Modular Level Design for Skyrim
161 likes63,869 views
The document discusses modular level design in video games, particularly in the context of Bethesda's Skyrim. It highlights the importance of reusable art kits for efficient level creation and the balance between complexity and functionality in artistic design. Various phases of kit development, from concept to ongoing polish, are outlined along with challenges and benefits related to collaboration between art and design teams.
Modular Level Design for Skyrim
- 1. Modular Level Design for Skyrim Joel Burgess Nathan Purkeypile Bethesda Game Studios
- 2. Where We’re Coming From ●Bethesda Games are Big
- 4. Where We’re Coming From ●Bethesda Games are Big ●We value high-mileage
- 6. Where We’re Coming From ●Bethesda Games are Big ●We value high-mileage ● In our systems, art, team - everything ●Beliefs + Circumstance + Experience = Culture ●How can our culture benefit yours?
- 7. Case Study: Modular Level Design
- 8. What is a kit? ●Kits are Systems
- 9. What is a kit? ●Kits are Systems ●Typically snapped to a grid ●More than the sum of parts
- 11. What is a kit? ●Kits are Systems ●Typically snapped to a grid ●More than the sum of parts ●In this context: architectural
- 15. “Modular Level and Component Design” Lee Perry – Game Developer Nov. 2002
- 16. Terminator: Future Shock (1995) Elder Scrolls 2: Daggerfall (1996)
- 18. Modular Level Design: The Pros & Cons
- 19. Pro - Reusable Art is… Reusable ●Reusable Art Mitigates Scope
- 21. Pro - Reusable Art is… Reusable ●Reusable Art Mitigates Scope ●Skyrim Environment Includes: ●16 sq.mile Overworld ● 300+ Dungeons ●5 Major Cities ● 140+ Points of Interest ●2 Hidden Worldspaces ● 37 Towns, Farms & Villages
- 23. Pro - Reusable Art is… Reusable ●Reusable Art Mitigates Scope ●Skyrim Environment Includes: ●16 sq.mile Overworld ● 300+ Dungeons ●5 Major Cities ● 140+ Points of Interest ●2 Hidden Worldspaces ● 37 Towns, Farms & Villages ●400+ cells of traditional “level” gameplay
- 25. Pro - Reusable Art Mitigates Scope ●Skyrim: 400+ cells, 8 Level Designers ●Scope as Restrictive?
- 27. Beliefs + Circumstance + Experience . Culture
- 28. Con - Art Fatigue ●Abolish copy/pasting
- 32. Con - Art Fatigue ●Abolish copy/pasting ●Disassociate setting from inhabitants & gameplay ●Encourage Asset Mix-n-Match
- 36. Pro - Low Artist:Designer Ratio TESV: Skyrim 2 Kit Artists 8 Level Designers 7 Kits 400+ Unique Cells 30 Months
- 37. Con - High Complexity ●Requires Art Talent & Technical Understanding Aesthetics Composition Design Art Workflow Workflow Modeling Kit Logic Snapping Rules Grid Multiples
- 39. Con - High Complexity ●Requires Art Talent & Technical Understanding ● Basically: Unicorn Artist ●Bug-Fixing can be delicate
- 40. Pro – Instant, Game-wide Art Deployment ●Art changes propagate automatically
- 41. New Art Incoming! Yay! kthx. Whatever.
- 42. Pro – Instant, Game-wide Art Deployment ●Art changes propagate automatically ●Artwork viewable in “real” context ●Zero impact on Design workflow ●Keeps Cosmetic Control w/Art ●Aesthetic Process Not Rushed
- 43. Pro - Iteration Speed ●High flexibility & agility ●Fastest Workflow Ever? ●No Turnaround Time ●1:1 Correlation w/Final Layout
- 44. Con - Total Dependence on Art ●Design empowered only if kits available and good ●Art/Design Relationship is Imperative
- 45. Common LD – Art Relationship ● Level Designers Build early version of level ● In-Editor: BSP, Graybox ● External Tools: Sketchup, 3ds Max, etc ● Abstract: Documents, Paper Maps ● Then: Hand off to art
- 46. LEVEL DESIGN
- 47. Common LD – Art Relationship ● Art does a visual pass on the level ● Sends back to Design for markup, scripting, etc
- 48. ART
- 49. Art & Design Collaboration ● Can we improve relations? ● Kits May seem impersonal ● Process is collaborative in dev
- 50. How Kits Are Born
- 52. Main Steps For Kit Building ●Concept ●Proof Piece ●Graybox/Functional Core ●Build-Out ●Ongoing Polish
- 53. A Note on Scale ●Abstract Units (Commonplace) ●128 units ~ 6 feet ~ 2 meter ●This talk assumes Z as “up” axis 1 2 8 32
- 54. Global Standards ●Establish Global Standards ● Doorframe Dimensions ● Minimum Width ● Valid Incline Grading ● Gameplay-relevant metrics ●Jump heights ●Cover heights
- 55. Concept Phase
- 56. Where Do Kits Come From? ●Concept ● Visual Theme, Gameplay Goals, Scope & Sub-Kits
- 57. Concept Phase ~1 week } Level Designer Concept Kit Artist Concept
- 58. Before You Begin ●ArtConcerns: ●Visual theme? Visual goals? ●DesignConcerns: ●Expected usage? Gameplay role of kit? Story?
- 59. Before You Begin ●How widely-used? Skyrim Kit Usage Cave Mines Ice Cave Nordic Fort Dwarven Ratway
- 60. Skyrim Kit Usage Cave Mines Ice Cave Nordic Fort Dwarven Ratway
- 61. Before You Begin ●What “Sub-kits” will be made? Cave – Used 200 times Ratway – Used Twice Nordic – Used 150 times Small Hall Hall Small Hall Large Hall Circular Room Large Hall Small Room Square Room Small Room Large Room Large Room Epic Room Temple Room Shaft Stairwell Balcony
- 62. Proof Phase
- 63. Where Do Kits Come From? ●Concept ● Visual Theme, Gameplay Goals, Scope, Sub-Kits, etc ●Proof ● Build out prototype pieces, test major concepts
- 64. Concept Proof Phase Phase 1-3 weeks } Level Designer Concept Stress Test Kit Artist Concept Proof
- 65. Pick Your Grid Footprint ●Grid Size is the Foundation of your Kit
- 66. Footprint Sizes ●Equilateral Sizes are common: ●512x512x512
- 67. 2 5 6 256
- 68. Footprint Sizes ●Equilateral Sizes are common: ●512x512x512 ●Play with Proportions ●256x256x768
- 69. 7 6 8 256
- 70. Footprint Sizes ● Equilateral Sizes are common: ● 512x512x512 ● Play with Proportions ● 256x256x1024 ● Sub-Kits on same grid size or multiples of it. ● 256X256x256 can match 512x512x512 ● 256x256x256 cannot match 384x384x384
- 71. Keep Your Grid Snaps Large (unhappy LD) (happy LD) Grid Snap Size
- 72. Common Footprint Problems ●Non-uniform XY sizes can be problematic Kit Kit Kit Piece Piece Piece Kit Piece Piece Piece Piece Piece Piece Piece Kit Kit Kit Kit Kit Kit Kit Piece
- 73. Common Footprint Problems ●Non-uniform XY sizes – use sparingly. ●Don't tile on all six axes. ●Huge Increase in Complexity
- 74. Understanding Footprint ●Footprint = full bounds of a piece ●Including Negative space ●Pieces should exist within the footprint ● Footprintis not the traversable space ● Avoid temptation to build outside footprint
- 75. Stay Inside the Footprint
- 76. Stay Inside the Footprint Outside Footprint!
- 77. Stay Inside the Footprint
- 78. Stay Inside the Footprint Overlapping Parts
- 79. Stay Inside the Footprint
- 81. Where Do Kits Come From? ●Concept ● Visual Theme, Gameplay Goals, Scope, Sub-Kits, etc ●Proof ● Build out one or a small number of prototype pieces ●Core Pieces ● “Graybox” versions of high-use pieces in all sub-kits
- 82. Concept Proof Graybox Phase Phase Phase 1-4 weeks Concept } Stress Test Concept Proof Core
- 84. “Graybox” Core Pieces ●Create a graybox of your primary sub-kit ●Figure out problems early on ●Focus on function, not aesthetics
- 85. Naming Conventions ●Be Consistent w/other Kits ●Balance Brevity & Meaning ●Avoid over-abbreviation ●But keep names recognizable ●Decide early - Difficult tochange ●Use “01” as a suffix to allow for variants
- 86. ● Utility Kit ● Bay Sub-Kit ● Corner (Inner) ● Mid/Tiling Piece ● WTF? Utl BayCorInMid PRTT01L01
- 87. UtlBayCorInMidRRRT01RL01 UtlBayGirderMountL01 UtlBayCorInBot01 UtlBayCorInMidRRTR01L UtlBayGirderMountR01 UtlBayCorInExSmL01 UtlBayCorInMidRRTR01R UtlBayGirderStraightReg01 UtlBayCorInExSmR01 UtlBayCorInMidRTPT01L UtlBayGirderStraightReg02 UtlBayCorInMid01 UtlBayCorInMidRTPT01R UtlBayGirderStraightSht01 UtlBayCorInMid01LR UtlBayCorInMidRTRT01L UtlBayGirderStraightSht02 UtlBayCorInMidDbl01 UtlBayCorInMidRTRT01LR UtlBayGirderStrutSingle01 UtlBayCorInMidFull01 UtlBayCorInMidRTRT01R UtlBayGirderStrutSingle02 UtlBayCorInMidPRPR01 UtlBayCorInMidRTTT01L UtlBayGirderStrutSingle03 UtlBayCorInMidPRPR01F UtlBayCorInMidRTTT01R UtlBayMidBot01 UtlBayCorInMidPRPT01FL UtlBayCorInMidTRTR01L UtlBayMidTop01 UtlBayCorInMidPRPT01FR UtlBayCorInMidTRTR01R UtlBayWallBot01 UtlBayCorInMidPRPT01L UtlBayCorInMidTRTT01L UtlBayWallExSm01 UtlBayCorInMidPRPT01R UtlBayCorInMidTRTT01R UtlBayWallMid01 UtlBayCorInMidPRTR01L01 UtlBayCorInTallBotL01 UtlBayWallMid02 UtlBayCorInMidPRTT01L01 UtlBayCorInTallBotR01 UtlBayWallMidL01 UtlBayCorInMidPRTT01R01 UtlBayCorInTallMidL01 UtlBayWallMidPR01 UtlBayCorInMidPTPT01 UtlBayCorInTallMidL02 UtlBayWallMidPR01L UtlBayCorInMidPTPT01F UtlBayCorInTallMidR01 UtlBayWallMidPR01R UtlBayCorInMidPTTT01L01 UtlBayCorInTallMidR02 UtlBayWallMidPT01 UtlBayCorInMidPTTT01R01 UtlBayCorInTallTopL01 UtlBayWallMidPT01L UtlBayCorInMidRRPR01L UtlBayCorInTallTopR01 UtlBayWallMidPT01R UtlBayCorInMidRRPR01R UtlBayCorInTop01 UtlBayWallMidR01 UtlBayCorInMidRRPT01L UtlBayCorInTopLR01 UtlBayWallMidRT01 UtlBayCorInMidRRPT01R UtlBayCorOutBot01 UtlBayWallMidRTL01 UtlBayCorInMidRRRT01L UtlBayCorOutEdge01 UtlBayWallMidRTR01 UtlBayCorInMidRRRT01LR01 UtlBayCorOutEdgeL01 UtlBayWallMidTR01 UtlBayCorInMidRRRT01R UtlBayCorOutEdgeL02 UtlBayWallTallBot01 UtlBayCorOutEdgeR01 UtlBayWallTallCorInBot01 UtlBayCorOutEdgeR02 UtlBayWallTallCorInExSmL01 UtlBayCorOutEdgeSm01 UtlBayWallTallCorInExSmR01 UtlBayCorOutMid01 UtlBayWallTallCorInMid01 UtlBayCorOutPR01 UtlBayWallTallCorInTop01 UtlBayCorOutPT01 UtlBayWallTallCorOutBot01 UtlBayCorOutTop01 UtlBayWallTallCorOutMid01 UtlBayGirder3wayReg01 UtlBayWallTallCorOutTop01 UtlBayGirder3wayReg02 UtlBayWallTallExSm01 UtlBayGirder3wayRegFree01 UtlBayWallTallMid01 UtlBayGirder3wayRegFree02 UtlBayWallTallMid02 UtlBayGirder4wayReg01 UtlBayWallTallMid03 UtlBayGirder4wayReg02 UtlBayWallTallMid04 UtlBayGirder4wayRegFree01 UtlBayWallTallMid05 UtlBayGirder4wayRegFree02 UtlBayWallTallTop01 UtlBayGirderBaseFreeWall01 UtlBayWallTop01 UtlBayGirderBaseFreeWall02 UtlBayGirderBaseReg01 UtlBayGirderBaseReg02 UtlBayGirderBaseSht01 UtlBayGirderBaseSht02
- 88. Common Naming Conventions Hallways Rooms Platform 1Way01 Corner01 Mid01 2Way01 Mid01 Corner01 3Way01 Wall01 ICorner01 4Way01 WallDoor01 DeadEnd01 ICorner01 1WayDoorL01 CornerDoorL01 1WayDoorR01 CornerDoorR01 1WayDoorB01 CornerDoorB01 DeadEndDoor01
- 89. Pivot Placement ● Consider designer workflow & usage of the kit ● Pivots at bottom-center are most common ● Exceptions exist, such as: ● Platform Kit (bottom of platform) ● Pipe Kit (hinge pivot) ● Alcove (edge of alcove) Pivot
- 90. Sidebar: Stress Testing & LD Role
- 91. Stress Testing Kits ● Don’t test in ideal conditions – this proves nothing ● Consider all possible use cases, likely and otherwise ● Then decide which to support ● Have answers for later
- 92. Common Problems: Loopback Issues • Double Kit on itself • Look for Gaps • Avoid Patch Pieces • Find Footprint Issues
- 93. Common Problems: Loopback Issues Bad! Good!
- 94. Common Problems: Unable to Stack • Uses full Vertical Footprint • Floor/Ceiling Co-Planar
- 95. Common Problems: Unable to Stack • Leave A Gap • Space for holes, gaps, partitions, triggers
- 96. Common Problems: The “Hall Room” Hall “3-way” Pieces Hall Corner Pieces Unintended Pillars!
- 97. Common Problems – The Hall Room ● Somebody tries it eventually ● Like, every time. ● Sometimes it’s ugly ● Decide – support it or no?
- 98. Stress Testing Is Teamwork
- 99. Building Out the Kit
- 100. Where Do Kits Come From? ●Concept ● Visual Theme, Gameplay Goals, Scope, Sub-Kits, etc ●Proof ● Build out one or a small number of prototype pieces ●Core Pieces ● “Graybox” versions of high-use pieces in all sub-kits ● Building Out ● Push for Functionally-Final, Begin Sharing w/Team
- 101. Concept Proof Graybox Build Out Phase Phase Phase 4-8 weeks Additional Level Designers Concept Stress Test Educate Concept Proof Core Build Out
- 102. Start Making It Real Art ●Art worries about this part - Design can start building ●Doesn't impact the flow/design of the level. ●Can be swapped out seamlessly. ●One “Visually Final” piece before building whole kit.
- 103. Building Out the Kit
- 104. Avoid Hero Pieces Which Matters more?
- 105. Helper Markers
- 106. Polish Phase
- 107. “Ongoing Polish” ●Review Usage Cases in Real Levels ●Respond to bugs, usability, special requests ●Continual visual refinement & development
- 108. Establishing The Kit Build Levels 2-5 Months Build Levels Build Levels Concept Stress Test Educate Build Levels Concept Proof Core Build Out Polish
- 109. Establish Building Real Content With Kit 1 2 3 4 Ongoing Polish
- 110. Going Off the Grid: Advanced Kit Building
- 111. What You Can Do So Far
- 112. Always Looks Like A Kit
- 114. Learn the Rules, Break the Rules ●With Fallout3, we begin bending our own rules ●Always keeping in mind the ideas behind them ● Choose and acknowledge logic quirks. ● “This one tiles vertically.” ● “These floors are placed independently.” ● “The halls in this one are curved.”
- 115. Rule-Bending: Always Remember the Cost ●Art/Design compromises ●Art –Looks better, takes longer ●LD – Harder to work with ●Worth it? Case by case
- 116. Snap-to-Reference ●Key technology ●Select any object, editor treats as origin
- 117. World Grid New Snap Reference
- 118. Snap Ref’s Grid
- 119. Snap-to Reference
- 120. Pivot and Flange ●Used for the Ratway in Skyrim ●Pivot actsas a hinge ●Allows for organic flow
- 121. Pivot and Flange - Organic Flow
- 122. Pivot and Flange - Archway
- 123. Pivot and Flange - Downsides ●LD build process is more linear ●Introduces non-obvious errors ● Be aware and communicate this to LDs ●Bad for Highly Specific Architecture
- 124. Looks fine...
- 125. …Holes!
- 126. Shell-Based Building: Skyrim Cave “Rooms”
- 127. Shell-Based Building: Skyrim Cave “Rooms” ●Organic Space + Orthogonal Kit = Problem ●Skyrim Caves used large “shells” ●Inset with tiling wall sub-kits ●Adds very few pieces
- 132. Directionally Restricted Kits ●Snap together only in certain rotations ●Some Examples: ● Unique in every direction ● “Front/Side” pieces ● Asymmetrical halls
- 133. Asymmetrical Hall A B A A
- 134. Asymmetrical Hall A B B A A B A B
- 136. Platform Kits ● Add “Planes of Play” ● Low Piece Count ● Bite-sized “LD art” ● Works with all sub-kits.
- 137. Kit Mashups ● Remember: Encourage Mix-n-Matching Kits ● “Glue Kits” – small kits made for blending ● Multiplier on visual variety options ● Requires compromise on art & design
- 140. Wrap Up
- 144. Games Are Made By People
- 145. Thanks For Listening! joel@joelburgess.com lplasma@gmail.com Additional Reading: ● Modular Level and Component Design -Lee Perry http://udn.epicgames.com/Three/rsrc/Three/ModularLevelDesign/ModularLevelDesign.pdf ● www.CreationKit.com