AI Generative Art

[Trike]

3 hours ago, Player2 said:

No, you get 50 free credits and can earn 100 more by joining their Discord... but you can use that feature for 2 credits each use.

I tried it but it wouldn’t let me click through. How do you do it?

[Herotu]

On 2/25/2024 at 2:21 AM, Trike said:

 

Look fellas, just watch me and you'll pick it up. The steps are easy, listen to the beat. And one, and two...

It goes a little something like this ...

[Player2]

15 hours ago, Trike said:

I tried it but it wouldn’t let me click through. How do you do it?

I signed up with an email to get the 50 free credits.  Drag picture to the upload window on the left, then click the blue button that says Create.  Under the button, there's a time estimate for how long it should take and some green text indicating that you'll be charged 2 credits for the generation.

[Robothenri]

19 minutes ago, Player2 said:

I signed up with an email to get the 50 free credits.  Drag picture to the upload window on the left, then click the blue button that says Create.  Under the button, there's a time estimate for how long it should take and some green text indicating that you'll be charged 2 credits for the generation.

I think they must have changed it now, i've got 43 free credits left, but it won't let me use creative upscale unless I subscribe, you can't drag anything in and the create box is greyed out.

 

[ThaOGDreamWeaver]

On 2/27/2024 at 4:48 AM, Techwright said:

I'd really like to know how you got it to create "Fury Wings" (game's name for 'em) on the helmet.  That's been one of a crazy number of problems I've been trying to solve to do Major Ray Gunn.

Unfortunately I've got no idea - just lucked into that particular combo. Here's the prompt:

Spoiler

Filter: RealVis XL

 

Prompt: A photorealistic middle distance still of a male superhero in full white body armour with blue decal details and helmet with blue visor, rocket boots, flying over a futuristic city, looking towards the viewer, firing a bolt of vivid electricity from his right hand

 

Exclude: ugly, deformed, noisy, blurry, distorted, out of focus, bad anatomy, extra limbs, poorly drawn face, poorly drawn hands, missing fingers

 

Model: Stable Diffusion XL

Guidance: 7

Quality: 30

Refinement: 10

Seed: Randomise

Sampler: DPM ++ Karras

 

On 2/27/2024 at 3:05 PM, Oz The Elderly Gaymer said:

Hey gang, just a reminder to include the descriptive text you used when creating your AI art.  Trust and believe, that really helps.

I'll try and remember.

 

Here's another take on the DW above, or as close as I can remember, using the Ominous Escape preset for Playground XL

(because it's busy right now, and the Photorealistic ones take too much queue time for free users...)

 

Here's the settings that were used:

Spoiler

Filter set: Ominous Escape

 

Prompt: A middle distance still of a 1950s pinup female superheroine with black messy hair with purple highlights, wearing a black leather corset with a spiderweb design, leather jeans, black boots, and a purple cape, walking through the ominous back streets of a city, Cinestill 512T

 

Exclude: ugly, deformed, noisy, blurry, distorted, out of focus, bad anatomy, extra limbs, poorly drawn face, poorly drawn hands, missing fingers

 

Size: 512x512

Model: Stable Diffusion XL

Guidance: 7

Quality: 30

Refinement: 10

Seed: Randomise

Sampler: DPM ++ Karras

 

Attempts required for DW happiness: 7

 

And now with the Masterpiece preset:

 

Again, the thing these seem to annoy me with most are the eyes - any specific ideas for nudging the prompts not to screw those up?

Tried adding blurred eyes, mismatched eyes, distorted eyes to the exclusions but doesn't seem very reliable.

Edited February 28 by ThaOGDreamWeaver

[Marshal_General]

1 hour ago, ThaOGDreamWeaver said:

Unfortunately I've got no idea - just lucked into that particular combo. Here's the prompt:

  Reveal hidden contents

Filter: RealVis XL

 

Prompt: A photorealistic middle distance still of a male superhero in full white body armour with blue decal details and helmet with blue visor, rocket boots, flying over a futuristic city, looking towards the viewer, firing a bolt of vivid electricity from his right hand

 

Exclude: ugly, deformed, noisy, blurry, distorted, out of focus, bad anatomy, extra limbs, poorly drawn face, poorly drawn hands, missing fingers

 

Model: Stable Diffusion XL

Guidance: 7

Quality: 30

Refinement: 10

Seed: Randomise

Sampler: DPM ++ Karras

 

I'll try and remember.

 

Here's another take on the DW above, or as close as I can remember, using the Ominous Escape preset for Playground XL

(because it's busy right now, and the Photorealistic ones take too much queue time for free users...)

 

Here's the settings that were used:

  Reveal hidden contents

Filter set: Ominous Escape

 

Prompt: A middle distance still of a 1950s pinup female superheroine with black messy hair with purple highlights, wearing a black leather corset with a spiderweb design, leather jeans, black boots, and a purple cape, walking through the ominous back streets of a city, Cinestill 512T

 

Exclude: ugly, deformed, noisy, blurry, distorted, out of focus, bad anatomy, extra limbs, poorly drawn face, poorly drawn hands, missing fingers

 

Size: 512x512

Model: Stable Diffusion XL

Guidance: 7

Quality: 30

Refinement: 10

Seed: Randomise

Sampler: DPM ++ Karras

 

Attempts required for DW happiness: 7

 

And now with the Masterpiece preset:

 

Again, the thing these seem to annoy me with most are the eyes - any specific ideas for nudging the prompts not to screw those up?

Tried adding blurred eyes, mismatched eyes, distorted eyes to the exclusions but doesn't seem very reliable.

Vibrant eyes

[Player2]

5 hours ago, Robothenri said:

I think they must have changed it now, i've got 43 free credits left, but it won't let me use creative upscale unless I subscribe, you can't drag anything in and the create box is greyed out.

 

I didn't use the Creative Upscale, I used this one:

[Trike]

They will become heroes one day, but this afternoon they are having a tea party celebrating the fact they are all 10 years old.

 

Left to right: Little Red Riding Hood, Jack Spriggins (soon of beanstalk fame), Hansel, Peter Pan, Goldilocks, Gretel, Alice Liddle (soon to travel to Wonderland)

In front: Little Bad Cub

 

 

Not what i asked for, but I can improvise: art station style, oil painting, sunny afternoon, English garden setting, children's tea party, Little Red Riding Hood age 10 with Goldilocks in green age 10 with Little Boy Blue age 10 with Alice in Wonderland age 10

[Ridiculous Girl]

this is fun! i usually use AI generators to do more abstract stuff. but i figured a would try this for kicks. so i made the OG Ridiculous Girl (fire/fire blaster) from live.

 

bing prompt: cel shaded comic book style, wide angle, very tall woman smiling, with orange hair, orange leather bustier, tight orange leather pants with a fire stripe on the sides, knee-high platform boots, standing outdoors amidst burning wreckage

 

 

i was most recently using AI art for generating images for my dreams; i don't remember the prompt, it was something like; couple, man and woman, sitting outside under a tree, it is nighttime, lit by gas lantern, with a farmhouse in the background.

 

 

another dream, AI prompt: snowman sitting at table in a kitchen, drinking coffee.

 

 

this all makes me want to experiment with my computer art again...

 

normally i do abstract mandalas using photo manipulation and some AI generation supplementing the abstractions... i have been using movie stills as a seed image.

 

seed photo: from Dune

 

initial mandala:

 

final image (that i could find)

 

thank you for getting me addicted again! 😄

[Techwright]

An attempt to interpret Dr. Nightlight.

 

On the one hand, some of the result were rather cool, I'll hand it to the A.I. for that.  On the other hand, there were lots and lots of problems I had to fight the A.I. over this time:

1. I sacrificed a freebee to see if the Bing AI understood what I meant by "Geritt van Honthorst art style".  Van Honthort is among my favorite painters, as is all the Dutch art school known as the "Candlelight Painters" for his ability to bring a feeling of life and warmth to candle-lit scenes.  Bing A.I. clearly understood, as evidenced by the single pic of a character holding a candle in the last row.  However, I could not get Bing A.I. to take what it clearly knew and apply it to the other creations.  Very frustrating.

2. Bing A.I. has a real problem with creating too much light.  I tried a variety of words to try to scale it back, but in scome cases, it chose to shine more brightly.  This was exceptionally frustrating as Dr. Nightlight, by concept, is suppose to be a sole light source at times.  Everything in the shot should reflect his light and the A.I. was constantly putting in alternate light sources (some without sources, just light) that was dramatically reducing the night vs. light impact.

3. Dr. Nightlight is supposed to not have his visage seen.  There should be no chin or nose definition, only the glowing eyes.  Think Star Wars Jawas.  The A.I. seemed to balk at this, and even with a full hood was trying to put eyes or nose in.  Only a few times did I achieve the effect and of those, even less with the glowing eyes.

4. Costume elements: Almost from the start, the A.I. refused to give me a black chain mail vest and a black silk Corsair's sash.  It did relent once or twice, and I've have examples here, but by in large it read "chain mail" and interpreted as "full medieval armor" giving the character a faceplate, breastplate, grieves and many other features.  Rarely did it give out the sash either.  It liked to use leather belts.  Lots of leather belts at times. Eidolons should be so lucky.  It also started shredding the cloak edge unbiddened.  I thought the good doctor was turning into Ragman for a moment.

 

There were a few that got closer to what I asked than the rest.  I've included those amongst this selection.

 

Most recent description:  Geritt van Honthorst art style, black background, floating athletic male hero all in matte-finish coal-black, hood hiding head in complete darkness except for eyes glowing gold-white, cover-the-shoulders clean-edged cloak, tight very black chain-mail across torso, round very bright glow from center of chest, loose black silk sash around waist, matte-black tight knee-high boots with knee guard, black elbow-length gloves,  faintly backlit by candlelight

 

 

[sbloyd]

So I gathered all my favourite Bing outputs of Horizon Twilight, and am having Stable Diffusion process them.  I'm interested in seeing what it does with it in the morning.

[sbloyd]

On 3/2/2024 at 10:34 PM, sbloyd said:

So I gathered all my favourite Bing outputs of Horizon Twilight, and am having Stable Diffusion process them.  I'm interested in seeing what it does with it in the morning.

Well THAT didn't work out.  It super-stylized the outputs. Weird.

[OldManMercy]

Laydeez an gentlmengs, I give you the best 4 from a screenie I took of my  fire/dark corruptor, Burnt Offerings:

 

 

[ThaOGDreamWeaver]

Odd question: anyone got a good formula description for someone holding a Reaper type scythe? For another hobby project I'm working on (building my own Tarot deck...)

[OldManMercy]

I'd use those very words. Grim Reaper Scythe.

 

[sbloyd]

On the Copilot (formerly Bing) Image Creator, adding comic book covered a lot of ground that I was formerly using line art, cel shaded and the like, freeing up space in the prompt for other things.

 

It turned a prompt of 

full body, style of artstation, phoenix, woman made of glowing fire, glowing bright orange skin, skin glowing with inner fire, bare upper arms, hair made out of fire, glowing red flaming eyes, wearing red tech bulky containment suit, shooting flames from her palms

 

 

to

full body, comic book, style of artstation, phoenix, woman made of glowing fire, glowing bright orange skin, skin glowing with inner fire, bare upper arms, hair made out of fire, glowing red flaming eyes, wearing red tech bulky containment suit, shooting flames from her palms

 

[sbloyd]

I need to mess around with LORA creation some more.  My experiment the other day was in the right direction but... flawed somehow.  It turned out a lot of very chunky lines when it generated images.

 

Meanwhile, I got Microsoft to turn out 

which is a lot closer to how I imagine Horizon to be.  

[Techwright]

I've stated it multiple times on the forums:  I'd love to see ethnic neighborhoods added to Paragon City: Little China, Little Italy, Little Greece, Czech Village, etc.  With their Hmong origins, I've felt the Tsoo were a natural for a Little Asia neighborhood.  So I attempted this look. 

 

For the sake of the A.I.'s feeble understanding, I made some compromises.  Instead of a "Little Asia" I went with "Little China".  Rather than waste limited text, I went with "ninja" for a general description.  As always, A.I. struggles with proper rendering of fingers and swords, and some of the rejects were hilarious. Even so, some of the finalists have these problems.  It was unavoidable.  Also curious, the system had a hard time figuring out what a ninja looked like who wasn't in black.  Some of the rejects look liked they formerly attended Ringling Brothers' Clown College.    I've yet to figure out how to get the figures to quit mugging for the camera and get into a real action shot.  Lastly, initial shots of the Little China were overrun by cars and pedestrians, so I placed limiters up front. Other than that, this wasn't so bad. 

 

Description: Low-population, 4 cars, street-level Little China district at night in New England big city, with male ninjas wearing a variety of dark colors with Asian tattoos on their chests, wrapped forearms and shins, holding a kitana and a tanto and wearing paddy hats and Asian demon masks, in action pose

(for the top left shot, I edited out the sword info and opted for "in martial arts action pose".

 

 

 

Edited March 8 by Techwright
Added description

[Trike]

6 hours ago, Techwright said:

"in martial arts action pose

 

I use "dynamic pose" frequently, which usually results in better action shots. Usually.

[GoldMath]

Quick test to make realist my main characters 

  

Edited March 9 by GoldMath

[ThaOGDreamWeaver]

Slightly off topic, but a couple from my tarot deck project...

[Raddy]

A few examples of the art I have generated for my 7ft Alien, Radostina 👽

[Techwright]

13 hours ago, GoldMath said:

Quick test to make realist my main characters 

  

 

What image generator did you use?

[OldManMercy]

So, still trying to find something apropos for my living garbage concept creature/character. Current prompts are: 3d CGI render of a large creature made of corpses and garbage.

I changed it to large amorphous creature and got...less than optimal results. Changed amorphous to shapeless and got that final one.

[Trike]

\

10 hours ago, OldManMercy said:

a large creature made of corpses and garbage

 

eep