In 1983 Atari inc. released ‘I, Robot’, the first ever publically available video game to make use of 3D polygons. Since then, a lot has happened in terms of graphics and although countless sprite and text based titles have come between now and then, the fact remains that the vast majority of titles released today still use the same polygon concept developed and put to use almost 30 years ago. Of course, it goes without saying that despite the same technology and concept, Metro 2033 does offer a little more than its ancestor, ‘I, Robot’, but regardless of this the same flaws remain: low poly-count, poor poly-shading and low-res textures. The obvious solution to these problems would be to increase visible polygons and decorate them with higher resolution images whilst creating better, more dynamic lighting. This obviously comes at a price however, and is limited by the processing power of the system and the amount of RAM it holds, which is usually, in the case of current gen consoles, seriously lacking.
So what about PC gaming, in which you can affordably get your hands on ridiculous amounts of RAM and processing power? Unfortunately, as sound as the above question may seem, there still remains an issue, or two in fact. The first of which is most likely the most obvious and that is that even with double, triple or ten times the processing power of consoles, circles and curves can never be circles or curves due to the very nature of polygons and although achieving near perfect results is possible, the strain is hardly worth the results. The second, and probably the more substantial of the two reasons, is that alienating console gamers and those who play on lower spec rigs is quite simply not financially viable, especially when you consider that producing those better graphics is more difficult and costly for the developers and artists to boot.
So where can we go from here? Waiting would seem to be sensible since it goes without saying that in time graphics will improve in much the same way they have done so far and at a similar rate. Consoles as well as PCs will be able to get more and more kick at feasible prices, meaning that at some point in the infinite future polygons could theoretically reach a point that they were so small that they were hardly noticeable. That, however, could take quite some time considering how little graphical development we’ve seen in regards to polygons these past few years, and most gamers are quite impatient by nature. It should come as no surprise to you then, that developers put quite a lot of money into workarounds to the issue and have come up with a whole myriad of solutions including Mass Effect 3s famous usage of sprites and Call of Dutys well-renowned repeating models. The results were not ideal and it was actually The Sims, of all things, that took a rather substantial step into the right direction; only render what you see.
It seems obvious really and most games do infact do this to some extent in regards to draw distance, but why bother rendering the back of that dumpster when you can’t even see it? This is the philosophy that The Sims took in order to make it’s game as playable as possible for those not-gamer types with their non-gaming computers and to be fair to them, it worked pretty well. The issue came when they remembered that someone has to decided when somethings in view and when somethings not, which of course, the computers job and much like every job, this takes processing power. One step forward, one steps back. The difference in processing power was noticeable, but not by much and it’s not like to concept could be properly used in free camered games, it would simply take too much juice. The concept however was spot on and is almost definitely the future of graphics, just not with polygons.
Voxels. A term most of us would have heard about by now, but likely not one you’ve explored to the fullest. With hundreds or even thousands needed to build a simple plane, it does seem queer to consider these an alternative to polygons, especially when you consider that the computer now has to process millions of objects as opposed to just one, but things aren’t quite as simple as that. Due largely to the nature of RAM and a whole bunch of other complicated stuff, it’s actually easier (much easier) for the computer to render and display 100 of the same image than it is for it to simply show one image 100 times bigger. The benefits this nature has on textures is obvious; a repeating pattern of sand composing of 100 voxels repeated across a dessert will look much crisper than the same image stretched across the entirety and will also be much kinder on your system. This alone isn’t the main benefit of voxels however.
Generally speaking a single voxel (regardless of size) will take a specific amount of processing juice regardless of it’s position or colour. What this means is that provided the rig can handle the amount of voxels present, perfect circles or literally textured surfaces can be as easy to create as a straight plane, should you be using the same size voxel in both cases. As well as this, each voxel acts independently of those around it and is perfectly capable of personal mechanics and physics, allowing for realistic destruction similar to that of physX or even perfectly emulated water, sand and smoke. In a lot of cases, minute details, such as the pin on a tac or a place of glass, would actually take less voxels to make the planes since by the very nature of 2D planes, you need at least three to make anything 3 dimensional. Voxels obviously don’t have that quality making them ideal for things without much depth.
If they’re so great, why isn’t everyone using them? Quite simply because computers cannot handle close to the number of voxels needed to make, well, just about anything of realistic detail and yet despite this the limitations are far less than that of polygons. Working in a similar way to real life, rather than using tricks and workarounds the lack of limitations is pretty much endless at a strain that isn’t exactly impossible for the rigs of tomorrow to handle. With plenty of viable solutions to limiting the amount shown at anytime, such as what we’ve seen already on the sims, or similar to how minecraft functions, it’s hardly as if the concept in practise is unimaginable. It’s regrettable to say but the day we see shells freely rolling across the floor, sand shift beneath our feet and walls physically eroding from gunshots is a very long way off and won’t be something we’ll be seeing on the PS3s half a gig of RAM. Nonetheless, it’s possible, feasible and most likely inevitable; voxels, ‘particles’ and free functioning physics is the future of videogame graphics and will revolutionise everything we know about gaming in much the same way the poly-graphics did almost 30 years ago.
452 total views, 2 views today
Articles related to this: