Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
Moongrave
Jun 19, 2004

Finally Living Rent Free
hey if you guys were as hosed up as i was back when SD was just a discord channel to prompt, old prompts from that still work VERY well in Dall-e 3

Adbot
ADBOT LOVES YOU

Icedude
Mar 30, 2004









XYZAB
Jun 29, 2003

HNNNNNGG!!
A friend and fellow goon who works at a radio station where I also used to work sent me a text today asking for a $30 donation for this year's fundraising marathon.

I cracked open Bing and replied with:

quote:

This you?


:owned:

But a handful of others were good, too.






feedmyleg
Dec 25, 2004
Some more spookies

























bees everywhere
Nov 19, 2002

Nyan Bread posted:

If you're looking to dive into the weird and wonderful world of AI art generation using Stable Diffusion on your local PC and have a video card with at least 8GB VRAM,

Thank you for posting this! Gonna quote it so I can find it this weekend and try it out :)

EVIL Gibson
Mar 23, 2001

Internet of Things is just someone else's computer that people can't help attaching cameras and door locks to!
:vapes:
Switchblade Switcharoo
I'm playing with FenrisXL and the quality gets crazy.

Guide from a good guy of how to get some crazy detail with it and upscaling: https://youtu.be/t6EF-vr3ydc?si=hd3HMCIdr8xicx14

A good thing about Nyans guide is it lets you incorporate CivitAI galleries so installing them is not a pain

Nyan Bread posted:

If you're looking to dive into the weird and wonderful world of AI art generation using Stable Diffusion on your local PC and have a video card with at least 8GB VRAM]

Hadlock
Nov 9, 2004

Lava vs electric eels



Brawnfire posted:

Very inspiring!

A speculative ecotopian future town square that serves as a public forum and collaborative holographic art display space. The historic brick streets and several preserved historic architectural pieces are still lovingly maintained.




This exists in Spain. Seville to be exact

Tarkus
Aug 27, 2000


Fabulousity
Dec 29, 2008

Number One I order you to take a number two.

"Ok, final test. Go into the server room and install a patch cable. Where? Oh, it doesn't matter. As long as you get it jacked in somewhere and come back alive you get your certification!"



Also a happy accident:

Tarkus
Aug 27, 2000


I like this one

Nyan Bread
Mar 17, 2006

Tarkus posted:


I like this one

Really captures the watercolor on paper canvas look, now just needs a frame.

Swagman
Jun 10, 2003

Yes...all was once again peaceful in River City.
















KwegiboHB
Feb 2, 2004

nonconformist art brut
Negative prompt: amenable, compliant, docile, law-abiding, lawful, legal, legitimate, obedient, orderly, submissive, tractable
Steps: 32, Sampler: DPM++ 2M Karras, CFG scale: 11, Seed: 520244594, Size: 512x512, Model hash: 99fd5c4b6f, Model: seekArtMEGA_mega20

black cat walking in front of your path while stepping on a crack and opening an umbrella indoors on friday the 13th, every bad luck symbol at once
Negative prompt: rabbit's foot, four leaf clover, lucky coin
Steps: 48, Sampler: DPM++ 2M Karras, CFG scale: 11, Seed: 1, Size: 512x512, Model hash: f65275abcf, Model: notreallyistic_v2, Denoising strength: 0.48, RNG: CPU, Hires upscale: 2, Hires steps: 32, Hires upscaler: 4x-UltraSharp, Version: v1.5.2

I tried but I couldn't manage to get them to walk under a ladder with an upside down horseshoe and spilling salt at the same time. Next friday the 13th perhaps.


Nyan Bread posted:

If you're looking to dive into the weird and wonderful world of AI art generation using Stable Diffusion on your local PC and have a video card with at least 8GB VRAM
This is an excellent write-up. Thank you for doing this. I know first hand how fast things move and can make these guides outdated, they are still extremely helpful for everyone that is just starting out their journey so thank you again for putting forth the effort.
Maybe AARD VARKMAN can put this in the first post of this thread? Since it still links to SD 1.4 there...

Illiterate Clitoris
Oct 24, 2011

I was looking for a point of re-entry, thank you for this.
Heads up, Visual C++ 14.0 seems to be a prerequisite for this. The installer crashes for me and points me to the C++ Build Tools. Even though web says it should handle everything.

Illiterate Clitoris fucked around with this message at 07:14 on Oct 14, 2023

TIP
Mar 21, 2006

Your move, creep.



















TIP fucked around with this message at 08:52 on Oct 14, 2023

The Sausages
Sep 30, 2012

What do you want to do? Who do you want to be?
Artist's portrayal of Microsoft Bing and Dall-e 3




Debunk This!
Apr 12, 2011




Phat Phingers
May 27, 2023

Ey Frito-Lay! FUH Q MANG!

oh that's sick, I really like how accurate it is to BOTW's style.

Gashroom
Jul 13, 2005
Blobfish goes shopping

ThisIsJohnWayne
Feb 23, 2007
Ooo! Look at me! NO DON'T LOOK AT ME!



Gashroom posted:

Blobfish goes shopping



Amazing

Brawnfire
Jul 13, 2004

🎧Listen to Cylindricule!🎵
https://linktr.ee/Cylindricule

I've got a mental image I'm trying to manifest of a cafe and organic market arrayed around a huge tilapia tank with aquaponic plants above.















Not sure what happened here. GMO-bombing by ecoterrorists, perhaps?

Mercury_Storm
Jun 12, 2003

*chomp chomp chomp*
children's storybook about a baby seal disguised as a loaf of bread that drinks only Mountain Dew and eats only Doritos in a magical forest of smiling happy rat families



baby seal disguised as a loaf of bread making repairs to the International Space Station in orbit above Earth

Mercury_Storm
Jun 12, 2003

*chomp chomp chomp*
Kinda wish I had gotten in earlier, everything else I want to generate is blocked lol

except this I guess:

a swarm of cheesy rats with cheesy bread stick tails in a multitude of boxes of cheesy bread stick rats being served by a rat waiter to a family at an intergalactic Pizza Hut on the outside of a space station

Mercury_Storm fucked around with this message at 21:17 on Oct 14, 2023

Fabulousity
Dec 29, 2008

Number One I order you to take a number two.

Gateway 2000 ad copy from the 1990s vibes from this one:


Fabulousity fucked around with this message at 07:17 on Oct 15, 2023

Tarkus
Aug 27, 2000

So, one thing I notice is that if the subject matter doesn't match the temporal context of an art style, it tends to do worse with particular styles.

For example the remainder are all midcentury modern style. perfect



Try doing it with just a cell phone. mmmm, not bad



try it with a man looking at a cell phone, no dice



And now with a man and a laptop.




I'm exaggerating a bit with my selections but it tends to blend the art style with particular aesthetics if the subjects don't match well. maybe it's just me. neat though.

Mescal
Jul 23, 2005

Nyan Bread posted:

If you're looking to dive into the weird and wonderful world of AI art generation using Stable Diffusion on your local PC and have a video card with at least 8GB VRAM, this weekend might be the best time to start with the newly released higher-res SDXL dataset. Here's a bunch of tips to help you on your way to setting up a basic workflow in the very flexible node-based browser-based editor ComfyUI:

Download and install Stability Matrix AI suite manager that will be used for downloading custom SD models and further customizations, as well as updating ComfyUI - https://github.com/LykosAI/StabilityMatrix/releases

Use the Packages tab to download the latest ComfyUI. Once finished it will be integrated into the Stability Matrix suite to allow using the Model Browser tab to download the trained datasets directly from https://civitai.com


Now onto the actual ComfyUI editor where the AI art generation will take place. First get the ComfyUI-Manager plugin:
https://github.com/ltdrdata/ComfyUI-Manager/archive/refs/heads/main.zip

Extract, and copy the ComfyUI-Manager-main folder into your "\StabilityMatrix\Data\Packages\ComfyUI\custom_nodes" directory. Relaunch ComfyUI and it'll download the dependencies and the Manager new button will show up in your editor.


Within the Manager click Install Custom Nodes and use the search field to install the following plugins for painless use of this noodle-based editor: (also make sure to relaunch the app so it gets the dependencies)
ComfyUI Impact Pack
Efficienty Nodes for ComfyUI
ComfyUI-Bmad-DitryUndoRedo


From the Model Browser tab of Stability Matrix I recommend you download the following startup components:
SD XL (get both the VAE Fix and the refiner VAE Fix)
SDXL VAE
Detail Tweaker XL
realESRGAN_x4+HAT (then copy realesrganX4HAT_v10.ckpt from your \StabilityMatrix\Data\Models\StableDiffusion\ directory into the \StabilityMatrix\Data\Models\ESRGAN directory)


Once those are downloaded you should be good to go as far as having all the raw image data to manipulate and the tools to do it with offline. Now here's some basic components of the actual ComfyUI workflow that you can build upon:


The compact "all-in-one" loader is used for loading the base model, the refiner, specifying the size of the generation as well as connecting to the Low-Rank Adaptation of Large Language Model (LoRA) stack which are like contextual modifiers or "smart filters" for the generated image:


The sampler is the engine of AI generation that performs functions on the dataset to come up with a coherent image. You can play with these settings to generate wildly different images from the same promt:


FaceDetailer is an automatic inpainting module that can fix messed up faces/eyes/hands but also slightly alters the entire image quality since it does contextual blending. Hopefully a smarter one comes out soon that only blends the image around the area of the mask:


Upscaler increases the size of the image heuristically and is pretty fast, more upscalers can be downloaded from https://huggingface.co


Lastly, you can import and export workflows in ComfyUI as .json files - here's the one in the example, enjoy your journey! https://pastebin.com/raw/EHvnh8y7

effort posts get quotes

edit: i still can't believe it writes text

Mescal fucked around with this message at 22:13 on Oct 14, 2023

Small Strange Bird
Sep 22, 2006

Merci, chaton!
Various Batsmen:

(It did a pretty good Brian Bolland on that last one.)


Prompt was 'batman pours jar of pickled onions over the penguin's head'. Well...

'knockoff star trek toys on a shelf in a store'

pzy
Feb 20, 2004

Da Boom!





Soviet thrift store

Hadlock
Nov 9, 2004


:golfclap:

Edit: imagine four balls on the edge of a cliff

Hadlock fucked around with this message at 00:12 on Oct 15, 2023

Tarkus
Aug 27, 2000

A variety of regal steed





Mescal
Jul 23, 2005






feedmyleg
Dec 25, 2004
ChatGPT is finally allowing for vertical orientation, but seems a bit confused about how to do it. Getting this sort of thing every time:



It's also flipping all of the Goosebumps books I've tried backwards, so I flipped this.

EVIL Gibson
Mar 23, 2001

Internet of Things is just someone else's computer that people can't help attaching cameras and door locks to!
:vapes:
Switchblade Switcharoo

Tunicate
May 15, 2012

feedmyleg posted:

ChatGPT is finally allowing for vertical orientation, but seems a bit confused about how to do it. Getting this sort of thing every time:



It's also flipping all of the Goosebumps books I've tried backwards, so I flipped this.

Yeah they seem to have transposed x and y coordinates instead of rotating, it does that for 100% of images when changing orientation

Tarkus
Aug 27, 2000

I've been trying to get a samurai warrior riding a kappa in the vein of my earlier images. It gives me Shiba Inu's and weird horse-dragons like below. Is there a trick?

Fabulousity
Dec 29, 2008

Number One I order you to take a number two.


:corsair:

LifeSunDeath
Jan 4, 2007

still gay rights and smoke weed every day
can it do funwaa?

EVIL Gibson
Mar 23, 2001

Internet of Things is just someone else's computer that people can't help attaching cameras and door locks to!
:vapes:
Switchblade Switcharoo





Edit: "a red headed child named 'Tasty Armageddon'"

EVIL Gibson fucked around with this message at 04:10 on Oct 15, 2023

Fabulousity
Dec 29, 2008

Number One I order you to take a number two.

EVIL Gibson posted:


Edit: "a red headed child named 'Tasty Armageddon'"


:golfclap:

I would try to do something somehow involving Pele and Oddjob but I'm more tired of seeing that drat "Oops" dog than the Duck Hunt hound.

Adbot
ADBOT LOVES YOU

Gawr Gooner
Mar 3, 2023
It's October but Bing cannot do Sadako without blocking the prompt. Except it can, if you describe her as a "girl with long black hair wearing a one-piece dress" and prepend with something like "creepy black and white grainy vhs still".

So here's something I thought would be fun.


A lot of the stuff I'm doing (including this pic) feels like it would take less time if I just opened an editor and did it myself... But I like the puzzle of doing something the AI doesn't want.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply