Hans Fuchs posted: You gotta use stable diffusion with comfyUI and models from civit.ai, the popular models EASILY beat DALL E. Just tell me if you need some more advice.

Yesterday I looked into image generators, stable diffusion etc. some others that can be used locally on a mac. Nothing was even close to what I'm getting with DALL E on open ai. What do you use for image generation?

Maksim @mxd · 1y

Replying to @_hans

ComfyUI looks like a lot of install work, is that what you're using? Is there a way to see and compare results? If the hustle worth it in the end.

Hans Fuchs @_hans · 1y

Replying to @mxd

Maybe you can try InvokeAI, I find it much easier to setup & use: github.com/invoke-ai/Invo… it's hard to really compare results since you're in full control with Stable Diffusion while all other solutions try to make it easier, but it's definitely worth it long-term.

GitHub - invoke-ai/InvokeAI: InvokeAI is a leading creative engine for Stable Diffusion models, empowering professionals, artists, and enthusiasts to generate and create visual media using the latest AI-driven technologies. The solution offers an industry leading WebUI, supports terminal use through a CLI, and serves as the foundation for multiple commercial products.

InvokeAI is a leading creative engine for Stable Diffusion models, empowering professionals, artists, and enthusiasts to generate and create visual media using the latest AI-driven technologies. Th...

github.com

Maksim @mxd · 1y

Replying to @_hans

What are you using for image generation and what are the images for?

Hans Fuchs @_hans · 1y

Replying to @mxd

I've used InvokeAI extensively, so I can definitely recommend it. I produced some AI images for websites in scenarios where you really can't tell the difference (smaller images), but also for hero sections for example. Tip: stay away from "edgy" cyberpunk themes, they're overused

Matteo Modica @matteomodica · 1y

Replying to @_hans @mxd

That’s interesting. How would you compare stable diffusion’s results to Midjourney v6?

Hans Fuchs @_hans · 1y

Replying to @matteomodica @mxd

Well, Stable Diffusion allows you to train your own LORAs so you can actually make AI images of yourself for example. Generally speaking, Midjourney is much easier while Stable Diffusion can technically produce much better results: youtube.com/watch?v=z-AoEL…

I Spent 1000 Hours Researching This - You Won't Believe What I Discovered About Stable Diffusion!

LINKS BELOW----- After over 1000 hours of research and testing, I have created the ultimate guide to achieving photorealistic images with Stable Diffusion. This 182 page prompt book contains over 350 tested images and 200+ prompt tags for getting perfect results every time. Learn the secrets the pros use for creating jaw-dropping, realistic images that look like photographs. Discover the best model settings, prompts, and techniques to take your stable diffusion images to the next level. See examples of prompts for different styles like glamour, street, documentary, and more. Get prompts for lighting, cameras, lenses and other properties to make images come alive. This book will transform the way you use AI to generate images. Everything you need to start making professional quality AI generated art is contained in this totally FREE book. All I ask is that you like this video, subscribe, and if you can spare it, donate $2 for my coffee fund. Get the book now: https://promptgeek.gumroad.com/l/photoreal ____LINKS_____ The book: https://promptgeek.gumroad.com/l/photoreal Models: UniverseStable: https://civitai.com/models/102001 AbsoluteReality: https://civitai.com/models/81458?modelVersionId=108576 Eyes LORAs: Female: https://civitai.com/models/116247?modelVersionId=127908 Male: https://civitai.com/models/118349/eyes-male-from-polyhedron Negative Prompt Embeddings: Bad-Hands-5: https://civitai.com/models/116230?modelVersionId=125849 BadDream & UnrealisticDream: https://civitai.com/models/72437?modelVersionId=77169 ____CHAPTERS_____ 0:00 Intro 0:54 Getting realistic images 1:20 Introducing prompt book 2:40 Models to use 3:46 Prompt settings 6:00 Prompt structure 6:26 Overview of book 6:56 The perfect prompt structure 7:00 Style of photo examples 8:13 Subject prompts 10:09 Posing prompts 11:00 Framing prompts 11:59 Background prompts 12:35 Lighting prompts 13:50 Camera angle prompts 14:04 Camera property prompts 14:42 Cinema camera prompts 15:11 Digital camera prompts 15:22 Retro camera prompts 15:34 Film type prompts 16:00 Lens prompts 17:02 Filter prompts 17:19 Photographer style prompts 17:48 Sharing images prompted

youtube.com

Maksim @mxd · 1y

Replying to @_hans @matteomodica

Wow this looks really good, thank you! I'll dig more into it.

Maksim @mxd · 1y

Replying to @_hans

Also what I find great about OpenAI is that it comes up with a prompt itself, for other local tools it seems like you have to really invest into the prompt to get anything decent, use negative prompts, parameters etc.

Hans Fuchs @_hans · 1y

Replying to @mxd

Yeah you gotta spend some time to learn everything but Stable Diffusion XL actually provides better results when you do so. This is a good place to start: youtube.com/watch?v=z-AoEL…

I Spent 1000 Hours Researching This - You Won't Believe What I Discovered About Stable Diffusion!

LINKS BELOW----- After over 1000 hours of research and testing, I have created the ultimate guide to achieving photorealistic images with Stable Diffusion. This 182 page prompt book contains over 350 tested images and 200+ prompt tags for getting perfect results every time. Learn the secrets the pros use for creating jaw-dropping, realistic images that look like photographs. Discover the best model settings, prompts, and techniques to take your stable diffusion images to the next level. See examples of prompts for different styles like glamour, street, documentary, and more. Get prompts for lighting, cameras, lenses and other properties to make images come alive. This book will transform the way you use AI to generate images. Everything you need to start making professional quality AI generated art is contained in this totally FREE book. All I ask is that you like this video, subscribe, and if you can spare it, donate $2 for my coffee fund. Get the book now: https://promptgeek.gumroad.com/l/photoreal ____LINKS_____ The book: https://promptgeek.gumroad.com/l/photoreal Models: UniverseStable: https://civitai.com/models/102001 AbsoluteReality: https://civitai.com/models/81458?modelVersionId=108576 Eyes LORAs: Female: https://civitai.com/models/116247?modelVersionId=127908 Male: https://civitai.com/models/118349/eyes-male-from-polyhedron Negative Prompt Embeddings: Bad-Hands-5: https://civitai.com/models/116230?modelVersionId=125849 BadDream & UnrealisticDream: https://civitai.com/models/72437?modelVersionId=77169 ____CHAPTERS_____ 0:00 Intro 0:54 Getting realistic images 1:20 Introducing prompt book 2:40 Models to use 3:46 Prompt settings 6:00 Prompt structure 6:26 Overview of book 6:56 The perfect prompt structure 7:00 Style of photo examples 8:13 Subject prompts 10:09 Posing prompts 11:00 Framing prompts 11:59 Background prompts 12:35 Lighting prompts 13:50 Camera angle prompts 14:04 Camera property prompts 14:42 Cinema camera prompts 15:11 Digital camera prompts 15:22 Retro camera prompts 15:34 Film type prompts 16:00 Lens prompts 17:02 Filter prompts 17:19 Photographer style prompts 17:48 Sharing images prompted

youtube.com

Posts