Project Evaluation

Our final game artefact, a trailer for a the game we proposed in our GDD's.

I am pleased with the above video, it took a great deal of work to even get that far. The original idea was just too complicated and we had to almost instantly tone it down. 2 months to make a functioning 3d game in UDK was just not enough time (especially as this was coupled with several other modules which we had to dedicate time to), so we lessened our scope to a more manageable level early on in the project. 

The module has been a great learning experience for me, I am now proficient in the basic functions of UDK and Kismet, which was probably my main goal at the beginning of the project. I began with almost no knowledge of UDK's functions and now feel confident when using the program and because of my experimentation with functions in the editor feel as though I can better understand any more complicated tutorial videos. I also got some experience in Photoshop and learned a few things in there as well. I had to make some of my own custom textures for the puzzle I designed as there weren't suitable ones within the content browser itself, I used Photoshop to make them to my specifications and then imported them into UDK to use, this taught me how the import functions work in the editor and I will be able to apply this knowledge in future. I also used Photoshop to design our game logo, this was interesting as I was able to further experiment with some of the functions and learn more that will help me in the future.

The decision by all of us to make a Dropbox folder for this at the start of the project is perhaps one of our most important. It allowed us to easily share work with one another over the course of the project and allowed us to swap ideas effectively in the initial stages. Using this kind of system, if you aren't all able to meet frequently is invaluable as without something like this we would have had less time to trade information and assets.

Working to a production schedule was very useful, the way we made it using Gantt project as a visual representation of how we should be spending our time made it easy to keep to deadlines and milestones throughout the course of the project. It also made it easy to co-ordinate between group members when assets had to be transferred between programs, for example when Greg had to use Ryan's Maya created models in the UDK environment. Because we knew when things would be ready, Greg knew exactly when he would have Ryan's models to import making the transition easy. This also came into effect when it came to rendering the videos for the final trailer, because we had finished each of our tasks early enough, we had plenty of time for video editing. The production schedule also helped with prioritising which parts needed completing first, and gave a clear direction for each group member to start a new task when one was finished.

The research and planning was all conducted during the first two weeks. This enabled us to have a clear idea of what we needed to create by the time it came to start production. The research gave us insights into the right ways of doing things in the our genre of choice, we were able to look at standout examples of games of a similar style and take inspiration from them, heading into our design process. The environment research we all conducted gave us all reference images and information which proved to be useful when designing the levels we intended to use. My personal research into puzzle design was invaluable to my own concepts of puzzles and then in the final one I made in UDK. It showed me how even simple puzzles could be effective within games.

I am overall happy with my puzzle design and it's implementation into the engine. I do regret some design choices however, I really think I could have made the puzzle easier to understand with just a few tweaks. I like how my vision for the puzzle is represented in our final trailer it seems to work exactly as I had envisioned it to in the first place even though the actual coding of the puzzle doesn't reflect this. It has piqued my interest in the UDK editor though and am eager to learn more over the summer so that I am in a good position to begin the main third year project.

It is disappointing that we had to scale down our original idea so much as I think it would have made for a fascinating game, time just wouldn't allow it. I'm happy with the trailer, but I just think that given more time we could have really done our initial ideas some justice. Still, it gives me hope for next year when we will have a whole year to expand on these ideas or just go for an entirely new approach using the skills gained over the foundation degree course.

This module allowed me to expand various skills that I have learned on the course over the past two years. I used skills from drawing modules (concepting my puzzle ideas, logo and storyboarding our trailer), work based learning modules (managing time effectively with the production schedule, allocated roles for each group member that are similar to actual in industry roles), game theory modules (invaluable in designing the puzzles), 3D modules (helped me with basic geometry functions to design my level in UDK and animation module helped with understanding UDK's matinee settings) and even managed to use some research skills I had gained from context modules (ability to sift through websites trying to find the right images that back up my ideas).  Expanding on all of these skills and managing to create an artefact that represents our original vision I would say has been a great success.

From UDK to our Trailer video

I had to get the level I had made from UDK into our trailer so I had to somehow record footage and export it as an AVI for use in a video editing suite at a later date.

I wanted the game to appear as though someone was playing it so I couldnt use the conventional camera export found within the program itself. Deadline was looming so I needed something of a quick fix to my problems.

I downloaded a program called camstudio which allowed me to select a region of the screen to record. I fiddled around with the options for what seemed like hours as I kept getting poor frame rate or poor quality footage. Each and every time I tried different capture and record settings until I finally found the right balance.

To record, because I am such a novice at UDK, I am unfamiliar with how to export the file as a level and the method I tried to do this just didn't work for me. I read forums to try and find some fix but couldn't find the right answer to my problem. I ended up recording the preview window that is accessible in UDK itself, and using camstudio to record that specific region of the screen.

I managed to get a suitable piece of footage and was really excited until I played it back for the 5th time and noticed that an error message from UDK displays all the way through the entire piece. As time is running out I may have to use this piece in our final video, I will endeavour to find a fix before then though.

But for the time being, here is my UDK level:

It still needs to be cut in a movie edit suite to remove the bits at the start and end, but could feature in our trailer as an example of gameplay.

First version of trailer after decompression and editing.

This isn't the full trailer for the game, I just wanted to have a go at making movie files myself. Ryan has created the actual trailer using his mac and Final Cut Pro. Which will appear in a later post.

 Here is the storyboard I have made to quickly design how the final video should look.

Logo

Here is the Lights title screen, I made in photoshop. It is quite a minimal design with a basic font with an underlay to give it depth. The canvas of stars is intended to reflect the otherwordly quality of the games design and narrative. A basic lens flare was added that sweeps across the entire title, it was intentionally minimal as we didn't want anything distracting attention from the word and the big light in the top left corner.

Underneath is my small concepts for the logo design, I have annotated each one and explained my choices for the final design here.

UDK Documentation

Using the basic tools on the left toolbar to create a cube.

Right clicking green cube icon on the left toolbar brings up properties so that I can select the exact parameters I wish, in this case I will also click the hollowed selection to make a room.

View of the newly created cube from the outside.

And from the inside.

With all the geometry built I now right click anywhere within the room and add a player start point.

I now right click on the surface and bring up the surface properties, I want the tiles to be larger so my custom textures will fit.

I open the content browser and import all my png image files I created in Photoshop to use here. When importing I made sure they had the "Create Material" box ticked.

Simply drag and drop the newly created materials onto any surface to apply them as a texture.

I finished texturing the level using materials from the UDK content browser.

The room with all my textures applied.

Now I add lighting, the same method as adding player start works here too.

I change all the light properties so each light that emanates from the symbols has a matching colour. To bring this properties window up I select the object and hit F4"

Adding triggers is the same process as adding lights and player start.

It's best that the trigger area is scaled up, so it isnt too difficult to find the right place to activate the switch. To do this I bring up the properties and change the collision height and radius in the collision properties.

The last part of setting up the room is adding a static mesh. I chose one from the content browser to fill the gap here and later animate out of the way when the puzzle is solved.

Change the static mesh's properties so that it fills the gap.

The most important part, and one that I won't forget again. Convert the static mesh into a "Mover" so it will be able to animate later.

I now open up Kismet, In the preview window I select the trigger I want to use and right click within kismet to add it.

Right Click and add matinee, this will be what powers our animation.

After double clicking on the matinee box in the kismet window, the matinee editor opens. The first thing I do here is add a new empty group. This could be a whole group of animations, but I will only be using one.

Then I right click the new group and add a new movement track.

The green triangles influence the length of the animation with the left being the start and right being the end. The middle black line icon with white line above indicates what point through the animation it is. For this purpose I will animate for 5 seconds. To do this I move the black line icon all the way to the very end of the animation point.

With the animation scrubber in the correct place I move the static mesh to its final point in the animation.

With these two elements complete I then went back into kismet and linked the trigger to the matinee, this would mean that when in game, the animation would play when the player activated the switch.

I was totally unable to make a full 4 switch puzzle the way I had wanted. It was just too much to learn in such a short space of time, that I am still trying to get to grips with. So at the moment it is coded with a kind of placebo effect in that the last switch only needs to be activated to open the door.

I really overlooked some things in the design of this puzzle, for a start I think the clue I have written for it is far too obvious. Additions like the lights going off or changing when they are activated would have made the puzzle more easy to understand too.

Overall, it's been an interesting learning experience, and one that I'm sure will help me with the final project of the third year. When I think that two months ago, I barely knew how to do anything with this program, I think I have learned a lot.

UDK Puzzle Design

So, I now need to figure out how to take the ideas for puzzles from the page and into the engine. I had little to no experience with UDK before this project so have had to learn as much as I possibly could to make the puzzles in a short space of time.

The first challenge I encountered was how to use Kismet in UDK to make switches activatable and for the switches to perform a triggered action. I did a little experiment with trying to create a useable light switch in game.

With this complete, I set about trying to use the same principles to make an object move after activating a switch. If i can work this out, then one useable puzzle will be available for people trying our game at the summer show.


I also had to work out how to make UDK play Ryan's music and sound effects at the correct times, because the tutorial videos I had watched for switches had covered this, it was an easier job than learning how to make a four switch puzzle!
 

PRE-PRODUCTION & RESEARCH DOCUMENT

Below is a link to my pre production and research document:

Pre Production and Research Document

GDD

Below is a link to my GDD for this project:

Lights GDD