Research into Game Engines

This is a bit of a late upload, i started writing this entry up on 13^th of January 2013, but didn’t finish all of my research work until later.

Having received no feedback from my learning agreement I have managed to get myself behind in my work. I have gone ahead and looked back at areas to improve on my work from last team.

BA4 re-work:

In my feedback for our game design of “Phenotype” the game was criticised for departing too far from the source material, though I have taken this in mind there is little I can do at this point, and as this has no bearing on Ba5 I will not worry about it. However I will make more of a conscious effort to stick closer to the criteria of the brief, and keep my focus on creating and implementing assets for a game engine.

I was also told that the overly convoluted back story detracted from the bits of the design that worked well. As such, I will steer away from the unnecessary details of science fiction lore.

Overall, the biggest criticism was the lack of research, this is an area that I can and intend to improve on for this project.

We were commended for our well-considered ideas surrounding game-play mechanics. I think this can transfer over to this project if I look carefully at what the brief demands. With regard to carefully considered ideas in the area of asset design and building, I think that I should take care when I am creating concepts, that the designs that I create are practical when it comes to inserting their 3d models into a game engine. I do not want to design assets that will inevitably require high Polly models and complicated rigs. That would not work when put in the game. It’s all about compromise between aesthetically pleasing design, and practicality.

I also need to go back and re-examine the relationship between my design research, the concept art, and the 3d models that we produced to illustrate our ideas.

Game engine research:

In our game design document from last team the team stated that we our game idea was theoretically designed to be developed for the next generation of games consoles from Sony and Microsoft. We felt that this would give our project more scope, and allow us to design a more expansive, varied, and visually pleasing game, as it is theorised that the next generation consoles (as has been the trend in recent years) will have higher processing, and graphics power to support such games.

Given that the game I am working from is 3d, I will obviously need a 3d game engine to import to.  I narrowed my choice down to CryEngine and UDK (Unreal Development Kit) both of which provide a free SDK.

Given that my practical studio work for this project is predominantly visual showcasing of my abilities. Aesthetic and visual fidelity of the engine is an important factor to take into consideration. I have found some samples of side by side comparisons of environments in UDK and CryEngine.

UDK tech trailer:

CryEngine tech trailer:

Obviously these tech trailers are a form of advertising for each respective companies products, and are not a none biased source of information. Nevertheless it is interesting to look at and compare how each company (Crytek and Epic) chooses to present its engine. It seems that Crytek is keen to show its engine as an impressive tool to create beautifully photo-realistic environments, whereas Epic Games seem more intent on showing how robust their engine is in terms of variety of games and applications.

(http://www.psu.com/a006835/CryEngine-3-vs-Unreal-Engine-3-Comparison)

(Posted March 31^st, 2009 at 20:18 EDT)

These images show that both engines are obviously able produce impressively detailed textured environments. Allot of the comparisons and debates that I have found online focus on particle systems technology and character animation. While these are important to consider when creating a fully working game, they are less important to me as I will not be using these features, as such I will focus more on asset rendering and texturing, and won’t let my decision be swayed by factors that aren’t immediately relevant to this project.

Looking at more side by side comparisons (video this time) gives me a better (and also none biased) comparison of how the same asset looks respective to each engine.

It seems to me (by looking at this clip) that allot of the impressive visuals in the cry engine are largely due to dramatic lighting effects that can be used to stage the game word in the most flattering way. Looking at comparative images of actual games produced in each engine continues to support my finding. I also feel that much of the benefit that comes with CryEngine comes with its ability to create and present realistic vegetation (organics) and soft surface modelling. Due to the interior, setting of my game design, and the industrial/clinical mix of design ethic I do not think I would be able to make use of these features either, within the parameters of my own learning agreement.

(A screenshot of “spec ops: the line”)

(Crysis 3 pre-release promotional screenshot)

Though my research has lead me to believe that my assets would have a higher graphical fidelity in CryEngine (producing a more “realistic” aesthetic), further discussion with my tutors about my own level of experience has lead me towards using UDK to import and showcase my assets. Being told (by CryEngine veteran Lothar Zhou) that as a beginner to the world of game engines, I would most likely pick up UDK quicker faster, was a large influence on my decision to choose that engine. I think that learning to use any game engine will be challenging enough without having to battle against the engine itself