Spaceship Generator 3D: I finished a program that makes spaceship models in OpenGL. (Link to free download in comments)
https://redd.it/1f02f27
@proceduralgeneration
My talk about layer-based infinite procedural generation from Everything Procedural Conference 2024 is now online (and the framework is open source)
https://www.youtube.com/watch?v=4oJGkx0K8UQ
https://redd.it/1ezclrd
@proceduralgeneration
Zoning blocks in a procedural city : second article in my Worldsmith Project series!
https://zero.re/worldsmith/blockassign/
https://redd.it/1ey2aat
@proceduralgeneration
I made a meditative website that simulates looking outside of a train window. You can customize lots of settings and draw northern lights in the sky.
https://redd.it/1ew1m2s
@proceduralgeneration
Call for Papers - Special Issue JCR Q2 (Multidisciplinary Sciences) COMPLEXITY (IF: 1.7) - "Complex Systems in Aesthetics, Creativity and Arts"
Dear colleagues,
Juan Romero, Penousal Machado and Colin Johnson will publish a Special Issue associated with EvoMUSART on "Complex Systems in Aesthetics, Creativity and Arts" and it would be a pleasure if you sent an extension of your contribution.
**Journal: Complexity (ISSN 1076-2787)**
**JCR Journal with Impact factor: 1.7 (Q2)**
**Deadline for manuscript submissions: 18 October 2024**
Special Issue URL: [https://onlinelibrary.wiley.com/doi/toc/10.1155/8503.si.941484](https://onlinelibrary.wiley.com/doi/toc/10.1155/8503.si.941484)
Instructions for authors: [https://onlinelibrary.wiley.com/page/journal/8503/homepage/author-guidelines](https://onlinelibrary.wiley.com/page/journal/8503/homepage/author-guidelines)
One of the main - possibly unattainable - challenges of computational arts is to build algorithms that evaluate properties such as novelty, creativity, and aesthetic properties of artistic artifacts or representations. Approaches in this regard have often been based on information-theoretic ideas. For example, ideas relating mathematical notions of form and balance to beauty date to antiquity. In the 20th century, attempts were made to develop aesthetic measures based on the ideas of balance between order and complexity. In recent years, these ideas have been formalized into the idea that aesthetic engagement occurs when work is on the "edge of chaos," between excessive order and excessive disorder, formalizing it through notions such as the Gini coefficient and Shannon entropy, and links between cognitive theories of Bayesian brain and free energy minimization with aesthetic theories. These ideas have been used both to understand human behavior and to build creative systems.
The use of artificial intelligence and complex systems for the development of artistic systems is an exciting and relevant area of research. In recent years, there has been an enormous interest in the application of these techniques in fields such as visual art and music generation, analysis and performance, sound synthesis, architecture, video, poetry, design, game content generation, and other creative endeavors.
This Special Issue invites original research and review articles which will focus on both the use of complexity ideas and artificial intelligence methods to analyze and evaluate aesthetic properties and to drive systems that generate aesthetically appealing artifacts, including: music, sound, images, animation, design, architectural plans, choreography, poetry, text, jokes, etc.
Potential topics include but are not limited to the following:
* Computational aesthetics
* Formalising the ideas of aesthetics using ideas from entropy and information theory
* Computational creativity
* Artificial Intelligence in art, design, architecture, music, and games
* Information Theory in art, design, architecture, music, and games
* Complex systems in art, music ,and design
* Evolutionary art and music
* Deep ;learning models to art and video creation
* Artificial life in arts
* Swarm art
* Pattern recognition and aesthetics
* Cellular automata in architecture
* Generative AI
Dr. Penousal Machado
Dr. Colin Johnson
Dr. Iria Santos
Guest Editors (EvoMUSART 2025)
https://redd.it/1evbjkt
@proceduralgeneration
Seeking Advice on Improving My Procedurally Generated Office Interiors
https://redd.it/1eupx23
@proceduralgeneration
Exploration of more artist friendly proc-gen systems
https://x.com/ga5p0d3/status/1824525840421470288
https://redd.it/1etxx0d
@proceduralgeneration
Procedurally generated dungeon gameplay
https://redd.it/1escufz
@proceduralgeneration
How Can I Make My City Generator More Visually Appealing?
https://reddit.com/link/1es7dy0/video/sq0s2tqrxnid1/player
Hey everyone!
I've been working on a city generator project here, and I'm looking to make it more visually appealing. I'm a big fan of low-poly and simplistic designs, but I've been told that this style might not resonate with everyone.
Does anyone have tips or suggestions on how I can enhance the visuals? I'm open to ideas on color schemes, textures, shaders, or even specific tools and resources that could help bring the project to life.
Thanks in advance for any advice!
https://redd.it/1es7dy0
@proceduralgeneration
What is preventing game devs from creating varied terrain gen?
Hi, procgen newb here. I am a big fan of space sims that involve planetary exploration like No Man's Sky. After the release of Starfield, I have come to the conclusion that game studios either can't make planetary procgen that is something besides procedural oatmeal, or else don't think it's important enough to dedicate serious resources to. I noticed that not many people had a problem with the game's half-baked, insipid topology, meanwhile I'm over here raging at Bethesda for failing to avoid the one problem they absolutely couldn't afford to have. The days of boring planet terrain are over, and Todd clearly didn't get the memo.
So, what is preventing studios from adding extra terrain features? For instance, a mountainy planet is fine, but wouldn't it be nicer if it had a river valley? What about a natural stone archway, or a plateau; something to break up the endless battery of mountains.
How hard is that to do compared with 'industry standard' procedural oatmeal?
https://redd.it/1es5bcv
@proceduralgeneration
A Tool to Visualize 4D Rotations - The Qualia Research Institute
https://www.youtube.com/watch?v=BRVH-LMCBkA
https://redd.it/1eriwfp
@proceduralgeneration
Influence between nearby chunks
I want to share a scheme for letting the contents of a chunk depend on nearby chunks. I'll give a grid example but it should apply generally.
The goal is an algorithm that is deterministic and local (we can generate an arbitrary part of the output without having to generate the whole output). We can get both properties by generating each chunk independently. But then we are blind to what's in neighboring chunks, which can lead to problems at chunk boundaries. For example, if we spread points using Poisson disk sampling, points in neighboring chunks might be too close together:
Chunk edges are visible if we generate each chunk independently
When generating a chunk, we could take into account any nearby chunks that happen to be generated first, but doing this naively will break determinism; the ordering of chunks will depend on which parts of the output we generate. So we need a scheme to make the order deterministic.
First, assign each chunk one of n labels, such that nearby chunks always have different labels. In our example, we have a grid of chunks and we consider chunks nearby if they share a corner. We can just spread 4 different labels over each 2x2 square of chunks in a repeating pattern. If the regularity produces artifacts, we can mix it up by adding a pseudorandom multiple of 4 to each label (at the expense of increasing n).
Now suppose we want to generate a set of chunks (green below). First add chunks (blue below) so that for each included chunk, any nearby chunks with a lower label are also included. This process won't extend more than n chunks from the starting set.
The set of chunks to be generated
Finally generate the chunks from lowest label to highest label, always taking into account nearby chunks with a lower label (we've ensured those will be generated already).
The generated chunks
https://redd.it/1erjhb0
@proceduralgeneration
A way to generate a noise function from a height map?
So, i might as well ask it here:
I'm doing a minecraft mode, and thus tinkering with its world generation engine.
My problem is that the dimension I'm trying to make has a set terrain shape. Now, i know you might be asking: If you have a height map, use that heightmap!
The problem lies in that the terrain is extense af. If i were to make the heightmap big enough for it to accurately represent the terrain, the file would end up being huge. So my next train of thought is:
Is there a way to make a noise function i can plug into minecraft that more or less accurately represents the terrain depicted on the heightmap?
Is it possible to do? Computationally feasible? Thanks in advance
https://redd.it/1ercj99
@proceduralgeneration
What do you think of my map generation? I've been working on these techniques for quite a few years.
https://redd.it/1eqlnos
@proceduralgeneration
Experimenting with original algorithms (+video)
https://redd.it/1ezog90
@proceduralgeneration
Limitless worlds in real time without chunks, tiles or noise
The Forest (free) runs in browsers. From that linked page you can download a PDF about how the terrain is generated.
https://redd.it/1ew4nfv
@proceduralgeneration
Need some clarification on perlin noise for procedural mesh generation.
I'm using perlin noise for mesh generation. I'm using it to control the z position of the vertices based on their x and y world position.
I've been told that the world position x y you input to the algorithm cant be a whole integer I've also been told that the coordinates need to be between -1.0 and 1.0. And that the algorithm will output between -1.0 and 1.0.
I dont understand how you could constrain an infinite number of world positions to be between -1.0 and 1.0. It would make more sense if you could just plug the world position in and get back a float between -1.0 and 1.0.
Can someone tell me what I need to do to the x y coordinates of the vertex before inserting it into the function? All the world positions of my vertices are whole numbers (floats with no decimal).
https://redd.it/1evpp9h
@proceduralgeneration
Ways to keep a fully procedural environment interesting - these work for my city project, but would they work for other environments?
https://youtu.be/4MZ5-KQW3pc?si=F5sZUXSRs0FL-HZz
https://redd.it/1evapps
@proceduralgeneration
Snowy Triptych (1 bit ASCII) [python + c]
https://redd.it/1ev3hqn
@proceduralgeneration
City map generation using tensor fields
https://redd.it/1etyjbh
@proceduralgeneration
LLM based Terrain Generation
Large Language Models (like Chat GPT) are generative functions using probabilty to determine outupt.
For those sick of AI fluff, I know a lot of hype is surrounding AI, people who maybe haven't had much experience coding without AI tend to claim they will provide some incredible all encompassing solution using LLM AI as the back bone of their infrustructure, however, I have found it a useful player in a traditional tech stack when used only for appropriate tasks.
#Example Prompt
You are game designer who's focus is designing game assets like maps and levels for a 2d top down village simulation game. Your game uses a grid based 2d top down Map where each grid cell is drawn a square tile image. Your job is to help design a realistic layout of these tiles based on real world geographic and geological norms and other logical placement of tiles. You will represent the tiles as a number (index of tileMap array), you will be supplied with a list of corresponding tile descriptions that map to each number. Consider the tiles neighborhoods up to 2 tiles away. You are generating an island, start by filling the top, bottom, leftmost and rightmost rows with the "open water" tile. Near the center place a grass tile. Build from there. Make sure the land tiles that have open water neighbors are replaced with beach or rocky shore tiles or something else like a dock that makes sense next to the water. TileMap: {"O": "Open Water", "G":"Grass", "r":"Rocky Shore", "s":"Sandy Beach", "c":"cliff shore", "D":"dock", "R":"rocky land", "T":"trees", "C":"inland cliffs", "R":"river", "B":"Bridge"} Add a dock that starts with atleasr one tile land starting on a non-water tile that borders the ocean, and is atleast 3 tiles long in a straight line into the water. Make the map 16 by 16. It is VERY IMPORTANT that you make sure to use each tile type at least once on the map.
When displaying each tile of the map, format the tile as the key or index of the tile map for that tile.
Chat GPT gives output along the lines of:O O O O O O O O O O O O O O O O
O O O O O O O O O O O O O O O O
O O O O O O O O O O O O O O O O
O O O O O O O O O O O O O O O O
O O O O s s s D O O O O O O O O
O O O r G G G D O O O O O O O O
O O O r T G G D O O O O O O O O
O O O r T G R R r s s O O O O O
O O O r G T C R r r s O O O O O
O O O r G G R C T G r s O O O O
O O O r T G G T G r r s O O O O
O O O r T G G G G T G r s O O O
O O O O s r G T G G T G r s O O
O O O O O s s r G G G G r s O O
O O O O O O O O r G T G r s O O
O O O O O O O O r r s s O O O O
https://redd.it/1es6va8
@proceduralgeneration
Nova Patria: Creating a Procedurally Generated Roman Steampunk World
https://www.youtube.com/watch?v=VO9jcWzvCBM
https://redd.it/1es5931
@proceduralgeneration
Good communities for shader-based workflows?
I've got about 3 years of C++ under my belt (mainly Unreal Engine), and I've decided I want to tackle a planet generating compute/vertex shader for large scale terrains. Are there any good active discords or communities specifically focused on shader development?
https://redd.it/1erp4n6
@proceduralgeneration
procedural generation topics sorted by difficulty
i'm new to procedural generation.
the thing i'm ultimately interested in is city generation.
yes, the only topic i can say for sure is the hardest in this question is city generation.
city generation should be broken down into road network/city map generation, actual road geometry/mesh generation, building generation, etc.
other topics could be maze generation, dungeon generation, terrain generation, biome generation, tree and plant generation, fractal generation, space filling curve generation, tessellation generation, and so on.
however... let's put my greed aside for a moment and start with something easy.
also how about following and learning from Junichiro Horikawa's excellent houdini tutorial series on youtube?
https://redd.it/1ereob4
@proceduralgeneration
Procedural Trees. Hi, this is an old code (I'm making a new one)
https://redd.it/1er1s1r
@proceduralgeneration