[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math

Search:


View post   

>> No.9677764 [View]
File: 103 KB, 381x997, why 2.jpg [View same] [iqdb] [saucenao] [google]
9677764

>>9677745
when people talk about 3d graphics they say it's matrix multiplication. here it is in this book in fact. if you dont want to read all this it's just making a matrix for (x,y,z)->(x/(1-z/d),y/(1-z/d),0). if I were gonna do this on a computer, again I would just do it directly without fucking with matrix multiplication at all. but they extend it out into homogeneous coordinates, and then finally say "to obtain the (x',y') coordinates you actually needed corresponding to ((x/(1-z/d),y/(1-z/d)), after multiplying (x,y,z,1) by the homogeneous 4x4 matrix, just divide the x and y of the vector product by the last thing in that vector". WHY THE FUCK WOULD I WANT TO GO THROUGH ALL OF THAT? that's effectively just undoing the entire process of constructing this homogeneous matrix in the first place. why not just have a function (x,y,z)->(x/(1-z/d),y/(1-z/d),0) DIRECTLY? this section is literally called "computer graphics". why would i want to apply anything in here to computer graphics if it's just more steps?

Navigation
View posts[+24][+48][+96]