atom feed19 messages in at.iem.pd-listRe: [PD] spread Gem-computation over ...
FromSent OnAttachments
Roman HaefeliFeb 27, 2007 3:19 pm 
cyrille henryFeb 27, 2007 3:43 pm 
Roman HaefeliFeb 28, 2007 1:08 am 
IOhannes m zmoelnigFeb 28, 2007 5:45 am 
cyrille henryFeb 28, 2007 5:53 am 
Kyle KlipowiczFeb 28, 2007 7:05 am 
Frank BarknechtFeb 28, 2007 7:25 am 
chris clepperFeb 28, 2007 7:47 am 
cyrille henryFeb 28, 2007 8:04 am 
cyrille henryFeb 28, 2007 8:15 am 
Roman HaefeliFeb 28, 2007 2:23 pm 
Roman HaefeliFeb 28, 2007 2:35 pm 
Roman HaefeliFeb 28, 2007 2:53 pm 
cyrille henryFeb 28, 2007 3:08 pm 
Chris McCormickFeb 28, 2007 6:28 pm 
Roman HaefeliFeb 28, 2007 6:34 pm.pd
cyrille henryMar 1, 2007 2:18 am 
Roman HaefeliMar 1, 2007 3:28 am 
cyrille henryMar 1, 2007 3:39 am 
Subject:Re: [PD] spread Gem-computation over several dsp-cycles (?)
From:chris clepper (cgcl@gmail.com)
Date:Feb 28, 2007 7:47:50 am
List:at.iem.pd-list

This is an optimization if your geometry doesn't change. A display list loads all of the vertex, color, texcoord and normal data onto the card and then just tells the card to render it right from VRAM. The model object uses display lists for speed. If you need 10,000 cubes it would be faster to make a model of the cube and load it that way.

Cyrille's trick is to set up a render chain and render that once into the display list. After that only the gemhead feeding GEMglCallList is rendered every frame.

On 2/28/07, Frank Barknecht <fb@footils.org> wrote:

Hallo, cyrille henry hat gesagt: // cyrille henry wrote:

don't forget you can optimized your patch with GEMglGenList /

GEMglNewList etc in order to create a display list so that you patch could resume in :

repeat 1000 [ GEMglCallList

I've never hear of this optimization, do you have an example for it?

Ciao