Grey - Game Project

We made some timeplan changes since we felt that the original goals for MS2 were misplaced. The original idea was to deliver both our enemies and the character and focus our third milestone period on the level. Once the actual work began we found that it was much more convenient to focus on modeling and texturing props and level details rather than modeling and animating the enemies. The changes are available in the timeplan document to the right.

Here’s a little progress.

Darko updated the texture on his lockers.

Modeling by Fredrik, texturing by Darko.

Modeling and texturing by David.

We’ve textured and lit the first room in the game. Some more props will be thrown in of course.

Texturing by Darko, lighting by David.

I also made some updates to the shotgun texture to make it more detailed.

Fredrik:

There has been a lot of work lately, but not very much posting. On Wednesday I made a grenade launcher, turned out well enough. As it turns out the university has a couple of zbrush licenses lying around which I made use of. I had never used zbrush before though, so I took a couple of hours to learn it, was surprisingly easy to learn, even if the UI is a bit clumsy. So then I spent two days adding details to the to the torso, arms and trousers mainly. After finishing up in zbrush the character is finally finished and was pieced together in Maya.

On Friday we also did motion capture for the game. We probably wont use the MoCap data straight of though. We will use it as reference when doing key frame animation.

I spent yesterday (Sunday) modeling a low poly cage for the normal map generation, did not finish it tough. I expect to finish it this week. But tomorrow I leave for my Christmas holiday. So I probably wont work very much until I get back, but i back on the 28e which is fairly early. So Merry Xmas or die trying.

Completed the character today, which feels great. However, this does not mean it’s actually done. It just means all the parts are there. Left is adding folds in the cloth and building a low poly mesh. I made a more detailed schedule for the next couple of days. Just what needs to be done before i leave Skellefteå for Xmas.

I also scanned some pictures of the human skeleton from an anatomy book and used as image planes. I was amazed and relieved at how well the matched up, in fact, I was unable to find anything actually wrong with any of the proportions of the character.

I was also able to make the sniper rifle today. Tobias sent me a sketch this morning but it was not what we had in mind, so I made a new idea by merging some of the ones we’ve had before.

I (Fredrik) Started modeling the final part of the character today by modeling the face. I made a previous attempt a few days ago, but did not get very far before realising that the reference pictures were not good enough and the result was miserable. So i asked Thomas who is also making a character if i could use his, and they turned out to be great. So today the result was a lot better and I’m proud of it for being just one days work. I did however fell i needed a fresh pair of eyes on it and so sent it onward to David to work on a little. Perhaps the face can be finished tomorrow.

It’s been a while since my last post, so I figured it was time.

First off I’d like to elaborate a little on the viewer part in Fredriks previous post. Getting the scale right is pretty much essential in a third person game. Simply designing your world after real world physical measurements is a one way ticket to some nasty remodeling and texturing. Thanks to a great article on Gamasutra we got a better view of things you have to keep in mind while designing your level. The programmers made the viewer so that we can fly around on our previsualization level and scale things just right.

Here’s a picture from the viewer:

I know, not a pretty sight, but despite its nudist excel spreadsheet appearance this is actually very exciting progress.

Today I textured the shotgun Fredrik modelled. Working with normal maps is a new experience, but I’m beginning to grasp the concepts and techniques and it’s starting to pay off.

Today we were supposed to record all the motion capture data for the main characters. Unfortunately we had to postpone this ones more. We were supposed to record it last Friday but were unable to do so. Now Arash says we can definitely do it next Friday, lets hope so.We made a lot of progress in other areas however. The programmers have now resolved several issues with our camera, and its no longer lagging around. So now we can walk around the level and start scaling all the rooms etc. Form there we can start modeling and textureing the level itself.

Darko modeled and textured an oven. We also modeled one of our weapons today, first out was the shotgun.

here’s some new items i’ve (darko) textured. Fredrik has modelled the silo, pillar and locker, I did the modeling on the wrench and hammer, the oven and the messhall table

Contrary to popular beliefs us programmers don’t just sit and write code all day. To write clean and efficient code you have to plan ahead and design the code on paper.

This is what I have been up to for the last couple of weeks. I am completely done with classes that will handle loading and applying of game configuration settings such as audio and control options. Right now I am working on designing the enemy and AI classes.

It’s not half as bad as it sounds, it doesn’t involve writing any code at all. Instead we use graphical tools, or pen and paper to design class diagrams that help visualize how all the classes are connected and what their interfaces look like.

Here is what I’ve got so far:

Config system

NPC & AI

Over and out!

Just an update on the character model. Its going great.

It has been a while since we (the programmers) busted out and blogged some.. And to be exact, our last entry was posted: 23 days, 12 hours and.. a lot of minutes.

While the CG-artists have been posting their pretty pictures and work results, we’ve been digging our own graves programming-style. The Prototype Phase is not a pleasing experience, with almost no structure, bad naming-conventions and a constant puddle-jumping from one sample to the next, I’m surprised that not more programmers go insane during this time… or bald..

Since our (programmers) Milestone 1 we have been testing samples and tutorials for the Yake API restlessly like little horny rabbits (without the sex). Some of the things we have been testing are Animations, Control Inputs, Scripting and Particles to name a few. There is still some problems to be solved and improvements to be made, mostly for the Physics pipeline but it’s looking good schedule wise.

This week we decided to move our Implementation Phase a few weeks earlier in the timeplan (right after new year we’ll be rocking), we changed this because we don’t believe there will be a whole lot more to gain from more prototypes so we might as well start a few weeks early.

Here’s some code snippets and stuff about Yake.
Yake is really big on namespaces, templates and typedef’s, neither of us had worked much with these perversely ritualistic parts C++ but it didn’t take long to get hooked on the idea and clever usage. Yake’s design is very generic, with a lot of flexibility through the use of generic- and meta-programming techniques.

Code such as this is a common thing for examples in Yake:)

struct TheConfiguration : public raf::ExampleConfiguration
class TheApp : public raf::ExampleApplication<TheConfiguration>

typedef uint32 IndexType;
typedef std::vector<IndexType> IndexVector;
	
typedef boost::function <void (void)> fState;
typedef fState* fStatePtr;

We’ll keep the updates coming whenever we have stuff to write about and time to write it, so hang in there.. I’m sure the CG-artists will post a whole array of new shiny images any minute now:D

Cheers!

We’ve done the presentation for milestone 1 now, you’re free to have a look :)

- Presentation