 |
|
|
click here to download our post mortem UBATU’S MASK POST MORTEM Ubatu’s
mask was a well scaled yet ambitious project taken on by the very first class of MOD team Students here at AIS. Ubatu’s
Mask is a third-person adventure-puzzle game for the PC utilizing the Unreal 3 engine. You are 8-year-old Tommy in search
of clues as to the whereabouts of your grandmother & mother after they mysteriously disappear in an old Victorian house.
Your adventure will span two worlds while never setting foot outside: one, in the real world; the other, in the Bogeyman monster
realm. The original Team consisted
of 19 Team members and Two instructors but as the two quarters progressed over 30 people and two different departments were
involved with this project. Originally
designed to be a three quarter long series class MOD team was cut to two quarters here at AIS (MOD Team I… and Mod
Team II). With high hopes we still decided to take on this ambitious project even though normal timetables were cut. In the
end the results were fantastic and we certainly learned a lot about a fairly large scale production process. Stats on the Game 23 weeks of production Over 9700 work hours committed to this project Aprox. 421 hours per week Over 200 Meshes Modeled and Textured 10 puzzles/quests 2 Fully Animated Characters 2 Hours of Game Play WHAT WENT WELL: Lots of quality Props: As I stated earlier over 200 meshes were made for Ubatu’s mask and a many of them
were at a professional quality level. The amount of content that some individuals cranked out while dealing with other classes,
jobs and families is a real credit to the individual artists and the school. Here at AIS we have been doing Level MOD’s for quite some time. While technology has developed
our ideas on game play have remained fixed on the first person shooter. It is important to attempt to try new Genres of game
play so we have a more rounded education. We took this idea and ran with it and made not just a level but a complete game
with a beginning middle and end. With puzzles being the cntral focus of the game play. Ubatu’s mask is one of the first
attempts from AIS to make a full nonviolent 3d computer game that is not just another FPS. A trend I hope future classes continue.
The team had decided to make a more stylized kind of game. We took a mix of Psyconaughts , Corpse
Bride and Team Fortress II and combined them into a visually appealing style. The
team that is assembled before you includes many of the most hardworking students here at AIS. Over the 23 weeks of this project
many gave more time to this project because the believed in the work that they were doing for it. On average 6 teammates would
commit to 9 full weekends to crunching on this project. These people would do anything that needed to be done to complete
the game on time at a high level of quality. Without their hard work the game would be nowhere as complete as it is today. With a team size floating around 20 members per quarter learning to work in a large team
environment was very eye opening. Learning to communicate effectively with this many people is challenging and while we had
problems we achieved more by relying on each other. The amount of cuts that were made to the game
has been minimal. Only2 rooms were cut from the original design and two slightly altered. The only other things cut were drag
and drop UI Puzzles (which was beyond our coding skills and timeframe) and cut scenes were dropped down to high quality Story
board Animatics. This project was full of large and small setbacks. The Dev team
and the Animator have become very familiar with the Unreal III engine over the course of constant bug fixing and crashing.
The team was relentless in tackling as many bugs as quickly as possible while putting a high priority on items that made the
most impact. WHAT CAN BE IMPROVED ON: At the end of Quarter 1, MOD II was no longer a required course. This caused us the most significant problem.
When it was announced that MODII was no longer required we had the departure of seven team members. While an attempt to fill
their positions in week 4 of MODII we were only partially able to replace them. The replacements while hard working were less
effective as well as less accountable to the project. The newcomers had the following problems: Lack of experience with tools and the team: Without prior working knowledge of most of the replacements
I had to take the time to evaluate their skill levels. The replacements had to be brought up to speed about the game and technical
specifications and most lacked any experience with the unreal III engine. While they were anxious to help most had very few
modeling classes under their belt. Without the appropriate skills I could not reasonably assign them all items of high importance.
Some Lack of Accountability As the contracts were laid
out only 2 contractors stayed on after 1 assignment (3 props each). Of the seven we received 1 also dropped the class. With
some creative effort overall we only effectively replaced the workload of 2 people instead of seven. This is because the normal
student was assigned 8 or nine props per quarter. Even
with some help cutting MODII as a requirement lowered my environment art staff by almost half. We had to reassign everyone
and rethink how we approached the game. This caused us to cut some aspects of the game like full animation cinematic scenes,
most particle effects, lack of life in the house and some sacrifice of texture detail. A more standardized work-flow process would have improved the quality of project
by 10 fold. The leadership team should have taken the time to standardize more of these procedures. Some of the specific items
that needed more of a standard or that needed to be implemented sooner were: Submission of Work More Logical Naming Conventions Creation of Textures After the
staffing cut after MOD I. The producer decided to simplify the schedule and assign specific rooms to individuals however by
simplifying things fell through the cracks that should have been taken care of. The Unreal III editor launched without the ability
to add custom characters. Even with a patch coming at the end of MOD I. No documentation was released by Epic games on how
to get one in game. Only by the sheer determination of Jeremy Lanum were we able to get them in the game. Because of the lack
of documentation and the patch not being installed on school computers; it took until week 7 of MOD II to get a custom character
fully functional in game. We had originally planned for about 5-10 more puzzles however we wanted a drag and drop functionality for
the UI in Unreal III This being our first attempt at this we found very little documentation on the UI editor itself and nothing
on how to achieve the types of puzzles we were looking for. After 6 weeks of attempting we gave up on these types of puzzles.
The art team had two leads for the first quarter
of the project. Both leads were well intentioned, but this was a mistake this caused some contradiction in the direction of
the project and took quite a while to hammer out the art direction. The
utilization of game flood was a good forum to share ideas & files but was taken too much for granted as a catch all. After
we eliminated the face to face leads meetings & implemented “everyone is responsible for a room”, we discovered
a lot of uncoordinated crossing of paths& wasted time/resources. Some serious
crashes happened during the last few weeks of the project. While some time constraints were out of our control more time should
have been planned for testing and bug fixing. Overall the
project is the most complete game seen here at AIS. Even if things did not meet our full expectations the team should be proud
of what we accomplished. We overcame tremendous odds to complete this game on time. Thank you for taking the time to read
this report if you have any further comments please email them to bear@jeremybearkim.com Jeremy “Bear” Kim Producer |
|
|