richardnovak77 opened this issue on Apr 25, 2005 ยท 14 posts
Dale B posted Mon, 25 April 2005 at 6:44 AM
(Steps up to the plate) Hows about me? A good description of Infinite is that it is a combination of all the goodness of Vue4Pro, Vue 5 ProStudio, and a few goodies all its own. Mover 5 -is- built into Infinite, as it all the other plugins. I've imported animations from Poser 4, 5, and 6 into Infinite (with the caveat that in P6's case I decompressed all the files, and decided not to use the external morph targets options, so that all the content is cross compatible between the 3 currently used versions), and animations built in previous Poser builds scrubbed through both P5 and P6 (and from some quicky tests, if you plan to animate, you need to check out P6; both cloth and hair sim times have been reduced, and seem more stable). Both dynamic cloth and hair import properly; you will need to do some tweaking on the hair material to get it to look its best, as Vue's shader system is just a bit different that Poser's. Infinite has the best behaved RenderCow system so far, and the HyperVue status window for network rendering has a lot more information (average frame render time for each active node, flags for what render options you've selected, Gbuffer settings enabled, etc). The -only- truly outstanding issue I've run across is the same issue that has existed since Mover 4 came out; the BVH rotation issue. Some capture software uses a form of interpolation that, at the point where you pass from -360 to +0 degrees, takes the XYZ rotation values into extreme ranges (Xrot might go from 23 to -98, Yrot from -358 to 27, and Zrot from 2 to -84. Just as an example). =Mathematically=, it is still a correct translation value for that part of the rotation, and Poser handles it just fine. Mover, however, seems to read those values in a sequential order, not simultaneously, and the result is that during a rotation (a dancer pirouetting lets say), as the object gets to the crossover point, it will flicker about to the extreme rotation angles, one after the other, from the root node of the figure (the hip). I haven't found a genuine work around for this one yet, no response on the thread I started at E-on, either. Aside from that it works fine on my system.