Image generator inside games (not AI)

I’ve toyed with the idea of making an image generator that would react to your choices in character appearance or actions that affect the environment and change the illustrations of the game based on that. Like a doll maker thingy. It’s a bit complex and would require way more coding and drawing skills than what I have. I’m just interested if any IF writer has already done this sort of thing.

2 Likes

Do you mean like graphical adventure games that draw a picture of the room, including any items you’ve left in it?

1 Like

Yes, pretty much that. Just that the choices would be made with the parser or by clicking hyperlinks. Some of those old browser based virtual pet games did this, at least. I don’t know if it’s even worth the effort, but it would be a cool gimmick (in my opinion).

1 Like

Emily Short’s Alabaster does this:

5 Likes

I think you can kind of see an example of this “paper doll” sort of thing in most every RPG with a character/avatar creator. Visual novels also do this: usually a character will have several stock silhouettes/poses, or even modular arm positions and expressions and different costumes to match that can be layered to create multiple versions of the character with moods and costume changes.

The simplest kind of thing you can do is divide each image into consistent “chunks” that match up like the paper-doll flip books for kids.

Here’s a YouTube tutorial with an artist explaining how Visual Novel sprites are created.

There’s probably an executive decision to make whether you want to “remix” the character appearance in-engine, or just externally create every variation and choose the complete image version that matches the choices the player makes.

2 Likes

In Cygnet Committee I used conditional layered images for the in-game map. IIRC the game checks off rooms depending on whether you’ve visited them, and it draws some hidden paths once you take them.

If you’re using Twine and Sugarcube, conditional images are just as straightforward as conditional text. You just need to surround images with if/then macros. Of course this is limited to what you know how to do with the engine in general.

However, you’ll need familiarity with CSS, particularly z-indexing (ie. layers) and positioning to make the images line up properly.

2 Likes