May 2nd, 2011
Posted in news | No Comments »
December 13th, 2010
Chris Foster has posted the second part of his Aqsis screencasts showing off realtime depth of field and motion blur in the new Aqsis core. This is very impressive work! Pixar has been experimenting with cloud computing, and prman is now supported by Studio GPU. Also have a look at Mitsuba – a new research global illumination renderer, and at the other end of the scale Rudy Cortes has a blog following of his development of his RenderRan renderer.
Allan McKay raises some interesting points with his article discussing overtime versus productivity. If you are interested in learning more about employment law and how it relates to vfx then check out Vfx Law and Vfx Soldier. If you have recently been made unemployed you might want to check out the Autodesk Assistance Program which offers free, non-commercial software licenses and training. SideFX has also released Go Procedural – a set of new Houdini starter tutorials.
Check out Marvelous Designer – a tool for designing clothing – which has very impressive cloth simulation. In other animation news, Movie Reshape is a paper about tracking and reshaping humans in a movie sequence with some very impressive results. There is a great demo of the Modular Character System rigging system for Maya, and Gianpaolo Fragale’s Surfomatic is a rig to create surfboard motion for improving you animation posing and balance for 3ds Max and Maya.
Jack Binks has several Nuke plugins on his blog for merging HDR images and ASC CDL grading. There are some great Nuke tutorials from Franz Brandstätter and tonnes of Nuke information from Nukepedia. Sam Hodge has a tutorial about creating point clouds in Nuke using 3delight and Cortex. And Cloud Compare lets you compare very large point clouds. MRV is a library currently in development to provide a fast, lightweight Python wrapper to the Maya API. It’s aim is to be more reliable and extensible than the standard Python API.
Disney has a lot of great information on their Digital Cinema website, including how to set up stereoscopic floating windows. Speaking of stereo 3d, PC Pro has an article about why bad 3d gives you headaches. If you’ve been having trouble find her, Kodak’s Marcie can be found here. If you are in need of colour charts information, there is an average Macbeth Color Checker chart and information, You probably don’t want to feature in this gallery of people with the famous chart.
Posted in news | No Comments »
November 8th, 2010
Aqsis has been viewed a one of the slower RenderMan compliant renderers for a while now, but that image is set to change soon. The Aqsis Blog has a very nice video of realtime rendering using Aqsis. This is being rendered on the CPU with no hardware acceleration and looks like it could give Aqsis a great core rendering framework. The guys at Luxology – best known for their development of Modo – have also been looking at the performance benefits of GPU rendering versus CPU rendering. Their video suggests that the GPU is not a silver bullet for magically improving performance. There is more discussion about this at Ompf.
A new photo modelling application called Insight has been released as Open Source for Windows and Linux. And a new online service called Scanner Killer has been released that offers photo modelling for $13 per stereo pair.
Chris Bond – the original founder of Frantic Films – has bought back the software department from Prime Focus. The new company Thinkbox now handles the development of the render farm management software Deadline and the volumetric particle toolkit Krakatoa. Also Alembic is at version 0.9 – and looking for feedback moving towards a version 1.0 release.
Gavin Rothery has written a in depth 3 page blog post that goes into fascinating detail about the vfx on the film “Moon” on which he was the Concept designer and VFX Supervisor.
And finally Alex Segal has been writing about the issues he has been facing in his Anti Mental Ray blog, and Mach Kobayashi at Pixar has a page about his 2010 Stupid RenderMan Trick RSL minesweeper program built in SLIM.
Posted in news | No Comments »
October 17th, 2010
Posted in news | No Comments »
August 8th, 2010
Posted in news | No Comments »
July 29th, 2010
Posted in news | No Comments »
July 20th, 2010
Posted in news | No Comments »
June 23rd, 2010
Posted in news | No Comments »
January 20th, 2010
You wait for ages, then 2 open source vfx releases come along at the same time!
Disney Animation Studios’s 3d paint format and API Ptex is available to download. Ptex avoids the need to UV complex meshes by saves out a set of texels for each face of a subdivision surface. These are independent, so it is possible to have different resolutions on adjacent faces. It does rely on adjacency information to filter across face boundaries, so you need to transfer your texture if mesh topology changes. I think it also only works if you have quadrilateral faces. Hopefully this will be picked up by the popular 3d paint packages soon – the author of 3D Coat has started experimenting with ptex.
And Sony Picture Imagework’s has released the first specification and code for Open Shading Language. This is a set of libraries and a compiler that can be used in any rendering engine, and will be used in SPI’s internal renderer – developed from Arnold. The language has been designed for raytracing from the start, and introduces the concept of radiance closures. These are functions that return a sampling pattern based on the material BRDF to the renderer. The renderer can then bundle coherent rays from several materials together to exploit raytracing caches better. Shaders can be linked in a similar fashion to coshaders, although the language does not define a scene description API (like RIB or mi) so the implementation is up to the renderer developer.
They also have some things which should have been added to RSL a long time ago such as more string tools, runtime determination of varying vs uniform variables, and arbitraryderivatives. They also have an interesting approach to secondary outputs called Light Path Expressions. This uses regular expressions to express which ray types you want in your secondary image. This looks powerful, but also a bit intimidating.
There is a good introduction to OSL by the lead developer, Larry Gritz, and a story by fxguide here.
Posted in news | No Comments »
January 5th, 2010
Posted in news | No Comments »