The Khronos Group wants to do for 3D what formats like MP3 and JPEG did for audio and video.
At Siggraph in Anaheim this week a small army of researchers, artists, programmers, and vendors are showing off the newest ways to fill an empty screen with technically complex and visually gorgeous or occasionally gory 3D imagery. But when all the work is done to create 3D graphics, at the end of the day it gets smashed down and trapped inside a 2D video or photo format, unless it is packaged inside a game. But even then the 3D nature of that data is trapped inside just that one game.
Audio has MP3; thanks to Napster it is a global standard. Photos have JPEG, allowing products like Facebook, Flickr, and Photoshop to make creating, sharing, and editing images a global activity. The video standard H.264 is not a household term, but YouTube made sure it is the world’s leading way to share videos. 3D needs its Napster moment.
Imagine the ability to view a video of 3D content on the web the way you watch a video on YouTube today, but at any time you can press pause and rotate a character, look behind a wall, or view from a different angle. Today’s games offer this, but only because every possible XYZ location has been rendered and stored, waiting to be discovered by the player. There are bits and pieces of this capability available, such as in Google Maps and Google Earth, and in the various viewing and editing products used by engineers, architects, and product designers. But these examples prove the need for a general standard. The data in these products is not sharable or editable outside a small cadre of products, and there are competing standards. Interoperability is not a problem using MP3 or JPEG, but it is a huge problem for engineering and geographic 3D.
If the average user thinks about any of these standards, it is generally in terms of the file type used to store the content. But all three began as research projects on how to transmit the audio/image/video data across the Internet and display or play it in a browser. So too, it will be programmers deep inside today’s leading hardware and software vendors, along with a few academics to keep things honest, who will eventually bring the world a useful way to first transmit, then also display, store, edit, and share 3D content.
Right now the best hope for 3D content is with The Khronos Group, a non-profit industry consortium that has produced WebGL, OpenGL, and other standards that have become the graphics backbone of the Internet. At a pace that may seem ploddingly slow to outsiders, The Khronos Group has helped pave the way for graphics to become universal. Their biggest success may not be in the creation of any one particular standard, but the recent acknowledgement by Microsoft that its next update of Internet Explorer will fully support WebGL after years of holding out in favor of its DirectX and Direct3D standards.
Neil Trevett, an Nvidia executive, is the current president of The Khronos Group. He told us last week a spec draft for a new format called glTF is available for review. The basic idea behind glTF, says Trevett, is to create a “shovel ready binary format” for streaming 3D content that matches the capabilities of WebGL. The first hurdle to cross is the ability to manipulate nodes and adjacencies; only then can they move on to enable streaming and compression.
The Khronos Group will be showing the initial draft at Siggraph this week. Trevett modestly asserts the draft “is not rocket science” but is important nonetheless; “glTF will be the launch pad for all the compression and streaming research and collaboration” that must happen for 3D to have its Napster moment.