Portfolio Development 5: Building The Artefact

Once I’d finished scripting the movie, I recorded it using FRAPS. At this juncture, I hit a slight snag in that I was unable to run the final map without some framerate issues on my own PC. I then decided that I’d take the opportunity of our lecture on the 27th to record it on the University computers, which I already knew would run it smoothly.

Alas, bureaucracy is the compost of indolence, and thanks to the lazy incompetents who are in charge of that sort of thing, we weren’t informed until that morning about the maintainence scheduled for that room (and we only found out then because I ran into Tim outside the door). Because of this, the raw video I had to use for the final product is somewhat laggy in places.

This aside, editing it in Premier Pro went quite smoothly. I chose to use a flaring transition between monochrome and colour footage to symbolise the dawning of realisation as the protagonist first discovers nature and selected the first movement of Schubert’s 3rd Symphony for its dramatic trumpet swell, which seemed an ideal accompaniment to this moment. As it turned out, I was forced to take a metaphorical meat-cleaver to the music (occasioning no small feeling of guilt), in order to fit it more ideally to the film. With this distasteful task complete, I brushed the mutilated semiquavers from my sleeves, and turned to the black-and-white opening section, which seemed to me a shade too barren.

Fortunately, a solution presented itself: poetry! While I felt that “I Wandered Lonely as a Cloud” was perhaps a shade clichéd, its dual themes of solitude and nature were too ideal for me to pass up, and so I dubbed it over the top of the opening section, where by pure happenstance it fit perfectly.

As a final touch, I decided to start the film with an appropriate quote, an opening bon mot, if you will. A friend suggested Descartes famous “Je pense donc je suis,” but I found it to be less than apt. In the end, I decided to quote Feynmann, ultimately for no better reason than that I was pleased by the almost tangential relationship of the quote to the subject of the movie.

Altogether, this was not only the easiest but also the most pleasant part of the project’s creation, and I find myself quite pleased with the outcome.

Portfolio Development 3&4: Scripting The Movie

The first part of the movie scripting was relatively simple; a matter of getting the camera motion working in preparation for creating all the different scenes. Since I could look up the code from the PerfTest map for a reference, this was easy.

The next part, however, was not. Managing to make a unit move to a specified location using scripting was quite horrible, and I’m honestly uncertain that I could’ve done it in any reasonable timeframe without help. (I’m sure I’d have managed eventually, but having it working in five weeks isn’t much good)

Once that was achieved though, implementing the camera movement and unit direction on a larger scale was trivial, and I was able to have the movie scripted almost in its entirety within two days.

By opening with a sweeping shot of the city, before cutting to the little robot, I’d hoped to create a sense of scale, and of insignificance in a vast world. Furthermore, the style of cutting between scenery and robot was one that I adhered to throughout the movie.

When the robot reached the oasis, a wide angled shot was used, pulling back to show the oasis as it came into view at the bottom of the screen. This ensured that the viewer was surprised by its sudden appearance, and helped induce a sense of wonder.

The next scene, showing the robot exploring the oasis, could have been better. Because there’s no capacity for facial animation, and using any kind of ‘voice’ for the robot would have seemed out of place, I had a hard time showing curiousity. In the end, I simply made the robot wander erratically and quickly around the oasis, in the hopes that that would be sufficient.

The following scene, in which the robot explores the city, is the first of the two primary scenes of the movie, intended to serve as a parellel for the nature exploration scene that comes later.  The seeds for the sense of déjà vu that I hoped to create were planted here through my camera-work, which I would then duplicate in the following scene.

The brief shot of the robot showing the city was, frankly, a disappointment. I’d hoped to be able to have the natural world shown sweeping out from the city’s edge, but due to my need to avoid showing certain things, I was forced to make do with a very close, boring shot, which just showed the very boundary as the robot left the city.

Next came the second of the two major scenes, the robot’s exploration of the natural world. This was essentially done in precisely the same way as the city exploration scene, just in a different location.

Finally, there was the grand reveal, where the robot finds the hulk of another that went before him, and the camera pulls back to show the overall shape of the map, and reveal the yin-yang design of it. There’s really very little to say about this scene: it’s a very simple shot, and it works beautifully. The one thing that almost became a problem was the game’s desire to display the little strategic unit icons when I zoomed out too far, but fortunately I was able to script the map to disable those icons when it l0ads.

That’s about all for the scripting: the map is done. Now it’s on to the video-editing part!

Portfolio Development 1&2: Map Creation

The first step in making my final artefact was creating the map itself. I did this in three major stages, the first of which was simply doing the main texture work to create the Yin-Yang pattern which is a central feature of the design.

This done, I decided to do the city half of the map first, since I anticipated it would be more time consuming than the nature half. I decided to use mainly buildings from the “UEF” faction to construct the city, since those looked most like a contemporary city. By placing different buildings in manners that caused them to merge together, I was able to create a surprising variety of different designs, and make a city that looked quite passable overall, especially from the low views I intend to use in the final product. Furthermore, by raising it up on a plateau, I ensured that it could be used for horizon scenery in later scenes if I so desired.

Once I had finished the city half of the map, I turned my attention to the nature half. This was fairly simple, since placing large numbers of props at once is one of the few things the Supreme Commander editor is well suited to. Due to their tiny scale and low polygon count, the overall effect is less pleasing than I would like, and the small selection of models is also a detriment to the appearance of the map; but it is sufficient not to embarrass me.

The next step in the production is to begin scripting the camerawork and NPC movement. When that is complete, the product should be essentially finished, though I may yet add more polish to the appearance of the map if I am able.

[FATAL EXCEPTION ERROR] – Post Regarding Unreal!

Creating a cinematic in UT2004 sounded like a relatively simple project at first, especially as the Matinee tool was fairly simple to come to grips with. Alas, I couldn’t have been more wrong. I elected to build my project scene by scene; first creating the set for a scene, then scripting the action before I moved onto the next one.

It’s very fortunate indeed that I took this approach, since it made me aware well in advance that the AI scripting and actor properties in UnrealEd are horrifically unreliable. Pawns that completely ignored their rotation values, actions that simply refused to be executed, timers that ran at a different speed to reality, the problems were innumerable. My personal favourite was the point at which a set of pawns I placed, which I’d copy-pasted from a set used earlier in the cinematic, all reset to the same default skin, completely ignoring the one that they were set to use. I could go on listing the things that were faulty, ommited, or obsolete, but I fear my wrists would cramp before I finished.

In any event, had I not taken the approach I did, I would almost certainly not have allowed enough time for the scripting. As it was, I managed to complete it two days before the due date (though I then had to work a couple of fourteen-hour days to do my Level Design project).

Looking back at the project, regarding the input I recieved… The most distinct flaw was the lack of music, or indeed of any real sound effects. I had intended to at least find an appropriate background track, but in the end I simply didn’t have the time to spare. If I’d had longer to spend on this project, I’d have seen to the sound aspects of it, as well as spending more time finding ways to work around the broken parts of the engine.

Of Ancient Air – Building A Museum

The hardest part overall for the matinee tool assignment was creating a suitable set for the camera-work. Because of the potentially disparate nature of the exhibits, a museum was ideal; it allowed me to make use of a large number of models without needing to worry too much about context. My original plan for the level was to have a few small, themed exhibits, as well as multiple individual items scattered about the museum in niches or on pedestals. After putting together a test-room however, I decided that this seemed too chaotic and random, and abandoned it in favour of a few, larger exhibits, each with a dominating theme.

The first one I created after I made the entrance hallway was an exhibit consisting of a diorama-like scene set in one room of an old spaceship, with a corpse lying in the middle. This was glassed off from the main room, which contained a few smaller items, such as a cleaver and meat-hook on a podium. The intention was that this exhibit was to do with some fellow who’d gone crazy on a long voyage and massacred the rest of the ship’s crew, but feedback from test viewers suggested that this was too obtuse. Because of this, and because the exhibit didn’t fit well with the others I had planned at the time, it was eventually cut.

The first exhibit I made that is included in the final map was the Japanese-themed one; because of the large number of statues, urns and bits of architecture in the static-mesh package, it seemed an ideal theme for a museum exhibit.

Leading off from that, I (undoubtedly like everyone else) made an Egyptian-themed exhibit. Once again, the massive quantity of appropriately artefact-like models made it a perfect choice.

I had originally planned to include a mediæval European exhibit, but when I searched more thoroughly, I found that there weren’t as many models as I had first thought, and those that there were tended to be quite bland and didn’t make for very dramatic exhibits.

For the overall structure of the museum, the corridors, entrance hall and so-forth, I chose to use dark stone textures, lit by a dim green light, mostly from floor level. By lighting the walls upwards, and placing buttresses between the light sources, I could create a very striking pattern of light and dark on the walls. Furthermore, illuminating the ceiling but not the floor served to create a skewed perspective effect, making the viewer feel as if they are very tall.

Choosing lighting for the exhibits was somewhat trickier; at first, the Japanese exhibit was lit in a cold blue, which meshed well with the green tones of the museum’s overall lighting. When I came to the Egyptian exhibit though, the natural choice was a golden-yellow lighting, reminiscent of the desert sun. I was so pleased with this effect that I found myself looking askance at my earlier effort, and wondering if there wasn’t a more appropriate choice in lighting. I had originally been intending to light the cut European exhibit in red, but a discussion with a friend reminded me of Japan’s nickname, “Land of the Rising Sun”, and so I reworked that exhibit to be lit in an orangey-red light, which turned out to suit it much better.

Once the museum itself was completed, it was only a matter of doing the camerawork. My previous experience with the cinematic tools in the Warcraft III editor came in useful here; although the tools were somewhat different, the basic theory was the same, and so I was already accustomed to working with panning shots and tracking targets. I decided to make my sequence a simple fly-through, almost like a video tour, and this turned out to be mostly simple aside from one inexplicable problem. About a third of the way through the sequence, when the camera was told to orient itself to face an actor, it rotated in the opposite direction to that which I had intended, turning almost a complete circle clockwise, rather than turning very slightly counterclockwise. I was able to rectify this by fiddling with the yaw settings in the orientation sub-action, but all this did was move the problem along to the next orientation sub-action. By going through the entire cinematic after that point and tweaking the yaw settings I was able to remove the problem, but I still have no idea why it occurred in the first place.

Overall, I am pleased with what I created, though the camera sometimes jerks inexplicably. If I’d had more time, I would have liked to try some more elaborate camera-work, but… c’est la vie.

In The Beginning…

…there was an empty blog.