<?xml version="1.0" encoding="UTF-8"?>
	<rss version="2.0"
		xmlns:content="http://purl.org/rss/1.0/modules/content/"
		xmlns:wfw="http://wellformedweb.org/CommentAPI/"
		xmlns:dc="http://purl.org/dc/elements/1.1/"
		xmlns:atom="http://www.w3.org/2005/Atom"

			>

	<channel>
		<title>Cycling 74  &#187;  Topic: point clouds</title>
		<atom:link href="http://cycling74.com/forums/topic/point-clouds/feed" rel="self" type="application/rss+xml" />
		<link>http://cycling74.com/forums/topic/point-clouds/feed</link>
		<description></description>
		<pubDate>Wed, 19 Jun 2013 12:20:00 +0000</pubDate>
		<generator>http://bbpress.org/?v=2.2.4</generator>
		<language></language>

		
														
					
				<item>
					<guid>http://cycling74.com/forums/topic/point-clouds/#post-24421</guid>
					<title><![CDATA[point clouds]]></title>
					<link>http://cycling74.com/forums/topic/point-clouds/#post-24421</link>
					<pubDate>Sat, 11 Feb 2006 23:23:41 +0000</pubDate>
					<dc:creator>Barney Haynes</dc:creator>

					<description>
						<![CDATA[
						<p>Greetings,</p>
<p>Has anyone worked with point clouds in open GL? I tried the brute force <br />
method of reading each point into jit.gl.sketch, but it took along <br />
time. The results were a tad sluggish averaging 8 to 9 fps on G5 dual <br />
2.7. Is there is a more efficient way of doing this? If not can anyone <br />
recommend a plug in or application for the mac that will convert point <br />
clouds to obj files? I&#8217;ve seen several on line but the cost is <br />
prohibitive enough to preclude experimentation.</p>
<p>Thanks</p>
<p>Barney</p>
						]]>
					</description>

					
					
				</item>

			
				<item>
					<guid>http://cycling74.com/forums/topic/point-clouds/#post-70567</guid>
					<title><![CDATA[Re: point clouds]]></title>
					<link>http://cycling74.com/forums/topic/point-clouds/#post-70567</link>
					<pubDate>Sat, 11 Feb 2006 23:43:00 +0000</pubDate>
					<dc:creator>Wesley Smith</dc:creator>

					<description>
						<![CDATA[
						<p>Well, I learned a new term today.  I don&#8217;t see anyway of efficiently<br />
doing this in jitter right now.  I think the best way to go would be<br />
to modify the jit.gl.gridshape SDK example and change the @position<br />
and @scale attributes to matrix inputs for drawing, say, thousands of<br />
spheres per frame.  You would pass in 2 3-plane nx1 matrices with<br />
position and scale data and on each render pass, it would call draw<br />
for every scale/position pair, just moving the sphere around.</p>
<p>wes</p>
						]]>
					</description>

					
					
				</item>

			
				<item>
					<guid>http://cycling74.com/forums/topic/point-clouds/#post-70568</guid>
					<title><![CDATA[Re: point clouds]]></title>
					<link>http://cycling74.com/forums/topic/point-clouds/#post-70568</link>
					<pubDate>Mon, 13 Feb 2006 02:03:02 +0000</pubDate>
					<dc:creator>Barney Haynes</dc:creator>

					<description>
						<![CDATA[
						<p>Thanks Wes,<br />
Will the jit.gl.gridshape solution run more efficiently on the graphics <br />
card once all the data points are loaded?<br />
Barney</p>
						]]>
					</description>

					
					
				</item>

			
				<item>
					<guid>http://cycling74.com/forums/topic/point-clouds/#post-70569</guid>
					<title><![CDATA[Re: point clouds]]></title>
					<link>http://cycling74.com/forums/topic/point-clouds/#post-70569</link>
					<pubDate>Mon, 13 Feb 2006 03:23:16 +0000</pubDate>
					<dc:creator>Wesley Smith</dc:creator>

					<description>
						<![CDATA[
						<p>You know,<br />
I&#8217;m not incredibly familiar with the point cloud litterature, so I&#8217;m<br />
going to refrain from answering your question as I don&#8217;t want to<br />
posture a wild guess.  That said, the best thing to do is avoid tons<br />
of max messages.  Point clouds seem like an ideal candidate for their<br />
own external as from what cursoiry knowledge I have they require a<br />
decently sophisticated data structure to manage their rendering.  From<br />
what I could tell, people were using octrees to choose the detail of<br />
the rendering.  If you set this up in a jitter external, you could<br />
have a float parameter for precision that would traverse the octree to<br />
a particular level of detail.  This way, you could adjust the detail<br />
to the machine you&#8217;re running on and get interactive frame-rates.</p>
<p>Hope this helps.  If you&#8217;re really into the point cloud thing, I would<br />
take this opportunity to learn how to make a jitter external.  Once<br />
you make one successfully, you realize how simple it really is and<br />
your jittering will be that much better since you can overcome hacking<br />
objects together and instead make your own, saving you much processing<br />
power.</p>
<p>good luck,<br />
wes</p>
<p>PS, on yahoo groups, there is a jitter-dev list.  you may find the<br />
archives handy when you start out.</p>
						]]>
					</description>

					
					
				</item>

			
				<item>
					<guid>http://cycling74.com/forums/topic/point-clouds/#post-70570</guid>
					<title><![CDATA[Re: point clouds]]></title>
					<link>http://cycling74.com/forums/topic/point-clouds/#post-70570</link>
					<pubDate>Tue, 14 Feb 2006 00:55:12 +0000</pubDate>
					<dc:creator>Barney Haynes</dc:creator>

					<description>
						<![CDATA[
						<p>Wes,<br />
Thanks again.<br />
Sounds like my summer vacation plans are starting to become a little <br />
more concrete.<br />
Barney</p>
						]]>
					</description>

					
					
				</item>

					
		
	</channel>
	</rss>

