![]() Los Angeles mental ray® User Group |
![]() | ||||||||||||
| |||||||||||||
Summaries from past years2006 Summaries2005 Summaries2004 Summaries2003 Summaries |
2004 Meeting SummariesJanuary 2004On January 27, at RFX, we discussed the current environment for jobs and training for mental ray in the Los Angeles area. We gained perspective from both the job givers and job seekers. As a followup, we are currently in the process of increasing the level of resources available on this site, as well. February 2004On February 24, in the Tiki Room at Jamaica Bay Inn, we discussed a few parameters of interest to people new to mental ray. Charlotte from Dreamworks talked about how parameters should be set for typical production usage. The slides from this will be turned into a section on the resources page. Then, we focused on hair and geometry shaders. Bart discussed details of the hair primitive and geometry shaders from a cross platform perspective. As an example, transparency was added to basic hair shader in the mib library. In order for such a thing to work, there does need to be some support from the 3D application to put out transparency numbers. Next, for an example of a geometry shader, a feather was created using the hair primitive. We also discussed the use of placeholders, how they could be input from separate files as well as from callback functions, and how this seems to be an untapped area of usefullness for production. Then, getting more app specific, Mark Schoennagel from Softimage showed us hair on XSI, and Bart showed Shave and a Haircut as a plugin for Maya. It can be noted that XSI also has integrated Shave functions. March 2004On Tuesday evening, March 23 at Digital Dimension in Burbank, we covered current topics from the mental ray mailing lists as they apply to local needs. Among other things, we discussed how to manage your component files with the object file placeholder mechanism. April 2004On Tuesday evening, April 27, at the historic Culver Hotel in Culver City, we held a post-NAB discussion. We had lots to talk about, from new release features to hardware acceleration. We talked quite a bit about that, and we will probably continue to do so, as we develop demonstrations of the acceleration abilities. We also discussed distributed rendering solutions for highly interactive lighting and look development turnaround. This was because an Altix was showing this in the SGI booth. This also led to discussions and eventually a page in the Resources section on setting up May 2004May meeting not held. (Too many last minute changes.) June 2004On Tuesday evening, June 22, at RFX, we discussed the latest version of mental ray (3.3), as it had been introduced in new releases of both Maya and XSI. After a generic feature overview, we showed how the new features integrated into both Maya and XSI. Thanks to RFX, Alias, and Softimage for contributing to the meeting. July 2004On Tuesday, July 20, at Digital Dimension in Burbank, our LAmrUG meeting featured a focus on skin and other subsurface scattering (SSS) related materials. First, we provided an overview for artists, covering what you need to know about shaders to use the SSS Phenomena. See the SSS tips Resource page to see the kinds of things we covered. SIGGRAPH 2004
October 2004On Tuesday evening, October 26, at RFX, Alan Harris, the Alias Senior Support Engineer specializing in mental ray for Maya, gave a presentation about tuning and optimization of mental ray parameters, and tips and tricks on everything from network rendering to dynamic/extra attributes. He has given us the presentation for your reference; see link below. Thanks to RFX and Alias for arranging Alan's visit and supporting the meeting. November 2004On Tuesday evening, November 23, at the Marina International Hotel in Marina Del Rey, we met to talk about hardware configuration and acceleration for mental ray. BOXX Technologies sponsored the event with food, drinks and t-shirts, as well as providing a speedy workstation with which to demo rendering. To begin, Ed Caracappa of Boxx Technologies presented the options one should consider when configuring a machine. We talked about processors, memory, storage, and video cards. It was interesting to note that CPUs for 64-bit processing don't affect cost decisions as much as support issues do. Additionally, architectures such as the Opteron have internal features which increase rendering capacity. With large memory, storage and dual GPU systems a current day reality, we now have great flexibility in the approaches to production, and more specifically rendering. Following this, Bart Gawboy presented the flexibility mental ray offers with respect to multiple machines, big memory machines, and GPU-accelerated machines. On one hand more and bigger machines can accelerate render throughput with less overhead than scene optimization and layering. On the other hand, GPU acceleration allows for extensive scene tuning opportunity. The options go both ways. After discussing how to mix software and hardware rendering, we showed some rendering demonstrating the ability to break up a scene and choose what to feed to the hardware for acceleration. In one example, we showed a ray-traced floor rendered in software combined with millions of triangles of objects rendered in seconds on hardware. We also showed some high quality, fast hair rendering using hardware shaders. We had some great discussion, and its clear that we've uncovered just the tip of the iceberg with regard to what is possible. Now its time to bring these possibilities into the realm of practical production methods. AcknowledgementsThank you to mental images for giving their full support to the formation of a locally-based group dedicated to mental ray users. |