Fall Quarter – Weeks Six and Seven
October 19, 2012 § Leave a comment
The last couple of weeks Farah and I have been busy getting ready for our lightning talk at ChiWIC next week. Last week (October 12) Farah and I met with Dr. Wolfe and we discussed our lightning talk and came up with a rough outline for our slides and script. After meeting with Dr. Wolfe, Farah and I worked together on creating the first draft of the slides. Over the weekend the slides were edited by Dr. Wolfe, Farah, and me. Farah also worked on writing the script to go along with the presentation, while I worked on using PowerPoint’s trigger function to create a slide that reveals the number of polygons in each area of the model once the audience guesses the part.
On Tuesday, Farah and I gave our presentation to Dr.Wolfe’s class. Overall the presentation went well. The presentation was within our time limit and the class seemed interested.
At our meeting today we made a few last adjustments to the slides. Then we returned to our work on reducing Paula’s polygon count. So far we’ve made an attempt at reducing polygon count using the ProOptimizer modifier that is native to 3DS Max. While ProOptimizer can reduce the polygon count to about a quarter of the initial face, it also produces harsh edges along the profile of the object. We discussed options for remodeling and believe that we can successfully remodel the teeth and eyelashes from planes using an opacity map. Instead of having each eye lash and to0th modeled we will create a curved plane and apply an image of the teeth (or lashes) to the plane. Then we will control the opacity of the plane with a black and white image of the teeth. The white sections will be the teeth and will be opaque and the black sections will be the negative space and will be transparent. The lashes and the teeth are the two objects with the greatest number of polygons. The lashes have 17, 356 polygons and the teeth have 8,926. By remodeling we should be able to get both of these to less than a thousand. This will reduce the total number of polygons by almost 40% and result in significant time savings during rendering.
To determine the extent of the time savings in wall clock time we have a script that saves the time rendering starts and finishes and then calculates the elapsed time. Using this script we will be able to compare the time different versions of the model take to render and judge our progress towards realtime playback.