Pixar released a
press item on Monday May 9th, 2011 that Pro Server 16.0 is now available. What makes this version so important is because of the competition Renderman has faced in recent years, primarily the Arnold renderer and the fact that ray tracing in films is actually quite affordable now, in fact in some cases needed in order to really achieve the photo real look. Ray tracing in film is not new, in fact Pixar and ILM were the first houses that were able to do so, however these shots were few and far between. One of the first uses of ray tracing by Pixar was during a couple of shots in "A Bug's Life", then full use of ray tracing and global illumination in "Cars".
The problem is that while it is possible, ray tracing and global illumination is horribly expensive, long render times, huge load on the CPU and memory, requires baking the illumination similar to traditional shadow maps and in general slow to work with. Pixar's flagship product and grandfather of the REYES family tree has been considered slow compared to other renderer's like Arnold, VRay and MentalRay when it comes to ray tracing. This coincidently helped these competing products establishing a foothold on the VFX and animation industry as viable production renderers because of that very fact.
From the press release :
"This latest release features a large number of innovative advancements in RenderMan's ray tracing technology, including a new ray tracing hider, a radiosity cache, and physically plausible shading. These milestones allow RenderMan to take full advantage of the ever-increasing processing power of multi-core architectures, while also delivering the tools to implement these new features with efficiency and elegance. Moreover, RenderMan's new progressive ray tracing provides interactive re-rendering for production shading and lighting."
While nobody can prove otherwise, it remains a bit suspicious that Pixar added a new interactive re-rendering method, possibly similar to the Aqsis interactive re-rendering that was announced last year, possibly it is just great timing. It is still cool none the less. Aqsis still holds the title of being the first REYES renderer that can re-render at near real time on the CPU.
"RenderMan Pro Server 16.0 fulfills every need on our feature request list. It lets us focus on the customizations we really care about for ray tracing physically-correct lighting in the cleanest way possible," said Dan Evans, Head of Shaders at Framestore, London. "The speed of the new radiosity cache makes ray-traced global illumination practical in a single render pass, and we can now refine our test frames live using the new progressive ray tracing. Multi-bounce glossy specular, importance sampling, area shadows, and direct lighting are now a breeze thanks to the renderer supporting them all directly. Better still, regular ray tracing is staggeringly faster at 8X on some of our more complex stereo renders.”
Of course the products are commercial with ProServer 16.0 at $2,000 a seat, so this is beyond the price range of most of the population. A full price list can be found
here if you want to check them out. This is one reason for this website's existence, much of the population that would be interested in Renderman simply cannot afford the software that these houses use, at best we get to use free limited use versions, so the importance of open source Renderman is very valid. While Aqsis still is not up to the level of features that PRMan does, it is far beyond just a hobbyist tool and as seen over the past couple of years, capable of producing film quality imagery.
Pixar always seems to kick it up a notch when it matters.
Weta also made recent news when it announced that it will be
adding deep image compositing to OpenEXR, while at FMX 2011. More can be explained
here at Colin Doncaster's website. Colin was the original developer of the Liquid plugin, a Maya to Renderman exporter that first saw use at Weta
(see a pattern??). The Lord of The Ring's trilogy's visual effects were primarily accomplished via this tool for instance.
Animal Logic first produced a paper that described what deep image compositing is and what it is for, it was supposed to be at SIGGRAPH 2010 but was rejected. Why, nobody knows, SIGGRAPH paper judges are not known for their generosity. Since then several applications have had this added, such as Side FX's Houdini. Now OpenEXR will get another very useful addition that will allow other applications to take advantage of deep image compositing, so one could expect to see this in Blender as well as Aqsis in time, when this happens is uncertain and certainly not planned officially anywhere.
"The concept of a deep image isn’t brand new; ultimately it’s just the technique of encoding more than just the RGBA value in a pixel. Many applications and systems already store multiple channels of data to enhance the compositing workflow as well as re-using calculations already performed by the rendering engine. Side FX software’s Houdini is an example of one of the more recent applications to utilize this workflow via it’s custom camera image format.", from Colin's site.
Exciting times in all areas of visual effects and animation!
