We’ve been working on portraits for quite some time. The approach is a blend of modern tech and AI portrait generation and photoshop to the finish line.
As the development was going forward, we quickly realized that a small indie team would require a whole department or a dedicated artist to achieve a high-level of non-generated portraits, so some solution in-between that utilized the AI tech of some kind came front and center.
And it’s been real fun playing with different portrait approaches, generators and engines, but the solution ended up many, not one. So, after almost all the portraits are done, we still can’t recommend one approach and engine over another.
However, while the brave new AI frontier has opened up a world of creativity, at the end of the day – “You gotta take that pen in your hands and make it all come together” as our lead artist & game designer Alex Djordjevic noted.
Alex continues: “The biggest problem is consistency and imperfections that are jarringly obvious between all the generators. For example one time you’ll get an ear blending with a hat and you gotta clean that connection up, not to mention the expression and the light/shadow play not always working as expected.“
“Other times details like eyes will be juuust a bit too off center from where you’d draw them on the face if you were doing it by the book.”
And sometimes, some of these imperfections are like imperfection all of us have, and you decide to keep them, because they just work. Other times, not so much.
In conclusion – while a lot of the AI tools will take you to a certain level and we can 100% recommend them – none will provide the consistent look and feel across the board.
So, take that pen and work on the last 1%. The last 1% of work takes 99% of your time, though, so keep it real, and make it beautiful!