|
 Features
Postmortem:
Pixelogic's The Italian Job
Pixelogic
Games was founded in 1996 and have been developing successful Playstation
Games for over 6 years. Under the management of Bryan Reynolds, Pixelogic
has successfully designed and completed three original PSOne titles on
time and within budget, our goal is to carry on creating many more great
original 'AAA' games.
 Armed
with a cult film licence and knowing that The Italian Job (TIJ)
already had a huge fan base which would mainly interest the general UK
public, we could only hope we could produce a game that would not only
do the film justice by not disappointing any fans and also come out with
a game worth playing.
When developing
any title, there are always tendencies to make the game-play too difficult
& hardcore, but due to the nature of this title, a decision was made
to go down the pick up & play route so the game was more accessible
to a larger audience. There are so many serious driving games out on the
shelves that cater for the hardcore racing fan; we wanted The Italian
Job to be a driving game with a difference rather than just an all
out racer.
What Went Right
1. The
Getaway The most important part of the game had to be "The Getaway"
sequence; to re-create all the stunts from the film was going to be our
biggest challenge, the aging PSX is going to have to work overtime to
display three detailed Minis, police cars, pedestrians, and general street
furniture in a grid locked Turin.
Development of our in-house Continuous Ordered Scenery Streaming (COSS)
technology enabled continuous loading of level data that allowed for huge
playable cities for the player to freely drive around in. Getting this
technology in place early on was a major factor on how the whole development
of project would run and how the finished game would play. Once this technology
was proven, the artists could then confidently begin to build the vast
environments required such as London, Turin and "The Escape Route".
The design and planning that went into "The Escape Route" was
probably the best design of all the levels. The art team had the job of
re-creating Turin, they started by mapping out the general road layout
before positioning all-important landmarks and escape route scenes from
the film, then all the extra bits from the movie that were relevant were
added and the art team tried to include all of them.
There were so many unknowns to deal with while implementing the actual
mission as not everything can be totally designed within a document so
one of our favourite techniques which we spend a lot of time on is to
design, create and judge game-play 'on-the-fly' during development. Certain
memorable scenes of the film would not have necessarily translated to
good gameplay, the overall number of scenes in the film was vast and included
hiding in car parks, outwitting police on rooftops, and driving down church
steps. When analysing the actual film, these scenes of the escape had
no connection to each other, it was just a series of stunts performed
in minis edited together and most of the stunt sequences weren't even
filmed in Turin. We had to fill in the gaps in-between each stunt and
scene location to make the whole escape route work as one complete mission.
Due to the length of the final route, our initial idea was to break it
up into three sections but after more testing and prototyping it was decided
that the player was going to have to drive the whole escape route in one
sitting in order to recreate the heavy tension from the film of getting
out of Turin in three Minis. Carefully positioned cameras were used to
capture main action sequences and cinematic scenes while trying to keep
it true to the style of the film; the pacing of the whole escape had to
be right so numerous cameras were created, moved, removed, recreated and
tweaked until the pace of the whole mission began to flow and the final
gameplay began to feel like the movie.
Once the route was finalised, the next job was to populate it with a grid
locked traffic jam. We ended up having to create a completely new level
based on Turin with all the unused parts of the city removed since the
amount of polygons onscreen was beginning to grind the game to a halt.
As more objects were added, the escape route had to be constantly optimised
right through to the end of the project to ensure a good frame rate.
The final Getaway mission consisted of over six minutes of constant gameplay
full of real time camera cuts and it was worth all the hard effort; this
was going to be our flagship mission so it had to impress. We knew that
anyone who plays TIJ will be waiting to get to this famous chase
sequence and we could not afford to disappoint, we all still enjoy playing
this mission now which must mean something about the amount of effort
and time that went into to producing it.
2. Voiceovers. About halfway into the project, everything was beginning
to come together; we were starting to have a good playable game on our
hands and were ready to start implementing speech to set the scene for
all the missions. Our publishers managed to get hold of Phil Cornwell
who has a huge repertoire of impersonations under his belt, mainly know
from "Stella Street", a popular comedy sketch series in the UK. Phil also
does quite a good impression of the lead character 'Charlie Crocker' from
the film. It was important that we had a decent voice for 'Charlie Crocker'
as we knew fans of the film would have moaned if this voice was not right.
The voice scripts for the whole game were decided quite early within the
initial design document by Bryan and myself and were only slightly reedited
to fit into the finished missions. More importantly, Bryan went down to
London's famous Air Studios for the voiceover recordings and since he
wrote most of the scripts himself, he could therefore explain the context
of the scripts and exactly how we intended the lines to be spoken. Also,
the foresight on the day to include extra one-liners and generally allowing
the voice actors to have a bit of fun and freedom while recording meant
we ended up with more speech samples than we bargained for. Some heavy
editing was required to fit over 400 long speech samples into memory with
all the other general SFX. The extra resources inspired the designers
to add extra speech to missions where appropriate and gave a more involving
humour throughout game; Lorna chattering about her shopping in the "Mafia
Mania" mission is a good example of this.
3. Where the crow flies. Our decision to guide the player by only
giving them an arrow that pointed in the general direction of each mission
goal was derived after much iteration.
Some people like to use maps and some don't and the inclusion of an onscreen
map was thrown out due to several other reasons, we were already pushing
the onscreen polygon count so it would have caused unwanted slowdown and
after watching more and more people play games which included a map, we
discovered they generally distracted the player from concentrating on
their driving. We wanted to create an instant pick-up and play environment.
We also had grand ideas of subtly colour coding roads and buildings as
you drove past but eventually introduced the current arrow system which
pointed the player in the direction of 'where the crow flies' This arrow
system seemed a bit too simple so we wanted to try and improve upon it
and made the arrow point down each individual road, showing the shortest
route to any given destination. This new arrow was in place for quite
a few months and the game just didn't feel right. People complained about
how linear it felt and even though time limits were quite generous and
the arrow automatically adjusted if the player missed a turn, the player
still thought they were playing the game wrong if they missed any of the
turns. Our decision to return back to the original arrow design was the
only way to go for the feeling of freedom and simpler game play - as long
as we minimised and prevented those annoying moments where you ended up
at dead end road, who needs a map or to be told exactly where to go? It
may not have satisfied everyone's needs but it suited our style of gameplay
and was well received by players.
4. Good marketing relations. It's so important to make sure that
any product you work on is portrayed to the public in the correct way.
Our publishers marketing department already had some knowledge of TIJ
and most of them have watched the film, they just needed to be convinced
about the game in development. We supplied them with hundreds of in-game
screen shots, level data, art resources and early playable missions to
project our vision of the finalised product. This proved to be a lengthy
but worthy exercise as the more information and data that was sent to
them during development, the more enthusiastic they became about the game
and the better they did of marketing. This also worked the other way,
as more decent press was released, the more it boosted the teams morale
when the going got tough and added more pressure to produce a great game
and live up to the expectations of previews and general press.
A marketing budget was put aside by our publishers which allowed for posters,
t-shirts, key fobs, bags, and PR involving sweet talking journalists,
etc which created enough hype that TIJ soon became one of the most
anticipated games on the PSOne. Even young kids who knew nothing about
the film were anticipating the game release, the only thing that could
let it down now was a delay in development or a badly executed game on
release.
One major market we had to satisfy was the mini owners community, this
involved good research and even getting the thumbs up from the Mini Owners
Club on final vehicle models and handling characteristics. One person
even sent an email after our first screen shots were released stating
that they were pleased to see that we had the correct number of spot lamps
on the front of the minis.
5. Team
dynamics. One of the main obstacles while developing any game is how
the team works with each other, you can't allow any one individual get
disheartened as it will begin to spread throughout the team and soon enough,
morale will drop. There were a few lows during TIJ involving reiterating
work and throwing a lot of it away, this is part of the development process
and you can begin to lose sight of the main goal although ideally this
should be avoided. Since the core members had a few years under their
belts working together, we already knew the most basic thing we could
do to prevent low morale was to back each other up, not to slag other
peoples work off, find ways of making good out of the bad and generally
use good communication skills to solve any problems. Unfortunately, not
everyone has the ability to communicate sensibly but you can always fall
back on keeping it professional while working. Importantly, and it is
encouraged, especially with the newest members is to join in the out of
work socials down the pub; here we can talk about things and discuss problems,
a sort of group therapy session, some of our best design ideas come from
a pub gathering. It became more and more important that everyone worked
efficiently together or meeting milestones would require late nights,
however we did hit all our deadlines except for a small accident in the
later stages, which is explained later.
Our work ethics at Pixelogic is to get everyone involved where possible,
all ideas are important, especially at the initial stages and as the project
developed, people began to stay more within their own areas of expertise,
it was good as we all respected each other opinions and working with talented
hardworking individuals pushes you to do your best and more, this also
has an impact on quality control.
Forging good relations with the publisher and having a project leader
who backs you up when your work might not live up to specifications of
the milestone agreement, which could be due to many different reasons
such as unfinished technologies or other requirements were more important
during that month, etc. The main objective was to avoid doing work for
something which wasn't ready to be implemented, this is partly to blame
on scheduling but you cannot foresee everything and some things in milestones
will need reshuffling during the project, for example, do we need all
the vehicles modelled before car handling is finished?
A breather space in the middle of the project was scheduled in, where
we all sat back and analysed what we had achieved and what was the way
forward, this proved to be extremely valuable as car handling and physics
came across as the weakest link to gaining good game play and therefore
an extra couple of months of programming time was arranged to correct
this. The outcome of the new handling technology improved the game no
end, we now could have highly customised handling characteristics, and
all 14 vehicles that feature in the game realistically behaved differently
to each other, which helped to mould the whole game.
|
|
|
|
|
What
Went Right
