FPrime V3推出

[moc]

http://www.worley.com/E/info.html
http://www.worley.com/Media/animations/ ... _baker.mov
http://www.worley.com/Media/animations/ ... ectric.mov

FPrime's design goal is simple, to boost LightWave workflow by continuously showing your rendered scene while you create and edit it. This transforms scene setup, lighting, and surfacing into an interactive process, not a cycle of repeated slow test renders.

LightWave 9 has advanced the fundamental technologies of Layout with powerful new Node and camera systems. FPrime's original preview emulated LightWave 8's "classic" surfaces, but the new Node system has a different level (and style) of complexity.

The solution was to replace FPrime's internal surface system with a much more versatile and extensive core. Instead of simple surface textures and classic shading, FPrime's engine now allows even portions of its engine to be swapped out, replacing shading algorithms and even allowing arbitrary raytracing and environment sampling. These hooks are now tied to LightWave 9.2's Node system.

The change to LightWave's workflow is again dramatic. Interactive node editing is essential for perfecting surface appearance, and FPrime 3 provides that instant feedback, even for Node shaders and materials. This includes even the powerful Conductor, Dielectric, Delta, Sigma, Skin, and Kappa SSS nodes.

As a quick example, the video to the left shows editing the setup and properties of a Dielectric glass jar. The most important feature to notice in the video is FPrime's immediate feedback. Even full resolution rendering is complete in FPrime in less than 25 seconds. The LightWave render of the same scene to the same quality takes seven minutes.

FPrime's raytracing engine has also been replaced, optimizing both speed and memory use. In particular, 64 bit rendering support required different internal data structures to prevent doubling memory use. The new core not only enabled compact 64 bit support, but also gives extra efficiency to the actual ray trace engine, boosting speed for 32 bit rendering too.

The original FPrime is single threaded, and one of our free major FPrime updates added multithreaded rendering. Since then, multiprocessor computers have become common, even expected. 4-core machines are now the bang-for-buck sweet spot, and soon 8-core will be the default for any professional 3D artist. Using all of your machine's power is important, and FPrime 3 now scales its rendering far beyond dual CPU support, and even 8 or more CPU machines are used effectively.

The internal changes to FPrime give new speed, new abilities, and new versatility to the engine. FPrime's integration with LightWave has also been tightened. There's no need to manage displacement or shader "trigger" plugins, reducing your distractions. Object display limitations are removed, showing subpatch and Catmull Clark objects with any displacement order, and even special subdivision level 0.

ACT cameras in LightWave 9.2 are fully supported in FPrime's preview and renderer. This includes not just the Perspective and Orthographic cameras, but also Baking, even as a preview. Multiple cameras, even of different types, can be viewed in simultaneous previews. The video at the left shows a scene with views of both a Perspective and Baking view.

FPrime 3's internal redesign pays off today with its full Node and ACT support, but more importantly, it means FPrime can be connected to the future LightWave advances more easily.

One last bonus is not an official feature of FPrime 3. As a temporary solution for the need for network rendering support, we now provide an unsupported tool allowing FPrime 3 to be used as a Screamernet node. While complex and sensitive to set up, it solves a major problem for many studios.

FPrime 3 requires with LightWave 8 or 9. FPrime Node and ACT support require a recent "Open Beta" version of 9.2.

FPrime 3 is available now. Upgrades are available to users of the original FPrime for $149, but we are offering a discount price of $125 until March 22.

From experience, we know that orders during launch can take several days to process. Thanks for understanding!


支援了所有shader,material node.....同時能邊調邊睇到效果....
較算圖經己高速很多的LW v9.2...更加高速......以幾倍計.....
解除了一些長久以來的限制....
同時支援了V9以後所有進階camera.....包括互動的Bake.....又系即時....
支援到8個內核CPU.....

[ed209]

終於support v9既node, 期待. 如能support sasquatch 就更好, 可惜要support sasqautch應沒有可能