Post PRoduction

This is a weekly breakdown of my work as I wrap up my senior year at Savannah College of Art and Design. I've added two scenes to my reel: an update to the sword scene and a new plasma cannon effect.

Week 1

I'm starting off the quarter by opening up all my old projects and cleaning them up. Consolidating all my files, organizing my assets (geometry, textures, animations, etc.) into appropriate folders and deleting any redundancies. Once I've cleaned all of this up, I can start planning my work around it.

The sword scene shows a lot of my strengths, but it lacks climax and urgency. I'm going to plan a clash of magic between the sword and the coffin. The sword acting as a seal for evil forces inside. Once it exerts its power to keep the evil away, it disintegrates.

Week 2

I'm happy to say that my previous work on this scene was pretty thought out in terms of longevity and reuse. All of the important events aren't hard coded together, and just need to be called by a level sequence. I'm clearing out old sequences, organizing all their assets, and taking notes on the required parameters for animation.

On the side, I'm starting to create new magic shaders to put on FX geometry in the level surrounding and clashing with the sword. The color scheme is up in the air for the "necrotic" power, but the sword is definitely a hot, sunny fire. I'm adding some edits to my UV distortion function for expanded use.

Week 3

I need to up-res my textures. Some of them were used for art tests and were optimized for gameplay. This, being a real-time cinematic, isn't going to be as intensive. I can be a bit more liberal. We're talking about an upgrade from 1k to 2k. Nothing drastic. There are some artifacts in the shaders beyond what happens from major UV distortion and coding.

Luckily, that cleared things up. Usually you can just scale up the textures, tiling them to get a better resolution. But at the scale I was working with, the repetitions in the texture was obvious and amateur.

Week 4

The geometry is in, and the animations are coming together once they're all plugged into the sequence. However, this week I'm starting to run into some trouble with Sequencer's keyframing. It tries to be a bit too user friendly and ends up making a lot of leaps in value, causing the camera to do full 360s and the like.

Now that I'm starting to see everything come together, I'm not terribly thrilled by the results. The magic needs more of a source. I also received critique from the art director at Sparkypants Studios and he mentioned a lot of things that need improvement in my environment, which I agreed with. I'm going to wrap up these shader animations and add some particles and see how far they take me.

Midterm

The project was well received in class, but I'm not happy with it. It's hardly an improvement from where it was, and I think shows more flaws than strengths. I'm going to shelf this project for when I have more time to overhaul the look into something stronger overall and design the clash of magic more thoroughly.

Now, I'm moving on to my next project. I'm designing a more modern weapon demonstration scene to act as another flagship for my demo reel. Much like the sword scene, it will demonstrate overall look development, my grasp of a wide variety of software, and some new skills like rigging.

Week 6

I'm at a crossroads. I need to decide between working on a stylized or realistic render for this weapon demonstration. Most of my work is closer to realism than stylized, but they all use shader manipulation. I don't demonstrate that much flipbook usage. On the other hand, I have no stylized work, and I've had that requested a couple of times from various studios.

For now, I'm designing the weapon types for each. A rather simplistic rocket launcher turret and a more detailed gauss cannon turret. Both of these are modelled to be rigged in the future and animated for their introduction to the scene or even for each firing sequence.

Week 7

I'm beginning the rigging process on the geometry for the turrets, and I admit that it's a bit sloppy, this being my first time and all. I'm having to reorganize groups, combine and seperate polygons, and arrange things differently than I would for your standard prop. I'm looking forward to exporting this for Unreal and studying that pipeline.

Week 8

I exported out a stand in mesh so that I can focus on the FX for the remainder of the quarter. My other work is taking a lot of time and I'm having to adjust the scope of the project. I'm playing around with After Effects again and I'm seeing how powerful it can be for FX alphas.

I also recognized the solution to a common issue I was having when using my noise textures from Photoshop. Photoshop renders out the Perlin noise as seamless, yet I was getting seams in some of my shaders. The reason was that I was scaling the textures at the center of the UVs as opposed to the origin at the bottom left. I also needed to scale by even multiples of 2, to ensure the resolution would line up with itself when tiled within the UVs.

week 9

I'm experimenting with using depth fade to give the illusion of lightning arcs clashing with surfaces. By inverting what is normally transparency used to fade out particles, I'm able to highlight the collision the particles have with geometry as a hot spot in the lightning bolt. I may need to experiment with using actual geometry as opposed to particles, to ensure that the pivot point starts at its spawn and goes out from there. Otherwise, the lightning will spawn centered at the middle of the arc - this will look like trash for surface interaction or controlling the source of the arc's power.

Finals

Unfortunately I had some family issues to return home to, so I lost all my work hours over Thanksgiving break. Everything is a bit under-developed. However, I have the controller set up and the FX all timed out with some fun experiments showing that they paid off. It will be easy to move forward with this project and continue polishing the effects.