Tuesday, 30 April 2013

Hindsight is a glorious thing

Well, this year seems to have flown by but looking back gives me a chance to assess the work that I've done and all the things I've learned. I'm probably a pain in the proverbial  when it comes to what the college is trying to assess and what i want to achieve in terms of my own goals. I guess i had a very specific idea of the path I wanted to follow, when I started, but the wide variety of tasks and briefs I've been set upon has broadened my skillset and made me realise where my strengths and weaknesses lie and what i need to concentrate on to grow as a Games Designer.


Concept art


My first project was always going to be the toughest, coming straight from a long (dull) career in retail/hospitality I was always going to struggle. Over the years I'd left my drawing/illustration/sketching skills to suffer and due to lack of practice my percieved ability of my previous skills didn't match up to what i could achieve in practice now.

For this brief i had to create thumbnails for both environment and character concepts for video game ideas. I wasn't short on ideas coming into the module, but presenting those ideas required a back to the drawing board approach. It's important to show the design process in its entirety, it allows the viewer to see the thought processes that i went through, laid bare.

While designing my character (Kung-Fu Drood) my initial sketches were stiff and didn't represent the fluid movement that i wanted to achieve in terms of game animation. After a suggestion by my tutor that i should research the work of Burne Hogarth and his 'Dynamic Figure drawing' it helped me achieve the poses and therefore movement of the character that i was trying to convey.

In terms of the environment i wanted to use the software Vue, as I'd seen it used for creating impressive environments in 3D, and it was something i really wanted to get to grips with. It's a great piece of software for creating ecosystems/biomes and in combination with World Builder to create the terrain the results looked promising. However creating good renders from Vue was something i struggled with so fell back into 3DS Max as it's software that i was more comfortable using. Vue is something i really want to come back to and try to incorporate it into my workflow.

Once the terrain was modeled it was relatively easy to build a small settlement relating to the initial Celtic/Asian style designs that i was going for, but struggled in terms of texturing and rendering, points that are weak in my skillset at the moment.

I also painted the line work of my character designs up in Photoshop, building up a series of highlights and shadows using a 17th Century technique used by the 'Dutch Masters'.

However as with most of my endeavors my ambition for what i was trying to create outstripped my ability and time constraints. Although happy with my character and environment designs i really wanted to do more, taking them into a game engine would have been the ideal outcome.


Sequencial Imaging


With my designs from the concept art module i was asked to produce an in-game cutscene, focusing on camera movement, timing, and mood.

To achieve this we had to create storyboards that were then turned into an animatic (functional animation). Again i had big plans for this project, concentrating on the background story of my character and to set up the game.

I got a chance to play about with video editing software Final Cut and Adobe Premier to create my animatic, and really loved doing the sound effects, but to take my animation sequence further i was asked to recreate it in 3D.

Always up for a challenge i worked on the Wicker Man in Max using Ivy-Gen to build the model and then delved into the wonderful world of the 3DS Max Character Animation Toolkit (CAT) to animate it. However things started to become unstuck due to the meshes massive polygon count, I'd never intended the Wicker Man to be animated, it was more of a concept piece to show what it could look like, so no pre planning had gone into it in terms of polygon budgets.

I was really impressed with CAT however, and strived to teach myself more about the software, as it seemed like a huge leap forward in terms of character animation, compared to packages I'd seen before.

I also needed my main character and repeatedly failed to get the style i wanted using 3DS Max so my tutor recommended Silo 2 for a more organic style of modelling. I created a stylised version of my character concept that looked perfect for my game style.

The terrain I'd built previously really helped when it came to producing the scenery for the animation but again I'd over extended myself. I felt i was teaching myself so much, but in doing so found it increasingly difficult to bring all the elements together to create a finished product.




Animation for Games


For this module i was asked to create a Game Animation specifically focusing on Pseudo 3D/2.5D techniques. My research led me to look at the many techniques used over the years take create 3D effects in 2D such as parallax scrolling, isometric viewports, ray casting, and sprite sheets.

I wanted to get my Druid character animated so headed back to CAT to teach myself more about the software and how to get the look and feel of his animation. I spent the time coming to grips with weight painting and animating, realizing that if this were to be a game model to needs more work in terms of optimisation of the mesh.

It was amazing to look at the history of game animation and discover the extensive work that went into creating sprite sheets of previous generation titles. This summer i'd like to create a working sprite sheet of my own including all the attack and jump animations for the character.

I was really happy with the character and animation, and received very positive feedback from tutors and fellow students. My ultimate goal was to get the character in a parallax scrolling environment scene, and although set up, time constraints meant this wasn't possible.
I think that turned out to be my weakness again, the ambition i had for the design was unobtainable within the set time frame. I need to simplify my workflow and set myself small achievable targets for the week rather than getting bogged down by the scale of the task.


Digital Skills/character design


This task was to learn new skills and use them in a brief. To create a character i did a lot of research into character creation (way too much!). The research focused on archetypes  and the hero's journey that establishes common tropes in game characters and story models.

We had to create a character turnaround sheet and this was where things were a little tricky for me as my drawing skills were still lacking. I'm really just experimenting and trying to find my own style.

I was starting to get a little frustrated as i felt i was treading old ground creating and animating another character which I'd done over the past few modules, but practice makes perfect! I'd never done a character turnaround sheet before and i loved using Adobe Illustrator to ink up my sketches. It was pretty easy to model the character based on the turnaround sheet so i gave myself the extra task of creating an organic character based on the design concepts.

I returned to using Silo to model the organic design, and was pretty pleased with the results, the mesh was deliberately kept quite low density, taking what I'd learnt from building and animating Kung-Fu Drood. I got more of a chance to use CAT for my animation cycles but I need to work on the gait of the character to suggest bulk/weight. It was a good deal easier to do though with the skills I'd picked up from the previous tasks.

The character really needed more detailing such as clothing (Bruce Willis - Die hard, style) and I realised this could be achieved using ZBrush, but with only a week to go until the deadline and without anyone to help me with some ZBrush tuition, this was just not possible. ZBrush is industry standard software these days so being on a Games Design course where this isn't part of the curriculum is a little worrying. It's something I'll try to get back into over the summer. 


Life Drawing


This was a module i was pretty excited about, I'd been a confident artist many years previously and my drawing skills had suffered due to a lack of practice. I was hoping the module would help improve my sketching and anatomy skills which in turn would help my confidence and feedback into my modeling and design skills.

We used a variety of techniques and tasks to allow us to concentrate on form and line such as; drawing while not being able to see where we are drawing, timed drawing sometimes no longer than ten seconds, drawing with a long stick, experimenting with different media, chalk, ink, pencil and charcoal. It was all good practice and enjoyable, helping break down some of my confidence issues around my perceived drawing skills.

I created a few images that I'm somewhat happy with, but it was much better as an exercise in experimentation. Foreshortening was the biggest stumbling block for me, achieving realistic proportions even after measuring, still tripped me up a little. As the weeks went by I did feel I was improving but it really was just a taster session. The second year students were able to join in with our sessions so I'm hoping to be able to join in with the life drawing sessions next year.



Work based/Environment Concept art


This module focused on working on a live brief. I was asked to design concepts for the BBCs' 'What If?' competition, which required entrants to design concept images based on their ideas of the future. It was a chance for me to practice my environment modeling and more specifically concentrating on texture work, as this is the area I feel was my weakest.

The module started with a mind-mapping session, without initial ideas it's easy to loose focus, so i sourced imagery and reference related to my chosen environment type. Inspired by my research I became very interested in the 1950's/60's style of the Googie and futurist movement. It was the atomic age where wild ideas of what the future may look like were fueled by the imagination of what atomic research promised. It was a very experimental era in terms of style and architechture, instantly recognisable for its optimistic, space age feel, but by today's standards looks a little naive, especially as we know the dark era of uncertainty that engulfed the world due to nuclear proliferation. Games such as Fallout use this era to base their narrative on, a world that destroys itself during the atomic age and what has happened to the remaining population that have had to scavenge, and cobble together the technology of the day to survive.

I discovered images of early nuclear testing that was occurring just outside the burgeoning Las Vegas strip in the 50's, with mushroom clouds towering above the mecca to capitalism and excess. I wanted to take this scene further to suggest what may have occurred if the Cold War had escalated into global nuclear warfare, or what effect the experimentation of the era may have on the environment. To me the mushroom cloud in the images looked like an enlarged tree, possibly mutated by nuclear fallout.

This was also a great project to experiment with texturing, i could model simple buildings and signs and use the textures to enhance them.. After setting up the simple shapes to represent the buildings i realised I needed some vehicles in the scene to ground it in reality. Using blueprints of a 1959 Greyhound coach, I produced an asset that I was really happy with and that could act as a 'Hero' object in the scene. I think most of my efforts went into the composition of the scene, trying to set the image up so the viewer was led through it and keep the focus where i wanted people to concentrate on.

Again the task was a great learning experience, I taught myself a huge amount in terms of texturing, lighting, and rendering, and am happy with the 3D side of the image as a piece of environment concept art. Due to the work I put into the image in 3D, I left myself very little time to experiment with over-painting the scene in Photoshop, which i feel it needs to soften the feel, and tie the elements in it together. 


Vehicle Design for Video Games


I was pretty excited about this module as I felt it fit with the skillset (building mechanical models) I  had before starting college. In the brief i was tasked with researching the top five, best selling games, of 2012, and the top five Indie games. Of course i had to refine this further to games that featured vehicles to study current design trends in games. This led me back to the work of Syd Mead, futurist, designer, and illustrator, who worked on vehicle and environment designs for films such as Tron, Blade Runner, and Aliens. I've always respected his work and it was clear he has a strong influence in the video game industry, the Mako and Normandy vehicles from Mass Effect are almost direct copies of vehicle illustrations I've seen in Syd's' work. 

The area i felt i had to work on the most in terms of the brief was presentation, it asked for a turntable render of the model, wireframe, ambient occlusion, and unwrapped textures. Initially I wanted to do a more organic design, based around the research I'd done into Halo and Starcraft. Each have factions that are insect-like in nature and i found the modern military, vehicle designs of games such as Call of Duty to be relatively dull. Sure it appealed to the little boy inside me who used to build, paint, and then hang his Airfix kits from the ceiling, but i wanted to push my own design ideas in a different direction. I came up with an alien vehicle based around the proportions of a wasp that i was pretty happy with and could easily model but with no one to show me the steps fro unwrapping and texturing my models i felt it wasn't the easiest shape to work with.

I continued to do a bit more research and kept going back to the designs of Syd Mead and his presentation style. Inspired by that and wanting to produce a model in a similar fashion i created a Tron inspired light cycle, purely for research purposes but also to practice a turntable render presentation. I was pretty happy with the results and it only took a couple of days, one of which was for the rendering due to screwing up the first render by not realising the glow effect I'd added was a post-processing effect that was added after each frame had rendered. At the setting i had it at it completely obliterated the bike. After fixing the glow, it's a really subtle but cool effect as you see it glistening off the edges of the metal as the bike turns in relation to the camera angle.

I used to love perspective drawing when i was at school so to practice the techniques i sketched a quick tank design, that could be used as the human faction in opposition to the insect-like designs. I thought this would also make for a much easier task to model, unwrap, and then texture compared to the organic nature of the opposing design. I really wanted this to be a model that at some point i could have running in a game engine so restricted myself to a very low polygon budget. Admittedly looking back I probably was a bit strict with the limit, it needed a bit more smoothing around some of the more angular edges, but that inspired a more high concept design that I'm really pleased with. Before I got to that point though, I used the wonderful texture set i had to help enhance the model. By using the textures I had, I could manipulate them around areas of the model to suggest extra detailing that otherwise i would have normally modeled. Overall it would make for a great model for use in a RTS game where players often control fleets of vehicles from a top down perspective.

As i was texturing the model I had the idea (based on my previous research) to turn it into a more high concept, ink and paint design. From this design the idea for my Rally Tank 2150 game was born. The turret section could move back and forth to dampen the recoil as the tank fires as it's racing, enhancing the design choices and giving it a realistic, it could work, feel.

No comments:

Post a Comment