AssetStore概要 优惠信息 beta

- 实时检查资源价格变化 -

The Materials Mega Collection: Vol 1 contains 376 high-quality materials & comes fully optimized to work out-of-the-box with PBR, Mobile, PC, Post-Processing, Built-in Render-Pipeline, URP, and HDRP.Materials Include: Ash, Bark, Beach, Brick, Carpet, Circuitry, Cobblestone, Concrete, Crystal, Desert, Fabric, Floors, Forests, Graffiti, Grass, Ice, Jungle, Lava, Leaves, Leather, Marble, Metal, Mosaic, Mountains, Mud, Paint, Paper, Pavement, Plastic, Puddles, Rocks, Rubber, Rust, Sand, Shattered Glass, Sidewalk, Stone, Straw, Tile, Walls, Wallpaper, Water, Watercolor, Wood, and More!Each material comes with the following texture maps:1. Base Color2. Normal3. Displacement4. Specularity5. Ambient Occulusion6. MaskEach Material Includes 10 Different Versions (Totaling 3,760 Material Files):Built-in Render Pipeline1. Standard (PC)2. Standard Specular (PC)3. Bumped Diffuse (Mobile)4. Bumped Specular (Mobile)5. Diffuse (Mobile)URP6. Lit (PC)7. Lit Specular (PC)8. Simple Lit (Mobile)9. Unlit (Mobile)HDRP10. Lit (PC - High-End)Post-processing for Demo scene is setup out-of-the-box for:1. Built-in Render Pipeline2. URP3. HDRPDemo Scene: In the Demo scene, there is a sample room for each of the 10 types of materials that automatically rotates through each of the 376 materials, allowing you to efficiently browse through each type.Documentation: Detailed documentation is included that covers how to easily view materials in Demo scene.Support: Please email brandon@fitfunapps if you have any questions or comments.Features376 Different Stylized TexturesEvery texture has a resolution of 2048 x 2048 pixelsPNG Image FormatTile pattern repeats seamlessly both vertically & horizontallyPBR, Mobile, & PC-ReadyTextures & Materials are already created for Built-in Render Pipeline, URP, & HDRP.10 Rooms in Demo scene to Display Materials: 1 for Each Type of MaterialYou can add the materials directly to your project without any configuration. To preview materials, Demo scene is setup to switch between different pipelines with 1 click.376 .mat Materials: Built-in Render Pipeline - Standard376 .mat Materials: Built-in Render Pipeline - Standard Specular376 .mat Materials: Built-in Render Pipeline - Bumped Diffuse376 .mat Materials: Built-in Render Pipeline - Bumped Specular376 .mat Materials: Built-in Render Pipeline - Diffuse376 .mat Materials: URP - Lit376 .mat Materials: URP - Lit Specular376 .mat Materials: URP - Simple Lit376 .mat Materials: URP - Unlit376 .mat Materials: HDRP - Lit3,760 Total .mat Material Files376 .png Textures: Base Color376 .png Textures: Normal376 .png Textures: Displacement376 .png Textures: Specularity376 .png Textures: Ambient Occlusion376 .png Textures: Mask2,256 Total .png Texture FilesBelow is part of the process I used to create this material collection. Python, ChatGPT, MidJourney, CrazyBump, and Materialize were all used to assist in the material creation process. I started with over 10,000 images and narrowed them down to 376 textures by manually picking out the very best of the best. Using MidJourney's "AI" to create the images was only about 1% of the entire process, which was also the most fun, easiest, and least technical part of the process. MidJourney saved me the step of hiring a graphic designer, and I spent hundreds of hours manually creating a perfect material package outside of MidJourney. I also took the time to develop a series of Python scripts to build sample scenes to quickly view and compare all materials within this collection, unlike any other collection on the Unity Asset Store.Steps:First, I used ChatGPT to generate approximately 2,000 MidJourney prompts. I fed ChatGPT a few examples that I previously used to generate nice-looking textures to train the model on how to generate prompts for these textures. This process can take a few tries of asking ChatGPT for the correct prompt format since the length of the prompt needs to be just right and needs to also include great keywords. These images consisted of approximately one-third of the overall textures.Then I wrote a Python script where I created two lists consisting of adjectives for how I wanted the textures to look as well as nouns for the type of material. The Python script then created every combination of these two lists. I then picked out my favorite ones and used these for the other one-third of the textures.For the remaining one-third of the textures, I reviewed the images from the first two steps, picked out my favorite ones, and generated additional variations based on the look of these images. I also then created hundreds of prompts manually to add in additional creativity.Using all three methods above to create the MidJourney prompts, I typed in a few thousand prompts into MidJourney. It gives back 4 image choices per prompt, so I had approximately 10,000 images to choose from. I reviewed all of these images and upscalled all of my favorite ones.After upscalling about 3,000 full-size images of my favorite ones, I reviewed each image, picked out my favorite ones again, and was left with about 1,000 of the best images. Then, I reviewed what each of these remaining images looked like within my sample scene within Unity and narrowed it down even further to 376 textures.To create the different layers of each texture, I used a combination of CrazyBump, Materialize, and a Python script that I wrote. Each of these 3 methods can generate certain types of texture layers better than the others, so using all 3 allows for a much better final material than if only 1 method were used to create all of the layers.Another Python script was used to upscale and compress all 2,256 textures without reducing the quality. This decreased the file size significantly.Additional Python scripts were used to create the materials for the built-in render pipeline, URP, and HDRP for each of the types.Another Python script was used to add the materials to each of the sample scenes to allow the materials in the scene to change prior to runtime and also to view the materials applied to the scene while they are automatically changing during runtime.After Mask layers were generated by Unity after converting them to HDRP, I then ran these new textures through the Python scripts in the previous few steps.