Welcome, Guest. Please login or register.

Login with username, password and session length

 
Advanced search

1411273 Posts in 69323 Topics- by 58380 Members - Latest Member: bob1029

March 28, 2024, 03:19:20 AM

Need hosting? Check out Digital Ocean
(more details in this thread)
TIGSource ForumsCommunityDevLogsCerebral Override - Cyberpunk Point & Click detective game
Pages: [1]
Print
Author Topic: Cerebral Override - Cyberpunk Point & Click detective game  (Read 849 times)
Perequin
Level 0
*



View Profile
« on: February 12, 2019, 12:18:02 AM »

Cerebral Override is a point and click adventure game in a cyberpunk setting.



The story
The story revolves around AI assisted animal uplift, in the wake of a near complete ecological apocalypse. Animals have been integrated into society, initially as a form of entertainment for rich people uplifting their pets, but they now clamour for animal rights as their different species are subjected to extinction.
The uplift technology is made possible by BioLift devices, created by a tech company called BioCorp. It is a device that attaches itself to the brain gradually with the use of a web of growing fibers from the device itself, accessing all brain functions.
The BioLift devices have become popular in black markets used on people as a way of boosting brain functions. It is not illegal to have a BioLift device in your brain, but it is illegal to perform the procedure.
You play as private detective Willard Sharpe, hired to investigate a murder committed by an uplifted animal, something that can potentially change the public opinion on uplifted animals, which is already waning thanks to the growing demands for animal rights.

The gameplay
I'm going for a point and click adventure game mainly relying on conversations and puzzles involving deduction of clues. There will be no inventory but a growing list of clues you have to sift through, and that you can use in conversations.

The art style
The art style is heavily influenced by the Lucasarts classics, more so from The Dig than say, Monkey Island or Sam & Max.
I've decided to have normal mapped pixel art enabling me to have dynamic lighting.
I made an art test last fall to test what I could do, and the result is a form of 3d-assisted pixel art, which I will try to explain in an upcoming post. The big drawback is the art scope which balloons almost out of control when you have to draw every animation frame 7 times (diffuse, 5 lighting directions, spec).  WTF Unfortunately for me the result has the potential to be astounding so I feel like I have to try and go for it.

Current status
I've been working on this for about a year, trying to figure out the story and art style. The game is made in Unity. For the past year I've implemented Ink into my project, which is a narrative writing compiler created by Inkle and I'm using it to write all dialogue and cutscene scripting. I've also written my own A* pathfinding system and I've also done some work on how I handle all the clues in the game both from a gameplay perspective but also the workflow for it.
Currently I'm trying to "white box" the entire game with a first draft of the story and basic gameplay which I will probably be stuck doing for quite a while. This will hopefully inform me of the art scope needed and I can act on that information once I know more.

Follow me on Twitter if you want to keep up on progress outside of this DevLog!

Cheers!
« Last Edit: February 12, 2019, 11:20:16 PM by Perequin » Logged

Perequin
Level 0
*



View Profile
« Reply #1 on: February 13, 2019, 12:02:08 AM »

The art test
So I'd like to do a little write up on the art test I did last summer/fall.
I knew I wanted the game to have a 2d look with orthographic cameras. I stumbled upon a program called SpriteLamp, which enables the generation of normal maps from 2d images. There are other more streamlined programs out there that does the same thing, but none creates the look, or gives you the control that this does. Mainly because it's all done by hand, which is a problem when it comes to scope, so we'll see if this is the one I'll keep using or not down the line.

The program requires you to input 4-5 images of the sprite lit from different directions. Left, Right, Top, Bottom and Front.


You also have to create the diffuse and specular map for proper rendering in the game engine, similar to a classic 3d workflow.

Here's my player character lit in-game.



Now this might be alright for characters but I knew I'll have quite a lot of environments to create as well, and they contain ALOT of pixels and complexity in terms of geometry and I'm not good enough of an artist to create that by hand. It would simply take too much effort.

So I came up with the idea that if I could create the scene in 3D and simply render out the light maps, I would get a lot of that information for free. Then it would simply be a case of hand-painting the diffuse on top of that information. It's obviously cheating, and probably will have me tarred and feathered in the pixel art community, but hey... it works.  My Word!


Environment lit from the front.

I also knew I'd want reflections from rainy streets since hey... cyberpunk. So I rendered out a reflection pass as well.




The reflections also receive the light information. The character reflections are just mirrored versions of their sprite sheets for now.

So that all worked out pretty much perfectly, there was only one problem, which was shadows.
The SpriteLamp shaders do have the ability to handle some self shadowing from the depth maps that it creates, but it wasn't nearly enough. Especially when it comes to character throwing a shadow onto another object which just isn't possible at all. I tried to render out the depth map from my 3d program, but that didn't work either.
The real issue is that you can't really have shadows dance across geometry because that depth doesn't exist.

Then I remembered that I had the scene in 3d already, and maybe I could utilise that in my game somehow to generate the shadows.
So imported the 3d mesh, and made a shader that only gives me the shadow information and rendered that on top of the rest of the scene through a secondary camera.
I also have a copy of my characters walking around on that 3d geometry, so they can cast their shadows onto the 3d mesh.



And it's still looks like it's in 2d, because it is!
Here's a perspective view.



The second benefit of having a 3d mesh match the background is that I can have VFX responding to the geometry, without having to bake it into some kind of background animation. And not only that, I can also have those effects get affected by light as well.



So that's pretty much it. There are still some issues with this technique that I haven't had time to solve yet, like the 3d version of my character gets smaller the further away from the camera he gets, which isn't the case in the 2d view, because it's orthographic, so I have to scale them accordingly the further into the scene they get.
I also haven't completely solved the geometry self shadowing itself, but I think that's mostly a case of picking and choosing what geometry actually cast shadows, right now none of them do, but we'll see.
This was good enough of a test to give me the confidence that this will work, though.

« Last Edit: February 13, 2019, 12:08:01 AM by Perequin » Logged

Pages: [1]
Print
Jump to:  

Theme orange-lt created by panic