Comments on: ACMR and ATVR http://www.realtimerendering.com/blog/acmr-and-atvr/ Tracking the latest developments in interactive rendering techniques Mon, 17 Jun 2013 03:17:13 +0000 hourly 1 http://wordpress.org/?v=3.4.1 By: Eric http://www.realtimerendering.com/blog/acmr-and-atvr/comment-page-1/#comment-47 Eric Mon, 30 Mar 2009 17:13:40 +0000 http://www.realtimerendering.com/blog/?p=83#comment-47 Yes, the optimum ACMR is 0.5 for a regular grid (or any closed mesh, for that matter - it's all from Euler's geometric formula; see pages 554-555 of the 3rd edition). I was indeed referring to the fact that for a single tristrip the ACMR approaches 1.0 at best (the more triangles, the closer to 1.0), since each vertex is shared by at most 3 triangles. Reading over my post, I made a typo, saying "meshes" where I meant to say "tristrips". Fixed, and sorry about that. At least we got it right in the book. BTW, the "non-bizarre mesh" reference is like what you point out, that you could have some silly mesh like a single triangle repeated a million times. With this pathological case the ACMR drops nearly to 0.0: 3 vertices/million triangles. Naty's the blogmaster, I've passed on your OpenID comment to him. Yes, the optimum ACMR is 0.5 for a regular grid (or any closed mesh, for that matter – it’s all from Euler’s geometric formula; see pages 554-555 of the 3rd edition). I was indeed referring to the fact that for a single tristrip the ACMR approaches 1.0 at best (the more triangles, the closer to 1.0), since each vertex is shared by at most 3 triangles. Reading over my post, I made a typo, saying “meshes” where I meant to say “tristrips”. Fixed, and sorry about that. At least we got it right in the book.

BTW, the “non-bizarre mesh” reference is like what you point out, that you could have some silly mesh like a single triangle repeated a million times. With this pathological case the ACMR drops nearly to 0.0: 3 vertices/million triangles.

Naty’s the blogmaster, I’ve passed on your OpenID comment to him.

]]> By: Arseny Kapoulkine http://www.realtimerendering.com/blog/acmr-and-atvr/comment-page-1/#comment-46 Arseny Kapoulkine Mon, 30 Mar 2009 05:36:26 +0000 http://www.realtimerendering.com/blog/?p=83#comment-46 Optimum ACMR for a regular grid is ~0.5, so 1.0 is kind of wrong. ACMR of 1.0 is the ACMR for any non-indexed strip sequence (without degenerates, if you exclude them from counting, and for any sequence if you include everything). Btw, registration on a blog w/out OpenID is kind of lame. Optimum ACMR for a regular grid is ~0.5, so 1.0 is kind of wrong. ACMR of 1.0 is the ACMR for any non-indexed strip sequence (without degenerates, if you exclude them from counting, and for any sequence if you include everything).

Btw, registration on a blog w/out OpenID is kind of lame.

]]>