Jul 15th, 2015 at 5:00pm PDT, Agenda and Minutes

From Web3D.org
Revision as of 18:22, 15 July 2015 by Cxo eoso (Talk | contribs) (added Leonard's comments via H-Anim mailing list)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

1. LOA2 and LOA3 H-Anim models

  - 10 Schematron errors from the validator in the LOA3 model. Don will review the reports and clarify the error messages. Don could not open either file with Instant Reality. 

2. SIGGRAPH Web3D standards draft meeting, August 9 (Sunday) at 1-5 pm

   - Waiting for reply from Anita who should facilitate with reserving a room for this meeting. Anita will arrange for the room at The Sheraton or a nearby conference hotel. WiFi, 

3. SFRotation interpolation between two SFRotation values

  - ex: if 10 SFRotations between 1 0 0 1.0 and 0 1 0 2.0. Don suggested using an Euler interpolation function. Don asked if MWL wants to compute the interpolation values herself. When you look at the "shortest path unit sphere" code examples you can back out the algorithm. Just look at the source code for the Open Source browsers (e.g., FreeWRL in C++) and search for the  implementation of the orientation interpolation node. Don suggested referencing the books Geometric Tools for Computer Graphics and Graphics Gems. MWL tried using X-axis, Y-axis, Z-axis independently but got poor results and rotation matrices didn't yield good results.

4. The specification v1.0 paragraph 4.9.5 Hierarchy and Table 4.4 Face Joint Object Names) defines eyelids, eyebrows and temporomandibular (jaw) joints but these joints are missing in Table A.9 — LOA Three: Default Site object translations. MWL will submit a Specification Change Form so the discrepancy will be captured in Mantis.

Joe and Don came up with the following Table extracts:


Joe, might you and I get back to a weekly call on topics like this? For example, maybe we should improve the table by adding counts for # joints, # segments etc.

5. H-Anim motion capture (Part 2 Working Draft)

  - Table of contents distributed by MWL. 

Leonard Daly emailed. In the original H-Anim standard, the default pose was an "I" shape (standing straight up, arms at the side). When mo-cap first started to become popular, the standard practice became a "T" pose (standing straight up, arms straight out). This cause some unnatural muscle shapes when the character was at rest, but it was east to rig and dress. Now it appears that the standard practice is a "A" pose (standing straight up, arms at 45 degrees out from the side). This relaxes the arm muscles, but still leaves room under the arms for proper skinning and rigging.

What will be the H-Anim standard pose in the next version of the standard? If it is not the "A" pose, how will the Consortium sell it to modelers? If it is the "A" (or actually any pose), what happens when standard practice changes (again)?

Note that "standard practice" refers to the standard used by people regularly model humanoids. This includes all of the work for film, TV, games, and other entertainment; and those who work in Maya, Blender, or other tools.

6. Scheduling next meeting

 (1st Wednesday) August 5th at 5:00pm PDT (6th, 9:00am KST)