Community
General and Everything Else => The Games and other Sierra Adventure stuff => Topic started by: cosmicr on August 31, 2022, 02:48:07 AM
-
I've been having a blast converting Sierra images to modern art pieces using AI.
See below some examples:
(https://i.imgur.com/5B5cmVW.png) (https://i.imgur.com/FPlF9Qs.png)
(https://i.imgur.com/cUYdmcg.jpg)(https://i.imgur.com/Oay9NIf.png)
(https://i.imgur.com/kgnG3FS.png)(https://i.imgur.com/XSWrW2W.jpg)
(https://i.imgur.com/gFihujj.gif)(https://i.imgur.com/AQ9lsda.png)
(https://i.imgur.com/2IRjMVV.jpg)(https://i.imgur.com/ajxYk8l.png)
(https://i.imgur.com/GqbASDm.png)(https://i.imgur.com/Lre3BJp.jpg)
(https://i.imgur.com/uOFwt0f.png)(https://i.imgur.com/RuX4tZn.png)
I have found that it works best for fantasy style scenes and landscapes. Luckily, a lot of sierra games feature these!
-
That looks amazing!
Can you further explain how did you do that?
What are your next plans?
-
You are aware, of course, that Maniac Mansion was not a Sierra game? 😸
-
I'd love to know how this was done!
-
I threw Maniac Mansion in there for fun, as I am also a big fan of LucasFilm(Arts) games. Looking for any other suggestions.
This is done using a Machine Learning platform called Stable Diffusion with Image-to-Image. You basically provide an image and a description, and the model will produce a result. You can then fine tune the result by feeding it back into the input, update you description as necessary, and repeat as many times you like. There are a lot of parameters that can be changed to influence your result. I also upscaled the images using AI and for the Passionate Patti Picture I ran it through GFPGAN which makes faces look more realistic. You need a decent GPU for quick results.
Here's a bonus image from Betrayed Alliance:
(https://i.imgur.com/7S61A26.png)(https://i.imgur.com/Qcu38ab.png)
If anyone has any requests let me know, I love this stuff. It's not limited to this style either. It can do painted, pencil sketch, 3d rendered, other styles I haven't thought of.
-
You could try the pics from KQ4 for a possible VGA conversion.
-
I like that skeleton Sludge pic and the skull cave turned out well.
How close can you get to the KQ4 box art if you start with Rosella, the Unicorn and Goons on the path to the castle?
-
Wow! Didn't expect to see a rendering from my game on here! Really cool stuff and thanks for sharing it. I really dig the KQ 4 one at the top, too
-
Those are absolutely exceptional.
-
Great renditions! Keep em coming folks. Would love to see some SQ3 scenes. Cheers!
-
Hehe, it certainly has come a long way since DALL-E and Stable Diffusion 1.5 (Which is what I used for those images in my original post)
As soon as the new OpenAI image generator came out the first thing I tried was some sierra screenshots...
(https://i.imgur.com/GOialH7.jpeg) (https://i.imgur.com/iPE5F4S.png) (https://i.imgur.com/CdoNc3X.png)
(https://i.imgur.com/x5IFoLq.jpeg) (https://i.imgur.com/KKbdQnC.png)
(https://i.imgur.com/axw4nXe.jpeg) (https://i.imgur.com/rhF1IWZ.png)
(https://i.imgur.com/0uhpzzZ.png) (https://i.imgur.com/nXHvpm3.png)
The only thing I don't like about the OpenAI generator is that it's not open source, and it seems to put a brown hue over every 2nd image...
-
(https://i.imgur.com/TDiYmi7.png)
awesome
-
The thing I've always wondered about this scene is if it's all automated, why are the lights on all the time?
-
Great stuff. I think such AI image conversions definitely can be used to create VGA-like versions of Sierras parser EGA games. Either by using SCUMMVMX (little work) or creating genuine SCI01 VGA parser games. Some challenges, though, like the generated images moving the camera, modifying objects, use different colors in different images, etc. But still possible, I would think.
-
There area other models like "Depth Anything" by Meta that could be used to re-create the priority screens. If you were REALLY keen on upscaling/recreating a game using AI, it could be done, albeit not without any manual work.
Upscaling sprites and animations (views) is much more difficult though.
-
Keep in mind that many of the Sierra artists are vehemently against AI art and might be offended if posted where this might be dropped in their laps.
-
*sigh*
-
Its crazy how much the narrative has shifted in 4 years... It's gone from "cool little toy" to everyone and their dog hating AI and everything it stands for.
I just like technology. I dream one day for a world like Star Trek - talking computers, holodecks, replicators. Can you imagine the push back today if replicators were invented!
I'm still into generating images but I don't share it with anyone anymore because more often than not it can bring out some surprisingly intense opinions...
-
I'd actually be interested in seeing how things go from the other way around - e.g. taking a photo of a real location and then generating a 256-color or 16-color image for use as a background in a game, with adjustments made as necessary to ensure things like walk paths and item hotspots are of a suitable size. Could be very useful for making fan games, replicating the style & charm of the originals. And I don't expect it would be too technically difficult; AI's involvement would be minimal here, basically just deciding what tweaks to make to the downsampled image, and ensuring stylistic consistency between all the images in an area/game.
Also FWIW I agree with Collector's point about AI & art for existing pieces, though I think these tools are promising for new games, esp. fan stuff like AGI/SCI where most of us simply don't have access to the time/resources/talent required to make a high quality game all by ourselves.
-
Your idea sounds pretty good, but I'd say if AI's involvement is that small you might as well leave those tweaks and checks to the user. I figure that could also help not make everyone's stuff look the same (1).
As for having the time and tools... FotoSketcher (https://fotosketcher.com/) gives pretty good painterly results for free and has custom preset support. Color grading and other content tweaks can be done in anything from Paint.net to Photopea (https://www.photopea.com/). It takes me more time to set up the original scenes and turn them into SCI backgrounds with priority screens and polygons, than it does to paintify them.
(1) Consider how you can pick out AI artwork so easily. Or the responses to DLSS5.
(2) haha you can type superscript digits all you want and they work in the preview but once they're stored... I hate this board's database settings I truly do XD