Evaluation

My overall impression of this module hasn’t always been positive. I’ve felt very much lost for large portions of it, and the bits I enjoyed seemed all too brief. Part of that probably stems from my own desire to be more specialized. One of the goals of this project seems to have been to give us a firsthand sense of the production pipeline used in industry. And while it’s certainly been interesting to understand a little better what challenges an animator, or a level designer might face, I came onto this course already fairly clear on the role I wanted to function in; pre-production & concept art. It can be frustrating to have to devote the majority of my time to areas that don’t interest me. More so because my inexperience in those areas tends to mean they take the longest. I suppose my biggest wish for this project would have been to allow us to work in teams. It would have been a closer approximation of what happens in industry, allowed for more specialization, and given us a greater scope for what we could achieve in the time frame. I also feel like collaboration brings out a better work ethic in me.

The brief started out with us designing a character. This was something I enjoyed. I don’t do nearly enough character work in my own time and it’s something I need to practice. I enjoyed coming up with a narrative and backstory for my character as well as deciding on the visual aspects of his design.

We also used the first couple of weeks as an introduction to Unity. I feel like this was too early, by the time most of us needed to work in Unity most of us had forgotten what we learned here. What I think would have been better would have been either more design sessions, or a more intensive reintroduction to Maya. Possibly beginning to model and UV map some of our environment props as practice for our character. Unity could have waited until after half term when we had actual models to import.

I also wish the design crit had been earlier, or perhaps the hand in date later. 2 days was nowhere near enough time between the two to make any major changes in response to feedback. I also feel like the brief’s stated goal of a rigged, animated, and textured character by that date was completely unreasonable. At that point our lessons in Maya had barely advanced beyond texturing cubes. To my knowledge most people, including those with prior experience in Maya, didn’t reach that point until nearly a month later.

After reading week we started modeling in earnest. Modeling a human is a big jump from our previous work in Maya, but I found it the challenge fulfilling and progress gratifying. My final model ended up a little basic, I wish I’d taken the time to attempt to model the face. However given my experience in Maya I was fairly happy with it.

UV mapping is where everything fell apart for me. I spent more than a week trying to unwrap my model’s UVs. Mistakes I’d made in modeling caused me so many problems. This is where I feel more time modeling props in Maya would have paid off. it’s hard to understand these kind of mistakes until you’ve already made them, you might not even notice them until later down the line. It can be a domino effect, and unless you can identify and fix mistakes early they just snowball. Eventually I ended up skipping parts of UV mapping to move on to rigging. When I came back to it later I just used planar maps to finish off, which hurt my texturing.

Rigging is an area where I feel it would have benefited to have more hands on advise from Matt. Following a tutorial can teach you the basic steps, but by necessity most people had to deviate from them to some extent to incorporate their designs, often leading to mistakes. Just following rote instructions it’s easy to make a decision without understanding it’s significance.

Controls and weight painting were both intuitive enough that I felt happy enough just learning from the tutorials. I wish Matt had used a more complex demo model though, to show us the kinds of issues we might run into there. But that’s a minor gripe.

Animating is an area I know a lot of people want to specialize in, and I feel sorry that got so little time to practice it in this module. Personally I was just happy to get it out of the way. I feel my animations were somewhat rushed, and I don’t much of a sense of timing. For a first attempt though, I’m happy just to have them work.

Texturing was something I enjoyed a lot. It was just painting without the design aspect. I found it relaxing just to be able to switch to autopilot and detail mindlessly to some music. It also adds so much to the final look of the game.

Building in Unity was something I didn’t expect to enjoy as much as I did. Scene composition, lighting, color, zooming in and adding little details, it’s a lot like painting in same ways, but with a greater element of problem solving. It was also gratifying to be able to put into use some of the tricks I’ve seen used in actual games.

I think my likes and dislikes with this module all come back to my own specialization. I enjoyed the areas that related to it and disliked the areas that deviated from it. It brings up a concern I’ve had for a while that this course may be too broad. 3 years isn’t enough to become an expert in everything this course aims to teach. Looking at industry I feel pressured to be working hard at my own specialism to reach that level, and anything that distracts from that just feels like a step backwards. Having a broad skillset is definitely an advantage, but I feel my individual skills are suffering for it.

 

Advertisements

Playtester feedback

As a way to round out the project and as a kind of final crit we spent an afternoon playtesting eachother’s levels. This is the feedback I received:

  • Nice view – Water is a bit choppy
  • Bottom of the shield is going through the floor, rotate a bit to sit on a flat edge?
  • Light source for in dark? Lantern?
  • 2D texture on coat hanger
  • You have modeled the other rooms but we can not get in them
  • Alter the other rooms so they don’t look copy/pasted
  • Wardrobe is kinda see through
  • Shame about the terrain edges
  • Cool environment – Static character
  • The wardrobe is transparent a little?
  • Terrain is really good – Adds to the concept
  • Scenery vista needs a focal point – Sun, mountain, another tower
  • Bird sound doesn’t work this high up in the mountains
  • Excellent landscape
  • Like the fire on the staff
  • The character should turn around a little more

Putting everything together in Unity

The final steps for turning my game assets into a functioning scene all took place in Unity. I find Unity a lot easier to wrap my head around than Maya, but it still has it’s share of frustrations.

First off I imported all my models and adjusted the scale so they measured up. Importing my character is something I had issues with before, but after getting help from Matt it worked flawlessly this time. I reapplied the textured that hadn’t been imported with the model for whatever reason, and played around with the shader settings to make my coathangers display transparency correctly. I then set a skydome and brought in a directional light to cast some shadows (which unfortunately don’t show up in my version at home). The interior at that point felt too dark so I created a second point light and set it to a low intensity inside the room.

With the room and balcony mostly sorted I used Unity’s terrain tool to start crafting the landscape. I used the default grass texture for most of the landscape, but I applied patches of snow to the various mountain peaks. I set a gentle blue-ish fog to give the impression of atmospheric perspective, and used a particle effect to add extra fog between the mountains. I made some quick fir trees in Maya as a spur of the moment addition. The were very simple, just a low poly cylinder tapered towards the base with planes facing upwards arranged around the outside. I applied a leaf texture to the planes and imported them to Unity. Up close they  aren’t much to look at, but from the balcony they’re convincing enough.

It was at this point, some time in, that I realized placing the balcony in the corner of the terrain may have been a bad idea. Looking to the side a player could easily see the edge of the world. After the work I’d put into sculpting the terrain for that viewing angle I was reluctant to simply move the balcony further in. Instead I decided to try and obscure the edges with more of the fog particle effect I’d used earlier. I hoped it would also give the impression that the player was looking down from above the cloud layer. I also cloned the balcony and added duplicates on either side to try and imply a larger building.

I then started trying to get my character to animate. I split up the timeline I’d imported with the mesh and named each segment after it’s animation. Annabeth had shown us how to use a box collider and a piece of script to make our characters swap between two animations in response to the player. However I couldn’t get mine to work. After asking Annabeth for help she diagnosed that I had applied the animation to the character mesh and not the root.Following a quick fix my character now switches between leaning on the balcony, and turning to face the player, whenever the controller approaches the window.

Lastly I added some ambient bird song from the BBC sound library.

I now have a functioning game environment. It’s small, crude, and barely interactive, but it’s a big step up from animating toy cars I feel.

Environment textures

I think this has been my favorite part of the project.Texturing is just mindless detail work, with none of the design pressures involved in concept art, and yet still well within my preferred skillset (Unlike Maya). Like with my character I went with entirely hand drawn textures rather than using photos as a base. I took colors directly from my concept art to make sure everything looked consistent.

I experimented with bump mapping too, but it just looked odd in my test runs. Maybe I’m missing something.

Environment modeling

My environment was very straightforward to model. I knew from the beginning that I wanted it to feel small and cluttered, and the room itself is made of mostly very simple shapes.

Dealing with UVs was also a lot easier here than with the character. Since you’ll only see the room from the inside once in Unity, planar maps are sufficient for most of the surfaces. I made a point of unwrapping the pillar first, since it repeats a lot in my design. By unwrapping it just once, then duplicating it, I only need one UV map and texture for all iterations of it.

The planes hanging from the front are where I plan to texture on several coat hangers. It seemed easier and more efficient than modeling them, but we’ll see how well it works.

Animating

While far from easy, animating is at least intuitive.

To start it’s best to find or record some reference for whatever you’re animating. I used a walk sequence I found online and footage of myself jumping as starting points. If the footage is the same framerate as your animation (in this case 24fps) then you can insert it as an animated image frame and copy the poses directly in Maya.

All the character’s animations go on the same timeline. Later you can split them up in Unity and assign them to play at specific points.

There’s not much to really say about the animation stage of things. I didn’t run into any major difficulties, it’s just very time consuming.

Skyrim

As you may have picked up on already I’m a big fan of the Elder Scrolls V: Skyrim. One of the strongest elements about the game, I feel, is it’s visual presentation. The game is beautiful.

To be clear I’m not talking about the graphics here. The term “graphics” gets thrown about a lot in regards to games, but essentially it’s the technical fidelity of things. The technology used to display the game world. On a technical level Skyrim isn’t that much more advanced than it’s predecessor Oblivion. The models may be more detailed, the lighting more advanced, and the textures higher resolution, but those aren’t what make the game gorgeous in my opinion. Modders will frequently buff up the technical areas of a game, but without applying a consistent art style to things it rarely makes it look much better.

I’ve also been trying to keep an eye out while playing for the different techniques we’ve been learning in this module and how they are applied in a full scale release.

The still water effects are similar to Oblivion’s. It’s just an animated plane similar to the one included in Unity’s standard assets.

Mist effects are just overlapping particles. I used the same kind of principle to create clouds for my character turntable.

Magic is also done with particles, but in this case it has multiple emitters overlapping to create this fireball. One for the Main body of flames, another to create sparks inside, a filter that distorts the scenery behind it to give the impression of heat rising, and a point light that effects the hands.

Magic in Skyrim generally feels more powerful in Skyrim than is Oblivion, despite having equal/lower damage, and smaller particle effects. The reason I believe is to do with the small charge up time introduced in Skyrim. It’s not much, barely a second, but it gives the projectiles more weight than the instant cast in Oblivion.

The show Extra Credits describes this  this curve of buildup to release more eloquently than I can in their episode on pacing: http://penny-arcade.com/patv/episode/episode-07-pacing

Foliage is just a series of plans with a texture, alpha channel, and bump map. This is particularly evident when the game’s textures bug out leaving just the geometry and normal maps visible.

Flowing water uses animated textures over non collider geometry. Several kinds of water effects can be layered together to create different effects but essentially it all comes down to well designed textures. In fact very few of the details in the game are actually modeled, the textures are responsible for so much.

The developers of the games have gone on record saying that Skyrim’s world map is approximately the same size as Oblivion’s. In fact if you use the dev console to break out of the game’s boundaries you can even see the height maps for the previous game worlds.  Yet Skyrim often feels bigger. Part of this is the way the game conveys distance. In Oblivion, if you climbed to the to of the tallest mountain you could see as far as your render settings would allow. In Skyrim fog reduces your vision of distant areas. This atmospheric perspective helps sell the illusion of scale. It also reduces the amount of the terrain the hardware needs to render.