利用者:DingTo/GSoC 2013/Weekly Reports/Week7

提供: wiki
< 利用者:DingTo‎ | GSoC 2013‎ | Weekly Reports
2018年6月29日 (金) 05:57時点におけるYamyam (トーク | 投稿記録)による版 (1版 をインポートしました)
(差分) ← 古い版 | 最新版 (差分) | 新しい版 → (差分)
移動先: 案内検索

Week 7

What I did this week

  • Brecht reviewed my branch this week (thanks a lot Brecht) so I could work on getting some things into trunk. I fixed some things in my branch which came up in the review (Blackbody node, Ray Depth, Vector Transform node...), and also added a new type for the Vector Transform node, which is "Normal".
  • I merged the Blackbody Node, Vector Transform node, HSV Combine/Separate nodes and the Ray Depth shading feature to trunk.
  • SSS Render Passes: I changed some variables there, which also came up during the code review and commited that. Brecht also gave me a hint how to proceed further. I have some (uncommitted) changes, which I hope to improve this weekend.
  • I also spend some time on trunk Cycles:
    • Added Presents to the Sampling panel, so people can quickly jump between render settings for final/preview purposes.
    • For the Non-Progressive integrator there is a new "Squared Samples" option which Matt Heimlich implemented (I just did some small tweaks), so the user can more easily tweak the sampling values. The idea here is, that increasing samples from e.g. 80 to 82 is not noticeable, a increase from 8*8 to 9*9 on the other hand is. This makes it easier to adjust and tune the different amounts of Sample types, to get a satisfying result.
    • Found a small improvement for Non-Progressive integrator, basically removes an if-branch to avoid checking for Transparent BSDF in one loop, which is not needed.

Next week

  • Continue on SSS render passes and hopefully some progress with the Motion Multiplier, which I have to admit is still a bit of a mystery for me.

Questions

None.