- Видео 26
- Просмотров 291 934
SiliconThaumaturgy
США
Добавлен 9 дек 2022
I make in-depth tutorials and educational videos on Stable Diffusion!
SDXL September Model Roundup - 30 SDXL Models Ranked and Reviewed
For this video, I did in-depth testing on 30 SDXL models from Civitai and will be ranking them on the old RUclips Classic, a tier list.
Spreadsheet for Model Testing Method and Data
docs.google.com/spreadsheets/d/1wLJ27ySec_pkSiMW1IJlAKlB_Gw3kTngpjpjAnGVwBU/edit#gid=116461310
Intro - 00:00
Anime ArtDiffusion Alpha 3 - 01:15
BluePencil XL v0.2.1 - 01:58
Copax Timeless v4.4 - 02:45
Counterfeit XL v1.0 - 03:25
CrystalClear XL - 04:08
Dark Pizza Origin v4.0 - 04:54
DreamShaper XL Alpha 2 - 05:46
DucHaiten XL v1.0 - 06:25
DynaVision XL Beta 0.4.1 -07:05
Explicit Freedom Alpha - 07:56
Hephaistos v2.1 - 08:45
Juggernaut XL v2.0 -09:22
Kohaka XL Alpha - 09:58
Lah Mysterious v3.3 - 10:32
LeoSam’s Hello World XL v1....
Spreadsheet for Model Testing Method and Data
docs.google.com/spreadsheets/d/1wLJ27ySec_pkSiMW1IJlAKlB_Gw3kTngpjpjAnGVwBU/edit#gid=116461310
Intro - 00:00
Anime ArtDiffusion Alpha 3 - 01:15
BluePencil XL v0.2.1 - 01:58
Copax Timeless v4.4 - 02:45
Counterfeit XL v1.0 - 03:25
CrystalClear XL - 04:08
Dark Pizza Origin v4.0 - 04:54
DreamShaper XL Alpha 2 - 05:46
DucHaiten XL v1.0 - 06:25
DynaVision XL Beta 0.4.1 -07:05
Explicit Freedom Alpha - 07:56
Hephaistos v2.1 - 08:45
Juggernaut XL v2.0 -09:22
Kohaka XL Alpha - 09:58
Lah Mysterious v3.3 - 10:32
LeoSam’s Hello World XL v1....
Просмотров: 7 284
Видео
Nine Tips and Tricks for Automatic1111
Просмотров 4 тыс.9 месяцев назад
Intro - 00:00 #1 -Load Settings from Image - 00:22 #2 - End of Job Sound - 01:39 #3 - Continuous Image Generation - 02:05 #4 - Quick Settings - 02:35 #5 - Quick Prompt Emphasis - 03:26 #6 - Change Prompt without Changing Composition- 04:27 #7 - Caching models to RAM - 05:19 #8 - Use Symbolic Link to save disk space - 06:08 #9 - Precise Inpainting - 07:39 Symbolic Link Helper Spreadsheet: docs.g...
Stable Diffusion Samplers - Which samplers are the best and all settings explained!
Просмотров 17 тыс.9 месяцев назад
This video will teach you everything you need to know about samplers in Stable Diffusion including how things change for SDXL and which samplers are which for ComfyUI. First, I break down the differences between the samplers by comparing processing speed, convergence, and the minimum steps required. Afterwards, show which samplers result in similar inputs and use the results from the comparison...
SDXL 1.0 blows away Stable Diffusion 1.5. And here is the testing to prove it.
Просмотров 21 тыс.10 месяцев назад
In this video, I will compare the newly released SDXL 1.0 checkpoint to both base Stable Diffusion 1.5 and top tier Stable Diffusion 1.5 checkpoints to see how they compare. I start out by discussing the architecture of SDXL compared with SD 1.5 to see how they compare on paper. Afterwards, I show the results of extensive testing in Automatic1111 and ComfyUI comparing SDXL to SD 1.5 in terms of...
Better Hands in Stable Diffusion 1.5 - Part 3 - Embeddings
Просмотров 8 тыс.10 месяцев назад
For this third and final part of my series on getting the best hands with Stable Diffusion 1.5. In this video, I test 13 different negative embeddings across several models to see which consistently improve hands and which do not. To round things out, I will do some testing to see which of the top 3 embeddings have stackable effects to help you minmax generation of hands and give you some guida...
Which Stable Diffusion 1.5 model makes the best hands? Part 2: Thirty More Models
Просмотров 3,9 тыс.10 месяцев назад
For this video, I rigorously tested 30 Stable Diffusion 1.5 models to see which models have the best hands including more popular models, viewer suggestions, and whatever I had lying around. Afterwards, I go through all the models in part 1 and part 2 by category using statistics to show which models are the highest performers in each category. Intro - 00:00 30 - 00:48 20 - 04:31 10 - 07:32 Sta...
Which popular SD 1.5 model makes the best hands? Plus a hand embedding that makes hands worse?!?!
Просмотров 3,5 тыс.11 месяцев назад
For this video, I tested 16 popular Stable Diffusion 1.5 models to see which made the best hands. After showing how testing was performed and scored, I show the models starting from worst and going to best with a little extra commentary on the strengths and drawbacks of closely ranked models. Finally, I test out two negative embeddings specifically for hands to see if they make a difference. In...
Customize the StableDiffusion Automatic1111 Interface - Change initial values, slider ranges, & more
Просмотров 2,6 тыс.11 месяцев назад
Link to Spreadsheet docs.google.com/spreadsheets/d/15r06uepR2OuxW3bf7IoD-_dUAuRuOAHkLIQH1D3mct0/edit?usp=sharing This video will show you how to customize your Automatic1111 interface for Stable Diffusion to your personnel preferences so you don’t have to change values each time you load it by editing the ui-config.json file in your Automatic1111 installation folder. First, I cover the nomencla...
ControlNet Reference-Preprocessor Comparison, Settings, Troubleshooting, and Style Transfer Workflow
Просмотров 10 тыс.11 месяцев назад
In today’s video, I will teach you everything you need to know about using the ControlNet Reference Preprocessors for Stable Diffusion in the Automatic1111 Web GUI. Starting out simple, I show you very basic usage of the tool then quickly move onto a comparison of the three Reference Preprocessors to identify what each does and doesn’t do. Next, I get technical to identify which settings like S...
Stable Diffusion Basics - When VAEs are and aren't needed and the ways to use them
Просмотров 14 тыс.11 месяцев назад
This video will teach you what VAEs do in Stable Diffusion and how to use them. First, we will cover what types of models require VAEs to get good results and which do not. Then, we will cover the two main ways to use VAEs in the Automatic1111 interface along with their benefits and drawbacks. The Stability AI VAE refenced in this video can be found here: huggingface.co/stabilityai/sd-vae-ft-ms...
Model of the Week Review - Lyriel - Like a gatcha game you pay for with electricity
Просмотров 1,2 тыс.Год назад
This is a review of the most popular model on Civitai for the week of May 17th, 2023. I start by showing what the model is good at and not so good at then break down which settings I used and why. I performed quality testing to show how often certain defects like bad hands or bad anatomy occur then go into how celebrities, styles, and artists work in prompting. Finally, I do a parental guidance...
Make Bigger Images in SD: Pros and Cons for Outpainting, Hires Fix, Img2Img, and ControlNet Tile
Просмотров 7 тыс.Год назад
Today, you are going to learn how to make bigger images in Stable Diffusion. There are many ways to do this, including adjusting the base resolution, outpainting, Hires Fix, Img2Img Upscaling, Ultimate SD Upscale with ControlNet Tile, and simple upscaling. For each of these, I will discuss the pros and cons along with potential pitfalls. Once we’ve covered them all, we will use that information...
ControlNet Deep Dive - OpenPose 1.1: Add Faces and Hands to OpenPose - Detection, Settings, and More
Просмотров 5 тыс.Год назад
This video is a comprehensive tutorial for OpenPose in ControlNet 1.1. If you are new to OpenPose, you might want to start with my video for OpenPose 1.0, since this video focuses on the additional face and hand functionality in OpenPose 1.1. OpenPose 1.0 Tutorial: ruclips.net/video/0RGYwnfxTRo/видео.html First, we discuss the five pre-processors for OpenPose as well as what the maps correspond...
Stable Diffusion Basics - Inpainting - All Settings Explained plus How to Solve Common Problems
Просмотров 19 тыс.Год назад
This video will provide a comprehensive tutorial for inpainting in the Automatic111 web gui. We will learn what all the settings in the inpainting tab do including mask blur and mask padding, recommended ranges for important settings like denoising, and how to troubleshoot common problems encountered when inpainting. Intro 00:00 Inpainting Basics 00:15 Masking and Denoising 01:28 Manual Editing...
ControlNet Deep Dive - Depth - Preprocessors, Weight and Guidance, and Generating at Max Resolution
Просмотров 9 тыс.Год назад
This video will teach you everything you ever wanted to know about using the Depth model in Stable Diffusion ControlNet. First, I will teach you the strengths and weaknesses of the two Depth Pre-processors, MiDas and LeRes, as well as how their settings impact the generated depth maps. Weight and guidance are the most important settings for using ControlNet. I will teach you what they do, how t...
ControlNet Deep Dive - OpenPose - What it can detect and output plus weight and guidance settings
Просмотров 7 тыс.Год назад
ControlNet Deep Dive - OpenPose - What it can detect and output plus weight and guidance settings
Testing ALL Artists in Stable Diffusion - Part 1 - A is for Artist
Просмотров 3 тыс.Год назад
Testing ALL Artists in Stable Diffusion - Part 1 - A is for Artist
How Much Information Can Stable Diffusion Handle in a Single Prompt? - Stable Diffusion Deep Dive
Просмотров 6 тыс.Год назад
How Much Information Can Stable Diffusion Handle in a Single Prompt? - Stable Diffusion Deep Dive
Stable Diffusion Basics - XYZ Plot - Test up to three variables at once
Просмотров 23 тыс.Год назад
Stable Diffusion Basics - XYZ Plot - Test up to three variables at once
Stable Diffusion Deep Dive - CFG - Don't Accidentally Fry Your Images
Просмотров 11 тыс.Год назад
Stable Diffusion Deep Dive - CFG - Don't Accidentally Fry Your Images
Stable Diffusion Basics - How to execute many prompts at once AND specify settings for each one
Просмотров 8 тыс.Год назад
Stable Diffusion Basics - How to execute many prompts at once AND specify settings for each one
Stable Diffusion Basics - Prompt Emphasis and Blending Concepts using your prompt
Просмотров 9 тыс.Год назад
Stable Diffusion Basics - Prompt Emphasis and Blending Concepts using your prompt
StableDiffusion Deep Dive - HiRes Fix - How to avoid twinning and losing composition
Просмотров 39 тыс.Год назад
StableDiffusion Deep Dive - HiRes Fix - How to avoid twinning and losing composition
Stable Diffusion Basics - Styles - How to save and (more importantly) delete styles
Просмотров 28 тыс.Год назад
Stable Diffusion Basics - Styles - How to save and (more importantly) delete styles
Stable Diffusion Basics - Prompt S/R - The most powerful way to test embeddings and prompts
Просмотров 7 тыс.Год назад
Stable Diffusion Basics - Prompt S/R - The most powerful way to test embeddings and prompts
Stable Diffusion Deep Dive - Samplers - which are worth using and which are redundant?
Просмотров 18 тыс.Год назад
Stable Diffusion Deep Dive - Samplers - which are worth using and which are redundant?
Where I can find this app you are using ?
dude, i just jumped in around Xmas 23 w/ 8g 3060 now its 500gigs of stuff & you just gave the best little darn talk ive seen so far. subbed.
Thank you VERY much for the time and effort you put into this! Most excellent!
This is incredible stuff.
Thank's alot
with all due respect, the narration goes into a solid C tier; I am absolutely sure you are able to pinpoint where it falls short when you review your commentary. Hint: Fast is not always better
According to Wikipedia, Leonhard Euler and Karl Heun are Germans, in their language EU reads as OY - OYLER and HOYN - it comes from pronunciation of Greek ϒυ - ύψιλον. It would make sense to pronounce the names based on the one principle, and the way their mom and dad called them.
I rarely subscribe anyone in RUclips but i subscribed you because your videos were jam packed with info. i hope you're alright now that you are not uploading.
please keep uploading man
how can i made that in comfyui
This is awesome! Would you also be able to turn all these matrices into a CivitAI article that could be quickly scrolled through? It would be great to be able to consult these via a page and scroll through them
You really went above and beyond with this spreadsheet, it’s fantastic! Thanks for all the hard work. 👌👌
Dude, I usually dont write comments, but this guide is awesome as hell. The explanation with the "a" in the name and the benchmarks PLUS your "good image" disclaimer were extremely well done and professional. (I am currently at minute 3, but had to post a comment first lol) I hope you get more subscribers and viewers. Thank you for your service!
a year later and i still have that little upscale vs denoising chart as a jpg on my desktop for quick review
Are you still out there? :) You stopped all videos about 6 months ago...
Extremely helpful video, thanks
The community misses you bro, your vids are still in a league of their own!
Your sound is too low.
Very clear, thank you very much. I just stepped up my SDXL game.
Any insight or help with images made in defourm? Colors are never consistent as more images are made. I have color coherence set to LAB. I don’t know what to do. Cfg 7
Best in depth video.
What were the prompts you used?
more of this please
Thank you brother, this will help a lot!
great video, could you tell me how to score the hands generated by AI?
It is sad that realistic models still trail behind anime/cartoon models. Like making something manga or waifu seems to be accepted in society then tasteful nudity and of course simple hands and feet.
The older models tend to do better, when they get recombined, or more censored, they seem to do worse. That is my problem with sdxl right now.
I know this video is 1 year old, but I wonder of through a similar technique can you achieve similar results as Korea or Magnific. Have you tried something like that?
I think it's far more just getting lucky than people want to admit. I've masked out the person's shirt because it's just a t-shirt and I want the shirt to be a different style. I changed the prompt (haven't seen anything saying this won't work) to the specific clothing item and it doesn't even come close to getting it right.
Fantastic video! I have only been messing around in controlnet without really understanding the differences. This all makes so much more sense now. Thank you!
This is a lot of work! Thank you for putting the effort in. Revisiting this in 2024 would be interesting!
barely audible
low volume
I came to similar conclusions with my own tips research and noob testing, as i have a 2070s im not going to make a big xyz grid so this video is good to see it all laid out. Wish i saw this earlier, would have saved me loads of time, but hey the experimenting was good for learning a1111, id still say im a noob though.
how can we use the batch with inpainting ? thanks for the video
Thanks!!
do you know if theres a way to use reference-only without using a prompt? i have an image and i want it to be replicated, but i don't want to prompt it
that was awesomely presented ! thanks for the good work !
This is amazing thank you!
SDXL is still hopeless at photo realism even with the latest models. I’ve given up with it now.
so thankful I found your channel, so informative <3
0:05 I prefer the one on the left. It looks way more interesting.
Using 'Euler a' as opposed to regular Euler, seems crazy fast & seems to produce very good quality images considering how quick it is.
Same experiences! It rarely messes up the quality and I love using it!
And Cyberrealistic is not even in the discussion, where I create almost best results all the time with that
SDXL epic fail...
Could you please turn the volume down more, that way I can officially blow my ears out on the next video instead of just going def.... thanks... Learn to encode audio.
where do you download the VAEs?
Short, sweet, and to the point. Earned a sub my friend.
The best channel on the topic. Only objective tests with x/y test matter, not some rando reddit flame war. Thank you! Ps you had my sub at "the doggiest dog :)"