4 Realistic Stable Diffusion Checkpoint In 10 Minutes - (Automatic1111)

There are many checkpoints for Stable Diffusion. This video will show you and review 4 realistic Stable Diffusion Checkpoints which you can download for your project. We will test different prompts to explore the checkpoints strengths and weaknesses.
-LINKS-
(When using an affiliate link, I earn a commission which is a fantastic way to support the channel)
➤ GENERAL ACCESSORIES - amzn.to/3Jl9zpe
➤ NVIDIA GRAPHICS CARDS - amzn.to/3TZNGRn
➤ AMD GRAPHICS CARDS - amzn.to/3PYIaNN
➤ INTEL CPU - amzn.to/3vRlYya
➤ RAM - amzn.to/3Jm53Hf
➤ CORSAIR VENGEANCE 16GB - amzn.to/3PY8JTm
➤ CORSAIR VENGEANCE 32GB - amzn.to/4aRTwLx
➤ CORSAIR VENGEANCE 64GB - amzn.to/3Q1NboQ
➤ GENERAL MONITORS - amzn.to/3xEk6td
➤GENERAL KEYBOARDS AND MOUSES - amzn.to/4aAaEpp
➤GAMING KEYBOARDS AND MOUSE - amzn.to/49wpYly
Support the Channel:
➤ Patreon: / bitesizedgenius
➤ Buy Me Coffee: bmc.link/bitesizedgenius
Links:
➤ 4x NMKD Upscaler: nmkd.de/shared/?dir=ESRGAN/Mo...
➤ Photon: civitai.com/models/84728
➤ EpicRealism: civitai.com/models/25694
➤ Juggernaut: civitai.com/models/46422?mode...
➤ Verisimilitude: civitai.com/models/10540
Chapters
0:00 - Introduction
0:20 - Photon
2:43 - Epic Realism
5:38 - Juggernaut
8:05 - Verisimilitude
10:29 - Conclusion

[@BitesizedGenius]

Gaussian Blur added at 0:10 to stop KZbin sending me to the shadow realms 👁👄👁

[@Feelix420]

epic realism and realistic vision 5 is all i need

[@mrowiegg6781]

Love ur content sir!

[@heikohesse4666]

hi, cool video - Juggernaut is insanely good!

[@KandooAI]

Thanks for checking out my Model (Juggernaut). That means a lot to me :) Juggernaut can do many art styles, but it´s best on photography and photorealism. Everything else is "okish" :D Buildings are indeed a problem, but Interior inside a building works really good with Juggernaut.

[@BitesizedGenius]

Hey no problem, keep up the great work! I used Juggernaut for the thumbnail as the characters are great, but i'll have to check out the other art styles at some point :)

[@KandooAI]

@@BitesizedGenius You should try out some "Movie Still, Film Still, Cinematic" Stuff. I trained it with a high amount of cinematic Images :D

[@pacorouget8571]

im just seeing all your videos, its fucking amazing how its done. Theres no smarter way to doing your "tutorials" than you do.

[@TheKamrenB]

Where can I find that list of upscalers you had at 3:43?

[@BitesizedGenius]

In description now

[@TheKamrenB]

​@@BitesizedGeniusamazing thanks!

[@johnleighdesigns]

great to see this really helpful - unfortunately for me my PC and GPU are too slow to be useable in the way you demo'd - it takes 15-20 mins to generate a rubbish image using automatic 1111 where my amd 2600 CPU and nvideo 1650 super are great for the mix of graphc design, retouching, 3D work and motion and video editing but for AI image generation is not good enough not just about the time it takes but what comes out is nothing like the galleries im seeing. For now im using the restrictive but speedy sites like payground AI, Lexica, and Clipdrop while i consider whether to spend out later in the year on a newer CPU and GPU

[@BitesizedGenius]

Do you find those speedy sites easier to use or do they lack resource? I've heard people in similar situations.

[@johnleighdesigns]

@@BitesizedGenius yeah for sure - Im totally new to automatic1111 and can see the versatility and power potential if you have the right setup - experimenting with Cliprop and others has been really useful as i couldnt do it on my pc and GPU

[@alanedwards1179]

If you enjoy using this application, then you should splash out on a new system with good CPU/GPU. My setup takes around 30 seconds to generate high quality output

[@luke-ev5rz]

so confused on how to use this! is it an app?

[@BitesizedGenius]

Software, i'll make a tutorial soon on how to install 👍

[@bozhidaratanasov7800]

Hi, Can you explain what the UI you are using is to tune your parameters/ enter prompts? What is the tool called?

[@BitesizedGenius]

Automatic1111 - Got tutorials on how to use it on the channel!

[@JimboSkinner]

Cfg=7 will never follow your prompt closely. If you want to see every term in your prompt you need to start increasing that number, at least 11, maybe even higher. 16 will rigorously follow your prompt, at the cost of model cohesiveness

[@WifeWantsAWizard]

Here's the problem: "realism" depends not on the diffusion model but on the dataset. If you only have six images in your library and you haven't broken down all of the meta data, you could have the greatest model in the world and it will generate crap because the library is broken. The only way to truly determine which services are offering quality output is if we have a universal standard for evaluating input datasets.

[@chickencombo7563]

what's the link for the website?

[@BitesizedGenius]

This one? civitai.com/

[@chickencombo7563]

@@BitesizedGenius thank you sir. I just started getting into Ai imaging recently. I want to learn how to prompt as a beginner.

[@BitesizedGenius]

@@chickencombo7563 No problem, i've got a prompting guide on my channel, its quite popular!

[@chickencombo7563]

@@BitesizedGenius I appreciate it sir. Thank you!

[@VigneshoViyan]

List out some art style models

[@BitesizedGenius]

You mean like anime, watercolour etc?

[@VigneshoViyan]

@@BitesizedGenius watercolor, sketch, pencil drawing that kind of model

[@SHARMATHEGAMER]

0.10 bro u just missed something

[@BitesizedGenius]

Thanks, just going to slap on some gaussian blur before people notice 😂

[@branddonefor21]

came here before it got edited

[@Gg_system]

you forgot mine! smh

[@BitesizedGenius]

No I didn't, Am I Perfection will have its chance to shine, don't worry 😂

[@Gg_system]

@@BitesizedGenius fantastic , V2 shines , photogenesis is my latest attempt at photorealistic portraits, and I'm honoured you know of my creations lol

[@byujkt3822]

Tried some model look too smooth skin. Only photon can generate realistic aesthetic textured skin.

[@MitrichDX]

Verisimilitude - badest.

[@tokatoofficial]

realistic vision?