Are you kidding? HOW?
(The material is a mapcap, since this is a ROBLOX asset)
The engraving on the front of the vault is a place holder until I have a name and slogan for my vault. Model is also not final.
did somebody say Fallout
Nope, no one said Fallout but you- oh wait I just did too…
Though this isn’t really going to be a Fallout project, I have a very different theme in mind for my “vault” then a Vault-Tec vault, closest thing to my idea except not as pretty would be Vault 88. Even then, its still a long way off from what I got in mind, no further details will be given.
ooo shrouded in secrecy I see
Re-doing all of my GUIs from scratch since I’ve gotten better at graphic design
And the old GUI for comparison lol
Thanks!
Btw, It’s not an Ascii renderer. It’s the console output.
Please add pants.
Please don’t add pants.
In my last post I complained that it was too difficult to take high res screenshots. Well I just made a tool for that.
Yesterday afternoon I made a program in C# using Visual Studio that receives binary instructions from studio using a grid of 1x1 frames that change color, and communicates back using a grid of 1x1 buttons.
I implemented a map stitcher, but the copying/pasting instructions are sent from lua, so I could even use it to generate high res thumbnails or panoramas without changing the C# implementation.
I made it draw in a spiral so things like terrain LOD don’t change violently, but I could always change the order from the lua end:
It can generate a high resolution image using a few hundred screenshots in less than a minute, but the seams can get pretty bad when using large chunks. I decided to see what happens when it takes 2 million smaller screenshots. It took over a day and generated a 21510x21510 image. I’d make it larger, but my program errors when the image’s dimensions are larger than about 22000x22000 because it tries to make an array of bytes that’s too long.
2,056,356 stitched screenshots, downscaled (the original image is 600mb lol):
Small section at full resolution (Trees need some work ):
I might be able to solve the resolution limitation by removing the alpha channel and using 64 bit uints, or by outputting as multiple 20k images. The most difficult part is fine-tuning camera.CFrame and camera.FieldOfView. Scaling down the CFrame’s rotation matrix solved a lot of z-fighting, and maybe scaling specific vectors will help with the seams. This will be especially tricky once I try taking thumbnail screenshots.
Edit: Replaced dead gyazo link with mp4
I have a bad habit of starting new projects and abandoning the ones I was previously working on. Here’s a new project I’m working on based on grimy, semi-run-down parts of Los Angeles.