prof.

Service Management for Smart Spaces (Intermediate T alk) Milad Khanibeik Supervisor: Prof.

A Master Thesis on: Service Management for Smart Spaces (Intermediate T alk) Milad Khanibeik Supervisor: Prof. Dr.-Ing. Georg Carle Advisor: Marc-Oliver Pahl, Dipl.-Inform. April 2014 1 Structure of the Presentation Smart Space

372 views • 22 slides

More recommend