[Bf-committers] GSoC Application :Physically Based Hair Shader

Brecht Van Lommel brechtvanlommel at pandora.be
Sat Mar 24 09:09:17 CET 2018


Hi,

In case you haven't seen it, there's a thread about this GSoC idea on the
bf-cycles mailing list which has feedback from Cycles developers:
https://lists.blender.org/pipermail/bf-cycles/2018-February/003237.html
https://lists.blender.org/pipermail/bf-cycles/2018-March/003292.html

It may be good to add more about the Cycles side and how you might test and
optimize the hair shader. Currently it's quite focused on the algorithm.
Since this is porting an existing implementation we hope there will be time
to polish the code and get user feedback.


For planning, here's some general advice that applies to most GSoC
projects. Students often make a plan that looks like:
* 3 weeks: read code
* 3 weeks: implement internals
* 3 weeks: implement UI
* 3 weeks: test and document

What usually happens is that by the time you can test the feature, many
things are broken. It then takes take a lot of time to identify which parts
of the code are causing which problems.

To avoid that, break up the project into smaller steps that can be studied,
implemented and tested one by one. A (basic) UI is useful to implement
first, so that you can start testing underlying code immediately when it is
implemented. For example consider these initial steps, where after each one
you can verify if things are working as expected:

* Add shader node in the UI
* Add node in Cycles kernel, creating a diffuse BSDF
* Create basic test files with this shader node
* Add new hair BSDF, with code just a copy of diffuse BSDF
* Port PBRT code
* Integrate into new hair BSDF eval() function
* ..

Regards,
Brecht.


On Sat, Mar 24, 2018 at 6:19 AM, uni sawano <sawanounicorn at gmail.com> wrote:

> Hello!
> I want to apply for GSoC this year. The proposal draft is here:
> https://docs.google.com/document/d/11_1fg0oC9gKSbMDxu_
> BczN5FlXy_3hG3YHvY5qA0AtM/edit?usp=sharing
> I will be glad to receive any feedback directly on the doc.
> Thanks
> Siyu Shi
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> https://lists.blender.org/mailman/listinfo/bf-committers
>


More information about the Bf-committers mailing list