Yesterday I presented my fiancé with a proper ring, a very beautiful and magically well manufactured piece that symbolizes the relationship the two of us are building. Of course, only a select few people on this planet (and maybe some x-ray goggled aliens who are interested) can actually say they are witness to this occasion. The rest of you have to take my word for it until you get the opportunity to see it yourselves.
But this is a reasonable claim, and many of us have friends who have done exactly this and so we accept that this situation is possible. It is entirely possible that I am lying.
Lets say that I have created a company, and that I have been in business for a few years. Lets say I claim to have created a new technology. This technology is not specific but I make claims about increasing the detail of games not by multiples, but by 6 magnitudes. More over, I will put these claims forward without evidence, reason, or examples backing them publically. You are only able to accept that these claims are real because of videos that I share every year or so.
This is exactly the set of information that are coming out of the Australian company Euclideon is making about their “new” rendering technologies, and there are plenty of skeptical opinions (and many more) to take into account. One of their claims is quite interesting, and actually lands in the realm of philosophy within Computer Science, and it is the name of their initial product – Unlimited Detail.
Every time a new 3d graphics card or games console (Nintendo, Xbox, Playstation) is built, they make it more powerful in order to run more geometry (Put more objects on the screen and make them look more rounded). Games today still have a long way to go before they look like 3d movies (Shrek, Nemo, Toy story); billions are spent every year on R&D to make hardware more powerful. Unlimited Detail is a software algorithm that gives unlimited geometry. When we say “unlimited geometry” we really do mean it. It really is unlimited, infinite, endless power, for 3D graphics.
Notch, programmer behind Minecraft brought things into scope pretty well, and he received quite an interesting array of responses:
They made a voxel renderer, probably based on sparse voxel octrees. That’s cool and all, but.. To quote the video, the island in the video is one km^2. Let’s assume a modest island height of just eight meters, and we end up with 0.008 km^3. At 64 atoms per cubic millimeter (four per millimeter), that is a total of 512 000 000 000 000 000 atoms. If each voxel is made up of one byte of data, that is a total of 512 petabytes of information, or about 170 000 three-terrabyte harddrives full of information. In reality, you will need way more than just one byte of data per voxel to do colors and lighting, and the island is probably way taller than just eight meters, so that estimate is very optimistic.
And then he elaborates
- One byte per voxel is way lower than the raw data you’d need. In reality, you’d probably want to track at least 24 bits of color and eight bits of normal vector data per voxel. That’s four times as much data. It’s quite possible you’d want to track even more data.
- If the data compresses down to 1%, it would still be 1 700 three-terrabyte hard drives of data at one byte of raw data per voxel.
Unlimited Detail is a bit of a pipe dream. It is always a shame when math and physics gets in the way of amazing dreams and wishes. Watching their own video leaves you wondering what the hell is going on.
We didn’t increase it by 10 times, or 100 times or 1000 times. We increased it so far we could abandon polygons all together and move to little atoms and run them in unlimited quantities.
This is outright bunkery, and more importantly it is unsubstantiated. Note above that we have been given a video. this video, although rendered at 1080p, it is by no means clear. Its pink and blue grey point clouds are innumerable sure, but not because it is impossible to count that high… but because of apathy and a lack of ability to discern what each model looks like. Even when they get down into the dirt, it is clear that there is a problem with duplication of content. Another key element is that their videos are rendered on their side, admittedly below our 24-30 frames per second to avoid page flipping artifacts.
If you have a background in the industry you know the above pictures are impossible. A computer can’t have unlimited power and it can’t process unlimited point cloud data because every time you process a point it must take up some processor time. But I assure you, it’s real and it all works.
Hang the fuck on, pictures? Rendering these pictures is impossible? That’s utter bullshit. I have seen pictures of JFK with an arm around Elvis and Bill Clinton, long-cats storming castles and starships moving faster than the speed of light. Don’t you dare tell me that the rendering of an image is impossible. We are only told and expected to believe they are in-renderer images, but we have absolutely no evidence through with to draw the conclusion that it must be any such thing. Hell, there are screenshots of amazing images from movies like The Incredibles that took a few hours to render and they contain more detail than is shown in their video or screenshots, albeit with shaders and texture manipulation.
The demo images they show off are primarily of duplicated elements. There are similar images available from real-time rendering systems like Ogre3D, which show off the ability to have thousands of meshes duplicated and there are tutorials for OpenGL DisplayLists which do similar things. There are images that have been rendered here as well. If you move away from real time elements you can push the amount of detail up with almost limitless bounds.
No matter my claims put forth, it is pretty clear that the course of action here is simply to remain skeptical and wait until we get some further information from their company, and by information I do mean something more interactive and representative of their claims.