| Tale of Tales “8” VAF Mile Stone Plan |
REPORT
|
Step 3: "Education" (Mar 1 - Apr 30) END GAME TRAILER & START GAMEPLAY DEMO |
| Introduction |
|
Despite of the rescheduling of a few things regarding the mile stone plan, it still makes sense to write a report now.
As foreseen in the previous step, we have not made an Autonomous Behaviour Demo in favour of concentrating on the main Gameplay Demo of which autonomous behaviour is simply one part. And as feared, we have failed to finish all the assets required for the Trailer on time. But we have made enough progress, both in terms of creation as in terms of organization, to justify this report.
After the hectic and nerve wrecking previous months, we have found a much more stable and peaceful working mode during this mile stone. We have removed the proverbial rotten apples from the basket. We made a much better schedule for the Gameplay Demo than the one we made for the Trailer. And we have taken more control over final decisions into our own hands rather than relying on expertise of third parties. While the month of March has been spent largely on creating the trailer for the game (of which we submited a work-in-progress verison together with our request for animation funding to the V.A.F.), April was the first month of three in the schedule for the Gameplay Demo. In April we have
|
| Team & Budget |
|
We have been quite disappointed in finding good people to work with. Apparently the creation of game and of 3D assets in particular is a black art that not many people master. We were sad to discover that our own competencies in these fields, despite of our being rookies, are often higher than the ones of the people we work with. But the dream of being able to make the whole game on our own is quickly shattered by the realisation of how much time this would take. And we have discovered quite a few people who are very talented and much better in their specific field than we are. In this mile stone, we have worked with such people exclusively.
Because of the enormous impact on time of communication between different team members, we have scaled down the size of the team and have decided to rely on well organized studios for the creation of most assets. Currently, the team consists of
|
| Schedule |
|
The schedule for the trailer was a failure. We had scheduled everything so tightly that we had to reschedule things too many times.
As a result we ended up with more things at the end of the schedule than we could finish properly.
We know it is not really possible to make a static schedule for this type of project. It needs to be flexible at all times. But if flexibility means pushing things back then your project will not get done on time. The schedule for the Gameplay Demo introduces two new elements: "buffer time" and "done or dead". For all assets that have to be finished by one person before another can start working, there is a week of buffer time. This allows for someone to be a little bit late, without serious consequences for the schedule. And if things are really completely over time, we just exclude them from the project, especially if they are things from first elements of the chain concept-model-texture-animation-programming. The trick to make this work is to divide up the work properly. And I'm afraid we have taken a few big risks in that respect in the current schedule (to reduce time lost on communication, we have scheduled each room as a single big asset that cannot be skipped without breaking the demo). Hopefully the dedication and the professionalism of our team will make up for this flaw. |
| Design "standards" and style guides |
|
The more we work on this game, the more we realize how big it is.
There are approximately 250 rooms in the palace in the game world. Considering that has taken us something between 2 and 4 months to make a single room up till now, it would take over 60 man-years to make the complete palace, if we were to continue like this. This is unacceptable.
The solution is to not design each room seperately but to develop "design standards" by which designing a room becomes a matter of applying certain rules and using mostly prefabricated elements. It shouldn't take more than a few days to design and finish an average room in the game this way. Of course there are a few special and unique rooms in the palace that will require the deluxe treatment. They should be clearly identified as key areas and alloted the time. Another motivation for standardizing is aesthetics. When we started working with trained architect and 3d animator Martin Maximilian Michl from Austria, he made us aware of the descrepancies between the descriptions in our design documents and the actual assets that we were producing. It's not so much that we had forgotten or abandoned these concepts. We just hadn't had time to think about how to execute them. The first thing Auriea and Martin did was to analyze the colours in the paintings that are to serve as the main inspiration for the look of the game: 19th century orientalism. They came up with some surprising results in the form of colour palettes that are going to be very useful when creating textures and light effects. The next thing is an experiment with "vagueness", at first sight a term that is irreconcilable with real-time 3D. We are going to try and create simplified "vague" shapes to represent elements of which the details are not important to the game. If we succeed, modelling assets will become a lot less time-consuming. We have also started writing a style guide with practical tips and techniques on how to achieve the style that we are going for. The aim is now to use the gameplay demo as a starting point for the creation of a library of elements and a set of guidelines for our collaborators to use when building new rooms. 
 Concept drawing of the backroom of the Tailor's store by Dragonfly and of the Tailor's workshop by Martin Michl |
| Gameplay Demo |
We have published
a description of the Gameplay Demo that we are working on.
The idea of this demo is that it is a fully playable little game that includes all of the main features of the full game:
|
| Artificial Intelligence & extra systems |
|
Ronald Jones has redesigned the complete game engine to make it more performative and extensible.
This has been accompanied by
descriptions of the way systems should work.
He also designed a sophisticated system for the Autonomous Behaviour (or Artificial Intelligence) of the main character of the Deaf Mute Girl in her Pretty White Dress. The system is based on the creation and analysis of the "mood" of the Girl. She will be able to act out part of the game autonomously in inpredictable ways. And she will respond to the player's requests to perform certain actions based on how she feels about these requests (and about the player). The player in turn can cheer her up or make her sad by doing certain things in the game. This will make her seem "alive" and will hopefully lead to amusing emotional interaction. This mood system has been developed "on paper" quite thorougly but since it is a complex system and since many other elements of the game need to be made as weell, it is still unsure if the full system will be implemented by the end of June. Some elements in the game would be extremely inconvenient or simply impossible to implement using just Quest3D. For those elements we have hired experienced programmers to do the job in C++. The result of their work takes the shape of extra "channels" that we can use in the Quest3D editor. Ferry Marcelis of Act-3D (the company that makes Quest3D) is developing a system for working with character animations in a very fluent and sophisticated way. This system will interact closely with the Artificial Intelligence and will allow us to build complicated animation blends easily and with much less bugs. As a result, the Girl will appear very lively through a wealth of expressive animations. This system should be finished by the end of May. Alex Mouton has designed a system for the "Wicked Branches", the thorny-forest-gone-bad that threatens to destroy the whole palace in the game. This forest is a massive semi-conscious creature that grows and blooms and tries to attack the Girl when she is near. Alex has just started programming it. It is unsure if this system will be finished enough to be included in the main Gameplay Demo. If not, we will make a small seperate demo with it. |
| Animation |
|
The attempt to make the animations in “8” with high-end 3D application
Maya
has failed miserably.
One of the biggest obstacles was that it takes an engineer to make a proper character skeleton in Maya.
Engineers, however, are not always good animators.
Maya doesn't make it easy to work with motion captures and sharing animations between characters (or between one version of a character and an updated one) is very cumbersome.
And last but not least, exports from Maya in the format that the game engine requires, had problems.
Also, we have discovered that we are quite picky about the style of animations. Apparently many people in the industry lack the talent or the skill to properly express a personality through character animation. When our fragile little girl started moving like a tank, we decided that we had enough. Enough of Maya and of incompetent (male?) animators. We went back to Character Studio and our motion captures of Martha Samyn. Character Studio is not a perfect application by far but it allows the artist to concentrate on animation more than conctruction and exports from it are usable. We found help to fine tune the old animations based on the motion captures and add additional motion with Laura Smith. She possesses the sensitivity and the skill to make our Girl move in the disarming way required for the emotional impact we want the game to have. We will also work with her on animations of other characters in cut scenes. |
| Models and textures |
Although the creation of models and textures will only get into full swing this month, we have continued the development of certain assets that were originally commissioned for the Trailer.
 The Bull staue, the Lamb statue, the Poet Prince, the Princess (without hair...) and the Sultan! All modeled and textured by Meerkat after concept drawings by Lina Kusaite |
| Conclusion |
|
With respect to the three goals of this mile stone plan, we can conclude the following for mile stone 3. In March we have created an enormous amount of assets for scenes that are visible in the Trailer. Several characters and architectural features have been dealt with, not to mention animations. Some of these will need to be further developed or even redone but overall a lot has been added. In April, the redesign of the game engine has been finished, partially "on paper", to be further implemented in May and June. A lot of concept drawings and some models for the Gameplay Demo have been created. And we have been developing style guides, mostly for modelling and texturing. Again, we have not contacted any publishers or investors. We are still focussing on creation and we plan to drop all of that in July to focus soleley on talking with publishers and investors. The Jan Van Eyck Academy is interested in publishing a small book about our project. This book would contain mostly stories about things that happened with the characters in the palace before the Girl ends up in it. It could be a first step towards the creation of the novel or collection of short stories that should accompany the game when it is released. This book may become an important element of the marketing of the game as it may help for “8” to reach an audience beyond the traditional gaming audience. We have learned a lot about team building and management. The current team is reliable and good to work with. But it is ultimately too small to carry full production of the game. Meerkat is a studio that will be expanding over the next year. But we would still need a few people who live in Belgium, who can lead some of the seperate processes. It will not be easy to find these without being able to offer them a concrete work arrangement. In the next two months, we are going to make a beautiful gameplay demo, as close to the design as possible. Hopefully we will find some time to finish more assets for the Trailer as well, as we think it can be an important first introduction to our project. |