Hi Hsuan-Hao, Having a dedicated hair solver would be great. The same code is currently used for particles as well as hair, with "fake" cloth data being generated in addition to that for hair dynamics. This is a very cumbersome and complicated system, where the exact use of the data has to be checked in almost every function.
Particles will get a replacement next year. This leaves us with the option of removing/disabling the particle code from the current system and turn it into a pure hair feature, but eventually the whole thing should be recoded. As a rough outline, these are the steps i would suggest: * Implement a dedicated data structure for hair. Atm the geometric model for hair is based on spline interpolation, with "particle paths" being used as the control points. But there are a number of more suitable models around, e.g. - Modelling hair as strands, which supports very detailed curling without much additional data [1] - Volumetric hair [2] I'm far from being up-to-date on all the methods being used in the industry at this point, a thorough investigation of available methods is needed. * Integrate hair dynamics into a general simulation framework. This is another part of the planned physics improvements for next year: We want to solve the problems with object interaction by moving away from the current use of "modifiers" as the basis of simulations. I've written down a few thoughts on why this is a problem and how it can be solved by adding a scene-level simulation framework at [3]. * Adapt current sculpting/particle edit tools to work with the new hair data structures. Ideally the sculpting tools should be generalized in a way that allows use with different types of geometry where applicable, such as surfaces (2D), "strands" (1D) and point (0D) data. Special hair features such as curl parameters may still require specialized brush types. Regards, Lukas [1] http://hal.archives-ouvertes.fr/docs/00/52/02/70/PDF/finalcoursenotes2008.pdf [2] http://graphics.pixar.com/library/Hair/paper.pdf [3] http://wiki.blender.org/index.php/User:Phonybone/PhysicsRoadmap#Simulation_Framework On Thu, Dec 19, 2013 at 8:15 AM, 張軒豪 <[email protected]> wrote: > Hi all, > > My name is Hsuan-Hao Chang and I am new to this mailing list. > I am specially interested in 'Hair' in Blender. > > The hair sculpting tools in Blender is amazing but I want more. > Since Blender is open source, I decided to build some tools. > For example, > brush to orient strand of hair without change the shape, > utility to transfer hair shape from one to others, > roller helps create curl hair like Disney's Brave, > etc... > > Before that, I think it would be a great idea to ask the project owner and > everybody here first for advices. > > Also, I am interested in hair simulation. > >From the source code, I found the current 'hair simulation' using an > implicit cloth solver. > I have gave it a try, but the behavior of collision is not easy to control. > > Will we have any plan to implement a real hair solver? > If we do have a plan, I want to do some contributes on that. > It would be a necessary if we just made a great hair style and want to > animate it. > > Regards, > Hsuan-Hao > _______________________________________________ > Bf-committers mailing list > [email protected] > http://lists.blender.org/mailman/listinfo/bf-committers > _______________________________________________ Bf-committers mailing list [email protected] http://lists.blender.org/mailman/listinfo/bf-committers
