Stable Diffusion Akashic Records
"Sic itur ad astra"
A compendium of information regarding Stable Diffusion (SD)
This repository is a collection of studies, art styles, prompts and other useful tools you can use throughout your exploration of the latent space.
As Stable Diffusion is still in beta and subject to lots of changes, the Records will often change to reflect new information.
📔 Table of Contents
General information [ToC]
The Records
SD is brand new. As such, not many resources exist which are solely based on it.
To differentiate between other AI imagery tools and SD, the Records will use the following nomenclature :
🔹 Resource based on SD
🔸 Resource based on other AI imagery
🔺 Resource not based on any AI imagery (such as photography terms)
Stable Diffusion
To use SD, you can access the official DreamStudio website or use the Open-Source Model
Keep in mind the following :
- Your prompt must be 75 tokens or less, anything above will result in "prompt truncated after tokenization"
- Your prompt is case insensitive
- There's about 30,000 tokens understood by the AI, this means it won't know about some weird word unused since the 1600s
- A token is roughly a word, a punctuation, or a Unicode character
- Same prompt, same seed, same modifiers will end with the same result
- To make variations of an image, it's recommended to keep the seed and slightly alter the prompt / modifiers
Guides
🔹 Stable Diffusion
Ultimate Beginner’s Guide
🔹 Stable Diffusion/Dreamstudio Guide
🔹 SD Launch Presentation
🔹 Creating cute animals with SD
🔹 Exactly How Stable Diffusion Generates Art
🔹 Seed Editing
🔹 SD Guide for artists and non-artists (csunberry#0594)
🔹 Understanding SD through teapots
🔹 How to maximize your DreamStudio credits
🔹 SD Windows Installation Guide
🔸 The DALLE-2 Prompt Book
Colabs / Alternative websites
🔹 HuggingFace Space
🔹 SD Notebook with KLMS sampling
🔹 SD with 🧨 diffusers
🔹 SD Researcher
🔹 Prompt-space Interpolation
🔹 SD with Queing and Google Drive
🔹 SD on Colab (Collection of colabs)
🔹 Optimized SD repository
🔹 Dezgo
🔹 Deforum SD
🔹 SD on CPU repository
Studies [ToC]
A study is the same seed with slightly different prompts / modifiers. It's mainly used to understand the impact of a word/modifier on the final result. It must contains the resulting images
Modifiers
🔹 CFG modifier studies
🔹 Sampler studies
🔹 Iban's Image size animation study
🔹 ESoS' Image size study
🔹 ESoS' Steps & Sampler study
🔹 ESoS' CFG study
🔹 Chris Allen's CFG animated study
🔹 Localstarlight's CFG animated study (localstarlight#3529)
🔹 Captain's ETA/scale study (Captain#4568)
🔹 Captain's ETA/steps study (Captain#4568)
🔹 Bacontime's sampler animated study (bacontime#6072)
Keywords
🔹 Fruit basket study (burnin#2452)
🔹 Park bench study (burnin#2452)
🔹 Camera study (Mr_Sumato#3369)
🔹 SD emojis study (Spake#1808)
🔹 4 Games Cafeteria study
🔹 Stylistic Lighting study
Misc
🔹 Multilingual capabilities study
🔹 Stable Diffusion Prompt Weighting study
🔹 160 celebrities study
🔹 One Hundred TV shows study
🔹 Film studies
🔹 Punctuation study
🔹 Secondary artist study
🔹 Prompt permutation study
🔹 Artist order study (Kitt Hirasaki)
Art style/Artist lists [ToC]
🔹 Suburban House, Art styles
🔹 Stable Diffusion Artist Studies
🔹 SD Artist Collection (SGreen#2222)
🔹 Bacontime's Artist comparisons, part 1 (bacontime#6072)
🔹 Bacontime's Artist comparisons, part 2 (bacontime#6072)
🔹 SD Styles & Mediums
Keyword lists [ToC]
🔹 Artists, Keywords, Art styles known to work with text to image
🔸 Offhand's Disco Diffusion Prompt Keywords
Misc [ToC]
🔹 Image dimension table (brbbbq#0731)
🔹 Image to prompt : CLIP Interrogator
🔹 SD Image splitter
🔹 Dream Channel Backup - SD prompts + image links / SD prompts only / Raw CSV
🔹 SD Image Manager
🔹 SD Prompts to sqlite
🔹 SD Prompts + modifiers sqlite DB
🔹 UMAP Maps of CLIP ViT-L/14 keywords representations
🔸 SD's Dataset search : CLIP retrieval
🔸 Prompt builder
🔸 Tools and Resources for AI Art
🔺 Camera distance terms
Prompts [ToC]
Here's a list of quick prompts to get you started in the world of Stable Diffusion
🔸 Prompt generator
🔹 GTA V cover character : [subject] in GTA V, cover art by Stephen Bliss, artstation
🔹 Quick photorealism (NOP#1337) : [prompt], Canon EOS R3, f/1.4, ISO 200, 1/160s, 8K, RAW, unedited, symmetrical balance, in-frame
🔹 Age changer (richservo#8465) : DSLR photo portrait still of [age] year old age [age] [person] at age [age]!!!, 85mm f1.8
🔹 Film character (richservo#8465) : film still of [character] [action] in [movie], 4k
🔹 Real life cartoon character (richservo#8465) : portrait photo still of real life [character], 8k, 85mm f1.8
🔹 Dish picture (richservo#8465) : DSLR food photograph of [dish description], 85mm f1.8
🔹 Muppet portrait (richservo#8465) : studio portrait still of muppet !!!!![person or character] !!!!!! as a muppet muppet as a muppet, 8k, studio lighting, key light -C 13 -s 150