This week I’ve been working a bit more on LODs and on a “photo booth”. The procedural faces I made just a little while ago aside from being part of the characters walking around on the station also needs to be in picture form on ID-Cards and such. To get that I set up a photo-booth of sorts for the characters to have their pictures taken in.
This week continued writing the plotpoints (in my storygen format that sees sentences autogenerated). I showed the work off to the team, and everyone seems excited. Adam called me a “mad scientist!”
Actually, Dave, I said your spreadsheets for the system read like the scrawlings of a madman, but they’re also pretty cool.
This week, I needed a change of pace so I focused on LOD integration. I started by creating a tool that could list our assets and show which ones had up-to-date LODs generated. This will allow Daniel to keep track of them to ensure we’re covered across the board.
In order to use those assets, we needed to incorporate a new step in our export process. Basically, for every section you can place in a module, Daniel has carefully laid out its contents in a test scene. The export tool can take those contents and package them in a way that allows us to load the sections as needed with minimal performance impact.
So, given the data that I generated for the tool mentioned above, I can go through the contents of a section and find all objects that have LODs and swap one for the other. The hope is to see a significant performance gain with little difference in graphical quality. Magic!
This week I started working on the concept for the Early Access trailer. Whoa, wait, hold your horses. This doesn’t mean that you should be obsessively refreshing the Steam site just yet (as I may or may not be doing every 2,5 nanoseconds for a certain Kickstarter involving big stompy robots right now). Putting together a trailer video takes a lot of planning to get right, particularly if you want to avoid having to waste a small team’s development time on work that won’t actually be used.
We started off by researching what trends are currently at play in best game trailers over the past few years, spending most of our time in our general genre. Let me tell you that we looked in our couch cushions to see if we could find a couple million spare dollars lying about because there is some gorgeous work out there. We also checked out trailers from old school sci-fi movies to find some ideas.
I then put together a quick storyboard laying out some shot ideas and a rough script which I turned into an animatic using a mishmash of in-game footage, game art assets, very basic geometry in Maya, and quick sketches hashed out in SketchBookPro.
These were brought into Final Cut and assembled with some simple sound and music overlaid to get a feeling for the pacing and the types of shots we will need.
The key point here is that it isn’t pretty. It’s a tool for discussion within the team. It’s a first iteration that we’ll be refining over time. Once we’re happy with it, we’ll be able to focus effectively on setting up the shots in-engine to capture what we need in a short time.
This is important, as the process may reveal that we’re lacking some art assets we’d need to fully convey what we want, or maybe there’s some other technical challenge we haven’t considered yet that will need to be tackled to achieve what we’re looking for. It’s best to have those squared away before we tick down to the last minute and go into panic mode.
The flip side of this is that we may realize that the initial vision for the trailer is too ambitious, and that we need to change our vision in order to keep the scope manageable within a reasonable timeframe and budget (e.g. how much more we’re able to eat starcat food sandwiches before developing a silky fur coat).