1
\$\begingroup\$

I have two point clouds of n points:

  1. A dense ball-shape point cloud with many points inside
  2. A spherical point cloud with points on its surface.

Is there (really) any difference between rendering (complexity/time) of these two point clouds?

How does the distortion of points affect the complexity of the rasterisation process?

\$\endgroup\$
3
  • \$\begingroup\$ Could you clarify what you mean by "distortion of points" and why you suspect it alone could affect the complexity? Also, ball vs. sphere usually means the ball has (a lot) more points than the sphere cloud. That alone could be a trivial difference. \$\endgroup\$ Commented Feb 19, 2015 at 11:33
  • \$\begingroup\$ @teodron, The question is self contained. Distortion of points may have impacts on depth and occlusion. please see the comment from ClassicThunder below. I am looking for a complete (detailed) answer. \$\endgroup\$ Commented Feb 20, 2015 at 13:01
  • \$\begingroup\$ thanks, but I am not in a position to answer your question. What I fear is that other people would not understand the terminology. Once again, could you explain in one sentence or even a picture what you mean by "distortion of points" that "may have an impact on depth or occlusion"? As I said, a sphere has way less points than a ball, so there's a higher amount of points to be processed. In a way, probably this might help you: http.developer.nvidia.com/GPUGems/gpugems_ch29.html. Cheers! \$\endgroup\$ Commented Feb 20, 2015 at 14:09

2 Answers 2

Reset to default
2
\$\begingroup\$

No, from rasterization point of view both are the same. GPU processes all points independently and doesn't care about their position.

\$\endgroup\$
5
  • 1
    \$\begingroup\$ Your answer, while being technically correct (the best kind of correct), isn't very useful. Explaining why the rasterization behaves the same in both situations would help. Having to ask this question demonstrates that OP lacks an understanding of the underlying technologies so diving into how depth does not effect the complexity of the rasterization process would be useful for him and future visitors. \$\endgroup\$ Commented Aug 29, 2014 at 17:42
  • \$\begingroup\$ Why does this post deserve downvotes? It doesn't answer the question? \$\endgroup\$ Commented Aug 29, 2014 at 17:50
  • \$\begingroup\$ I didn't downvote. But I would assume who ever did also though your answer was not helpful, probably for the same reasons I specified. \$\endgroup\$ Commented Aug 29, 2014 at 18:01
  • \$\begingroup\$ Down votes happen. Try not to take it personally and resort to name calling. Thanks. Based on your other answers, I'd suggest putting a little more effort into expanding your answers. \$\endgroup\$ Commented Aug 29, 2014 at 19:00
  • 1
    \$\begingroup\$ @JeffMiles, I appreciate your effort to answer my question (in brief). However, please note that I am looking for a complete (detailed) answer to clarify my doubts (I agree with @ClassicThunder). I really appreciate if a senior member shed lights on me. \$\endgroup\$ Commented Aug 29, 2014 at 20:24
2
\$\begingroup\$

In theory, it will take the same time to render regardless. However, if a point occluded points in the back, then some pixels may be discarded before they are rasterized (because they failed the depth test), and I can imagine that that would be slightly faster.

\$\endgroup\$

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.