I've been having some fun running experiments on NightCafe using the new Stable Diffusion XL model. Here are some collections of studies so far: - SDXL Beta: Artists, AG 275 creations - SDXL Beta: NC Preset (Grid only) 35 creations - SDXL Beta: NC Presets 105 creations - SDXL Beta: VFX Environments 96 creations - SDXL Beta: VFX Scenes … Continue reading SDXL Beta Test Collections
Category: Text to Image
Pulled from a CLIP dataset, here are 1,000 prompt modifiers tested on Stable Diffusion on NightCafe AI text-to-image generator. Landscape Creation Settings Text Prompt: Shown under each image Initial Resolution: Low (16:9, 832*448px) Runtime: Short Seed: 8377729 Overall Prompt Weight: 70% Sampling method: K_EULER_ANCESTRAL CLIP Guidance: NONE 1/ 160s landscape 1/160s landscape 150mp landscape 16k … Continue reading 1000 CLIP modifiers (Landscape)
A visual reference guide on the effects of some artistic movements when used as modifiers in a prompt on NightCafe's Stable Diffusion text-to-image generator. Jump to a section beginning with: <A> <B to E> <F to K> <L to N> <O to R> <S to V> All prompts are the column heading + caption and … Continue reading Art Movements in Stable Diffusion (Portrait, Sci-fi city, Landscape)
There is no excerpt because this is a protected post.
**This post is a work in progress and currently unfinished, but may still provide useful information** What's the difference between the various sampling methods currently offered in NightCafe's Stable mode, and are there any advantages to using a specific one? TL;DR Continue reading to find out. Click #here to jump to the results. You can … Continue reading Study: Stable Diffusion Sampling Methods (WIP)
A new page has been added to the blog: https://tdraw.art/collections to showcase all of the collections that I've made on NightCafe. This will be where you can access all of the individual creations contained within my studies and galleries. I'm planning to add a preview to each one eventually, but for now, it's only the … Continue reading New Collections Page
A study of using different artists as modifiers on NightCafe in both Artistic and Coherent text to image generation.
A study of how NightCafe's Artistic Text to Image algorithm generates images when given the prompt "Animal painting by [artist]"
Use the slider below to see the difference between Overall Prompt Weights 1 & 0 when using NightCafe's Coherent Text to Image generation option. Weight: 1 vs Weight: 0 If this page helps you then please consider supporting the artist via PayPal.
Comparison of how the NightCafe Text to Image Artistic (VQGAN+CLIP) and Coherent (CLIP Guided Diffusion) algorithms generate images using 3 artists with 6 prompts. The initial Artistic and Coherent creations (columns 1 and 2) are generated without a start image, while the final image in each row was created with Coherent using the Artistic output … Continue reading Algorithm Study: 3×6