D&D General DALL·E 3 does amazing D&D art

[Jfdlsjfd]

D&D spell illustration challenge, day 40 (12 days late): Contingency

A single 6th level spell could have saved the CERN tens of billions of investiment expenditures.

Prompt: Inside a cluttered wizard’s study—books on quantum thaumaturgy and multiverse harmonics lie open—a robed archmage with well-trimmed grey hair and starry robes lounges in a velvet armchair, sipping tea. A large glass orb glows faintly overhead. Suddenly, the orb lights up bright blue. The wizard freezes mid-sip, eyes wide with awe and triumph. Behind him, a chalkboard reads: “Contigent Light: if a Higgs Boson is present in the Universe.” A cat jumps, startled, from the windowsill.

Commentary: this one was tough, how do you illustrate a spell that does absolutely nothing? Also, yes, Contingency is being nerfed over and over (in 3.5 you could have Contingency, Chained Contingency and Simbul's Spell Sequencer stacked, now you need to wait for level 11 to get a single Contingent spell...) but they didn't specify the limitations on what the trigger is. Contingent Light triggered by "the assassin of Lord Greystaff is within 5 feet of me" as you tour the suspects, "Contingent Gaseous Form whenever I am about to be hit by a physical attack" or "Contingent Teleport whenever I am holding my staff parallel to the floor in the exact direction of the Lost Temple of Xzurg..." (triangulation is the worst enemy of old ruins) still remain attemptable by players.

 

[Jfdlsjfd]

(side discussion before I resume doing the spell list)

For those who like AI image but are reluctant to use models trained on copyrighted artworks, the works to train model on licensed art is progressing.

• Firefly (inside Adobe's suite) is now, in arena matches, where people are asked to tell which image is better when confronted to the same prompt, is on par with Flux Schnell, the lightweight edition of open-source models you can run on your own computer.
• If you prefer models you can run at home, Bria 3.2 is having SDXL-like performance, which is where the best performing model was 8 months ago.

It is far from perfect and below current models of course, but if one wants to create tokens for their online games without regret for uncompensated authors, it might be interesting.

I have played with it a little:


A warforged with a leather belt and a leather satchel on his torso

A wizard with a fireball over his palm:

(the 4 fingered-hand was a classic blunder back then and the improvement comes from a different training, not from using more images)

Two space marines in front of a glass tube holding a green fluorescent liquid.

A floating castle in a gloomy forest, in front of the moon

An orange large industrial truck with large wheels and a futuristic design, the driver is standing next to the vehicle, parked in a desert environment next to some black metal structures.

A 1920 british gentleman studying over a tome in his library. Besides him, a brain floats in a jar.

a spaceship flying in front of Jupiter

a middle-aged woman sitting on a bench in a public park, wearing red dress, pink hat, white shoes. A dog sits at her feet.

While these images will be a throwback from what is recently presented on the thread, it's because it's a 4 billion-parameters models, while current models -- which are more costly to train -- are 12 to 18 billions parameters. The nice thing with small models is that on current harware it takes 5 seconds to generate an image, so you can easily get a bunch of them, choose the closer to your need and correct the details. Maybe there will be interest for this line of models, whose main attraction, for those concerned, is that it can be used guilt-free (and for everyone, it dismisses the myth that AI images need to be trained on copyrighted work to get anywhere).

 

[Jfdlsjfd]

And in the realm of videos...
View attachment 1753719557619.webp

(same prompt as the Contingency spell).

Can't say I'm a fan, but it exists.

 

[DragonLancer]

Again, not D&D art related (sorry) but I've been playing around with ChatGPT's art generation to create some artwork to draw some wargaming friends away from 40K to a system I'm trying to generate interest in (OPR's GrimDark Future, for reference). I've asked to make some art to help me promote it. I'm not a fan of the standard AI art as it feels too... plastic. I've had good results across the board if I ask for the art to be photo realistic as though taken by a National Geographic photographer. It struggles with some images though as it says that there isn't any photos of the relevant contents to create it. Sounds like a bit of a cop out but I guess AI art is still in it's rough infancy.

I'm going to try my hand at some D&D art for my own campaign setting soon.

 

[Ulfgeir]

Made in Lexica.art

Character name: Amy Clarke / Pandora Rosenstar

Some images of a a character that I first played in a forumbased post-apocalyptiv freeform game, then reused in Dresen Files. She has also appeared as an NPC in Daring Comics rpg, and in Scion 2e.

The prompt were variations of: "Attractive voluptous gothic witch with long black hair, pale skin, tattoos, black makeup, lots of silver jewellry. She is dressed all in black [Long fingerless lace gloves, velvet skirt, leather corset]. she is reading the future through her tarot cards."

 

[Jfdlsjfd]

D&D spell illustration challenge, day 40 (30 days late...): Control Weather

Role model for 15th level wizard for high school male pupils: Kwarg, the Dreaded Necromancer of Xang, leading an undead army.
Role model for 15th level witch for high school female pupils: a weather girl. At least they're not just announcing the weather.

Prompt: A TV weather forecast set, bright and modern, with large digital screens showing rainy clouds systems and temperature maps. At the center stands a glamorous weather presenter, seen from the bac, in front of a map of France. She's facing the map display. She wears a sleek, professional outfit infused with magical elements — shimmering fabrics, glowing runes on her sleeves, and a wand ending with a glowing red gem instead of a pointer. Besides her, a small pseudodragon flies.

Commentary: yet another high level spell (8th) that require concentration for 8 hours to cause the temperature to go from warm to hot. That's brilliant display of anthropic warming isn't global, it's just for a 5 miles radius. Tell me no one ever memorized that spell? Also, I specifically insisted in the prompt to have the main character look at the map so it is drawn from behind, not looking at the viewer, but after generating it I remembered that news anchors are not looking at the map but toward the camera...

 

[Jfdlsjfd]

D&D spell illustration challenge, day 41 (29 days late): Create Undead

You can start at level 13 by casting your daily Finger of Death at an unsuspecting peasant. After 20 years...


Prompt: A side-view of a powerful Greek-inspired sorcerer riding a nightmare horse (fiery-eyed, obsidian-black, flaming hooves). The sorcerer wears flowing robes reminiscent of ancient Hellenic garments, embroidered with arcane symbols, gold trim, and laurel motifs. His face is stern, partly shadowed by a bronze circlet. His staff crackles with necrotic energy. Behind him: a line of undead in decayed hoplite armor, eyes glowing blue, some carrying ancient Greek weapons. Beyond them, a massive, indistinct army of zombies stretches into a grey mass — endless, faceless ranks, their forms blending into one another. The atmosphere is heavy, apocalyptic, with drifting ash, torn banners, and the reddish haze of some cursed battlefield.

Commentary: this is Kwarg, the Dreader Necromancer of Xang, leading his undead army. As mentionned above, he kindly volunteers in high schools to present his line of work to aspiring young pupils. When your parents aren't into necromancy, you often miss the opportunity, and somehow carreers advisors are reluctant to offer Evil Overlord as a goal for ambitious teenagers.

 

[LuisCarlos17f]