Search found 56 matches

by hobold
Tue Jul 31, 2012 4:21 pm
Forum: General Development
Topic: Quadrics kinda sorta work for real time
Replies: 34
Views: 25211

Re: Quadrics kinda sorta work for real time

A subdivision algorithm based on the ZP-element does indeed exist: http://www.cise.ufl.edu/research/SurfLab/pre99-papers/9697.sss.pdf . But I dare say it results in clearly inferior surfaces than the other constructions mentioned so far. And no word if the ZP-element based algorithm actually makes i...
by hobold
Mon Jul 30, 2012 8:38 pm
Forum: General Development
Topic: Quadrics kinda sorta work for real time
Replies: 34
Views: 25211

Re: Quadrics kinda sorta work for real time

I am as new to this as you are. :-) The box splines were a proverbial "rose by the wayside" so to speak. That rose does have an intoxicating smell, though. For example, my gut feeling about similarities to wavelets is backed up by publications from the mathematical community. All box splines can ser...
by hobold
Sun Jul 29, 2012 10:42 pm
Forum: General Development
Topic: Quadrics kinda sorta work for real time
Replies: 34
Views: 25211

Re: Quadrics kinda sorta work for real time

Reading up on subdivision surfaces, I learned about box splines. A particular one of those, also known as Zwart-Powell element, can be reproduced exactly with quadrics: http://vectorizer.org/boar/basisNeu.png It is only C^1, but has some other interesting properties that remind me of wavelets. It mi...
by hobold
Fri Jul 27, 2012 5:42 pm
Forum: General Development
Topic: Quadrics kinda sorta work for real time
Replies: 34
Views: 25211

Re: Quadrics kinda sorta work for real time

well... if you can figure out a "good fit" to a Loop SubD, you can subdivide the mesh until the approximating quadrics fall below some error threshold. This is actually a very good idea. A much more interesting potential dead end than those that I had on my list. The subdivision process ought to lo...
by hobold
Fri Jul 27, 2012 1:25 pm
Forum: General Development
Topic: Quadrics kinda sorta work for real time
Replies: 34
Views: 25211

Re: Quadrics kinda sorta work for real time

Shinji Ogaki published a paper last year about how to ray trace Phong tessellations. I had skimmed it. As far as I understand, he basically renders Bezier triangles of degree two. This works out to an implicit degree four surface (a quartic). There is also a good paper by James Blinn et al on how t...
by hobold
Thu Jul 26, 2012 5:32 pm
Forum: General Development
Topic: Quadrics kinda sorta work for real time
Replies: 34
Views: 25211

Re: Quadrics kinda sorta work for real time

Seems like a fun project :) In a way, the fun is the only remaining reason behind the project (the hopefully humorous tone of the project home page is a consequence of that). I had much grander plans ... imagine if realtime raytracing was capable of not only doing more realistic lighting with less ...
by hobold
Wed Jul 25, 2012 2:00 pm
Forum: General Development
Topic: Quadrics kinda sorta work for real time
Replies: 34
Views: 25211

Re: Quadrics kinda sorta work for real time

I am making slow progress on the impractically slow tree building. I finally understand the geometry and math behind a sweep type algorithm based on the original box-quadric-intersection test. To find all relevant candidate split positions (at most four for a general quadric primitive), it is suffic...
by hobold
Thu Jul 19, 2012 8:18 pm
Forum: General Development
Topic: Discussion on GPU ray tracing
Replies: 3
Views: 3654

Re: Discussion on GPU ray tracing

The fundamental reason why GPUs require some sort of coherence to achieve their best performance lies in their basic architecture. All current GPU hardware is based on the SIMD principle ("Single Instruction Multiple Data"). This means that, regardless of all the syntactical sugar provided by CUDA o...
by hobold
Tue Jun 19, 2012 2:34 pm
Forum: Links & Papers
Topic: Intel Xeon Phi
Replies: 9
Views: 7683

Re: Intel Xeon Phi

As it's x86 chips, I'd love to plug it in and just see them in taskmanager... And then just say "run more threads" to your favourite rendering app of choice to spread over to there. but i guess it won't be that easy. but.. it should be. No, it won't be that easy. The three main reasons are: 1. No S...
by hobold
Thu Apr 19, 2012 10:25 am
Forum: General Development
Topic: Quadrics kinda sorta work for real time
Replies: 34
Views: 25211

Re: Quadrics kinda sorta work for real time

I miscalculated my initial figure for clock cycles per ray ... apparently I divided the number of rays by the number of threads once again. After a few more tweaks for better load balancing and faster code paths for a few special cases, I am now getting between 1.55M and 1.75M rays per second per co...