Beware of Google Desktop Search...

....if your diskspace is low, because it takes up considerable space. In my PC, the index file size is about 1 GB + . I thought i would be much smaller, but maybe google puts something more into its indexing algorithm so that the size grows bigger. But on the other hand, the search time is definitely faster than native windows XP searching program, so it's still worth the cost.
Anyway, since most of us have hundreds gigabyte of harddisk space, it wouldn't be a big problem, except that you're like me who like to install a lot lot software into the system =.=

digimax > all

lately, i was approached by Asus, Quanta, and some others, and i have rejected the chances to have an interview = =". Hopefully my decisions are right ones, and the work permit for digimax can be solved immediately. Otherwise maybe i have to call them again to ask for interview =.=

Computing barycentric coordinate

I find a good website that teach me how to compute a barycentric coordinate of point P in the triangle T(v0,v1,v2). The formula can also be used to find is whether the P is inside T, which is the problem i want to solve.

Here is the link to the full explanation.

In short, the formula is like this:

u = V₁-V_0;
v = V₂-V_0;
w = P-V_0;

The parametric coordinates (s,t) of P is the solution of the equation: w = su + tv, and the barycentric coordinates of P = b₀V₀ + b₁V₁ + b₂V₂ are equal to: b₀ = (1-s-t), b₁ = s, and b₂ = t which satisfy: b₀ + b₁ + b₂ = 1.

We can compute the solutions for s and t using only dot products as:

Thanks very much to Dan Sunday at http://softsurfer.com/index.html for providing the solution.

眷村

剛看了康熙某一集，邀請了一些在卷村長大的藝人．感覺真的很像我的家鄉，雖然我沒他們慘，但還是有那種ｆｅｅｌｉｎｇ．希望以後有出頭了也可以給我的家鄉一些貢獻!

社會局核銷真離譜!!

規定也太多了吧!! 真的很不合理!
連電話費都要著名是打給誰的,然後工作人員交接加減也要報告給他們,我覺得社會局根本就沒有信任這概念.我們有不是拿這些錢來做生意,是在幫災民呢!!
海嘯是2004年發生的,現在已經2007年了.我想如果捐款者知道他們捐出來的前三年過後還呆在社會局帳戶裡,會很氣吧!這樣捐款的benefit 就沒了.

老闆也一樣不合理.問題這麼嚴重了還不想直接下手,一直叫根本沒什麼經驗的人去處理,然後又希望趕快弄好.還說什麼 " 你已經要進入社會了" ,我有不是去做行政的工作,我是寫程式的勒!!
咳,不想管那麼多了,反正先把薪水要來吧.

translucent happy buddha

Here is my newest rendering of translucent material using PBRT.
The rendering time is quite slow, about 2000 seconds using P4 with 768MB RAM.
I am quite satisfied with the result. But maybe i need to add more surface reflection to make it more jade-like.

mood is temporary, class is permanent

can you learn to be classy?

panorama stitching

This was my VFx assignment #2. We're supposed to combine multiple images into one panorama image. 
Here is one of my better results:


 For the full-size image you can go here:
http://r4inm4ker.m3.ntu.edu.tw/ok2.jpg

digimax has just published a cool short movie

the story is somewhat similar to "Night at the Museum" comedy, but it's full CG movie. I just saw the trailer some days ago, and i was impressed as it is (IMO) as good as most of the current US CG movies.
Here's the trailer.

fwd: importance of writing

Cyy sent us a link to a very good article about the importance of writing. I paste some of the interesting content:

「我的朋友叫張懷恩他是我難亡的朋友他有很多很多專長他的專長是跑步跑很快、跳遠跳的很遠很遠、打電腦長長說有什麼遊ㄒ一ˋ他知道很多很的遊ㄒㄧˋ他很短可以他很跑很快的。老師一定有ㄧˊ問說：「他為什麼是我最今我難亡的朋友我也不知道我知道我跟他跟我是很好很的朋友所以他是令我難亡的朋友。」

這是桃園某小學五年級學生作文：「我最難忘的朋友」，這篇作文沒有標點，錯字多，內容顯得空洞貧乏、毫無思路可言。但這個小朋友是來自中產家庭，他也是民國九十年，政府推動「九年一貫」教育改革，降低國語課時數後的「產物」。



Here's the link for you who's interested to read the full article.

成功加入太極影音了!

昨天去面試了第二次,跟R&D主管 chris 聊了一下(大部分跟工作無關 =.=").雖然薪水起跳有點低,但相信我一定會學到很多有趣的東西.我覺得還是很直得啦.好比在我沒什麼興趣的公司工作,雖然有較多錢可以拿但感覺也沒什麼意義 . 七月開始工作,非常期待呢!!

Improving last semester's project

After unsuccessfully simulating jade's appereance last semester, i try to improve my codes now.
Here's the current result:
















I will try to improve my formula to create a better one as i am also working to implement multi-layered subsurface scattering for translucent materials.
I will try my best to match the results from original paper's authors. The images below were linked from http://graphics.ucsd.edu/papers/layered/

Subdivision Surfaces with Loop's algorithm

My first assignment on Geometric Modeling was about subdivision surfaces. I implemented Loop's algorithm, and work on VC7 using CGAL library. Most of the time on this assignment i was dealing with how to work with the library, since i found it very confusing = =" (not the library's fault though). I also use some of the code from PBRT for counting the vertices' weights.
Here's the result of my program:

original shape



















after 3 levels of subdivision

new chapter..again!

For the n-th times,  i start all over for this blog. Hopefully this time i can fill it with interesting stuffs!! ^ ^
Some weeks ago, i finished my first assignment on Digital Visual Effects about High Dynamic Range (HDR) images creation. I use Paul Debevec method for this assignment. Anyway, here's the results :


It's still not very satisfying, partly because i didn't take enough pictures and my camera's lens was a bit dirty. I'll try to take better pictures when i have spare times.