The Art of Draconis
AI used as an artist’s tool, NOT an artist replacement.
Draconis features 321 beautiful new works of art!
Every piece is created by Wise Wizard Games incredibly talented Lead Artist Antonis Papantoniou.
Antonis has been working full time at Wise Wizard Games for many years.
He created these pieces by combining his painting skills with new AI tools.
As AI art programs came onto the scene, Antonis was interested in seeing how they could be incorporated into his artistic process.
Through experimentation, reviews with the team, and a lot of hard work, Antonis made new art creation techniques for Wise Wizard Games.
The results, which required the skilled hand of an artist, are truly stunning!
Here Antonis describes his process:
I have been working as a game artist for some years now, using various tools, techniques and processes to develop visuals and assets for video games and tabletop games. During this time I have used traditional tools, digital painting tools, and 3D modeling and sculpting tools, often merging these together to explore various approaches.
AI tools have been a recent addition to my toolset, combining them with my other tools to explore a wider variety of options for concept development and visual storytelling.
When I work on a visual, I do have a general end-result in mind, but I like to keep this mental image flexible and open to changes. AI tools allow for this flexibility and exploration of forms, as they produce a vast number of iterations when developing a visual idea.
By exploring, art-directing, and merging these iterations, unexpected visual solutions occur, keeping the process of visual storytelling dynamic and interesting to me.
AI tools also take care of time-consuming polishing passes, keeping the overall process at the speed of concepting and sketching, while rendering polished results.
Step-by-step example
1. Rough hand-painted color sketch, to establish colors and overall type of composition
2. Use the rough sketch as an image prompt to generate thumbnails and explore ideas
3. Pick one of the generated thumbnails to develop
4. Paint-over areas of the thumbnail, to develop the dragon and the background
5. Use the resulting rough image as an image prompt to get a polished version
6. Paint-over, making corrections
7. More paint-over along with AI iterations. This step requires an organic, constant back and forth between painting, generating AI in some areas and manipulating the image using Photoshop’s tools, including its native AI tools
8. The process described in the previous step continues as I detail and polish the image further
9. Intuitive change: As I work on the image, it occurs to me that the environment and this type of dragon are more closely associated to the Command ability of a Gold Dragon rather than the Confuse ability of a Green one. So I use the image as an image-prompt, instructing the AI to change the color of the dragon to gold and making additional color changes in Photoshop
10. Use Photoshop’s image manipulation tools to make the dragon leaner. Painting and smudging over the image and using layers of painterly filters to unify and diffuse low quality artifacts. Also adding effects like the Dragon’s breath and light flares
AI tools used in this process include Midjourney, DALL-E, Photoshop AI Tools, and a local installation of Stable Diffusion and ControlNet.