Jimmy Swanick | A Wigglier Dream

We had a lot of meetings, and I did a lot of research about shaders.

So, one of the biggest shortcomings of the game we built last term was that while it had a handful of nice touches, it didn't look very good. The lighting was far too dark, the lights were unfortunately not volumetric which made it often difficult to realized which way a screenie was facing, and most areas in the demo were extremely dark. It was hard to see many of the things we did to decorate the wall and the environment, and even harder to figure out where you were going.

We also had a lot of particle effects, few of which compared well to each other in the same environment. I set out to learn more about particle effects and shaders in order to eventually unify the appearance of the game and make the cool aesthetic-emotional moments we were planning possible. 

Last term I'd learned a lot of basic things. The most noticeable thing I developed shaders for was the vision portal that used the stencil buffer to record the positions that should be masked out on the walls the player was looking through. That was more functional than aesthetically pleasing, though. There a lot of things that need the same attention.

My first goals was to create an effect for Rem's character model that would communicate better that she is a ghost, and not just an especially blue human. 

Programming shaders isn't at all like anything else I've learned about art, modeling, or programming, so it took a lot of experimentation and research to make any sense of it.

I found three resources which, when I finally found them, led me to accomplish more progress in a few hours than I had in two weeks of reading. 

The first was a game developer who posts amazingly helpful shaders tricks & example code on twitter:

The second was this free, web-based, interactive textbook about shaders, called the Book of Shaders. Its examples and explanations were clear and directly useful. It even lets you play with the code in the examples and see the results change in real time.

The third was this graph editor that shows you the shapes that are produced by the various shaping functions of HLSL, and includes a super concise reference of all the HLSL shaping functions available. 

My goal was to make some kind of effect that would help communicate Rem's ghostliness. Maybe something to make her mesh wavy or ethereal?

I started with one of @Joyce's shader examples that made use of a wave, with hopes to modify that wave to control vertex position in the mesh in a wavy motion, instead of controlling color.

There some strange looking errors along the way, but I didn't think of capturing them until this one, which turned Rem into some sort of much more menacing creature:
Note: videos in blogger aren't currently working for us. Stay tuned for the visual that's supposed to be right here.

But eventually I managed to turn it into a more subtle ripple moving upward over time, which is super closed to what I want to end up with!
Note: videos in blogger aren't currently working for us. Stay tuned for the visual that's supposed to be right here.

Along the way I realized that I could use the same code I'd developed to modify the environment and make more dreamlike, distorted spaces:
Note: videos in blogger aren't currently working for us. Stay tuned for the visual that's supposed to be right here.

I'm getting comfortable with using sine curves, fract functions, and step functions to create new patterns, and using those to modify mesh vertex positions in space over time. (sorry for the word soup) My goal for the next week: clean up what I've made into reusable effects for making things wiggly so our "dream world" can start feeling like a dream, and start experimenting with color using the patterns and wave manipulation I learned this week.

Bonus: I drew a cartoony dust poof animation to help make it more convincing when characters touch the floor when they run. We'll see further along if I should change the style on this one, but I'm liking it so far:
Note: videos in blogger aren't currently working for us. Stay tuned for the visual that's supposed to be right here.

CONTENT WITH HOURS

• Programming Meeting ( 2 hours )
• Level Design Meetings (10 hours)
• Art Meeting (2 hours)
• Story / Design Meeting (4 hours)
• Shaders Research (20 hours)
• Creating new Particle Effects (2 hours)

POSITIVE OUTCOMES

• We're starting to move into production, now that we've established our workflow(s), updated our software, and cleaned up our project files
• Our plans for progression and level design are becoming interesting! We've managed to develop strong ideas from several of our designers and blend those ideas into something all of us are excited to make.
• Having many more programmers has meant that I've had room to focus on learning how to do creative things with shaders!

NEGATIVE OUTCOMES

• A whole lot of our level and environment design is going to end up depending on my ability to produce particle effects and shaders to fit the environments - lighting and texture was a point of mixed success before and it's going to take even more learning to succeed.
• A lot of our workflow, especially re: lowpoly, texturing for toonshaders, and perforce, is going to be new to most of us, and that's going to require some extra learning time up front.

NEXT TASKS

I've learned a whole lot about shaders, and it's time to put that together into effects for game, especially improvements to the old ones.

TOTAL HOURS LOGGED THIS WEEK: 15  hours

TAGS

#{Jimmy Swanick}

#{Week 2}

#{Making Stuff Wobble}

#{Design}

#{Art}

#{Story}

#{Programming}

#{PPJ}