<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	>
<channel>
	<title>Comments on: Screencast frame-rate tips</title>
	<atom:link href="http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips/feed" rel="self" type="application/rss+xml" />
	<link>http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips</link>
	<description>The personal weblog of Peter Hosey.</description>
	<pubDate>Mon, 08 Sep 2008 06:33:39 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.5.1</generator>
		<item>
		<title>By: Peter Hosey</title>
		<link>http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-12055</link>
		<dc:creator>Peter Hosey</dc:creator>
		<pubDate>Wed, 14 Mar 2007 20:01:16 +0000</pubDate>
		<guid isPermaLink="false">http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-12055</guid>
		<description>Yeah, Snapz Pro is looking mighty at this point. They're definitely doing something right.

And from the output of nm 'Snapz Pro X.app/Contents/MacOS/Snapz Pro X':

&lt;blockquote&gt;         U _CGRegisterScreenRefreshCallback&lt;/blockquote&gt;

So it looks like they are using that. And iShowU isn't—no such line.</description>
		<content:encoded><![CDATA[<p>Yeah, Snapz Pro is looking mighty at this point. They're definitely doing something right.</p>
<p>And from the output of nm 'Snapz Pro X.app/Contents/MacOS/Snapz Pro X':</p>
<blockquote cite="http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-"><p>         U _CGRegisterScreenRefreshCallback</p>
</blockquote>
<p>So it looks like they are using that. And iShowU isn't—no such line.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Simone Manganelli</title>
		<link>http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-12054</link>
		<dc:creator>Simone Manganelli</dc:creator>
		<pubDate>Wed, 14 Mar 2007 19:53:11 +0000</pubDate>
		<guid isPermaLink="false">http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-12054</guid>
		<description>Oh, by the way, I forgot to mention: Snapz Pro X isn't even a universal binary, and yet it still can capture 30 fps full-screen captures (I forgot to mention my resolution: 1280x800).  This seems to me to be a pretty impressive feat on the part of Ambrosia.</description>
		<content:encoded><![CDATA[<p>Oh, by the way, I forgot to mention: Snapz Pro X isn't even a universal binary, and yet it still can capture 30 fps full-screen captures (I forgot to mention my resolution: 1280x800).  This seems to me to be a pretty impressive feat on the part of Ambrosia.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Peter Hosey</title>
		<link>http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-12053</link>
		<dc:creator>Peter Hosey</dc:creator>
		<pubDate>Wed, 14 Mar 2007 19:41:08 +0000</pubDate>
		<guid isPermaLink="false">http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-12053</guid>
		<description>&lt;blockquote&gt;Peter: from what you've written here, it sounds like iShowU is very unoptimized for screencasts (I think I'm using that term correctly -- "screencast" means a movie of your screen, right?).&lt;/blockquote&gt;

Yes, it does.

It's quite possible, as iShowU is the newcomer. Also, Snapz may be using the Quartz Services callback that Ken mentioned. I'll give Snapz a try and report back.

&lt;blockquote&gt;I'm taking a full-screen 30 fps screencast on my Intel Core Duo MacBook right now with no perceivable lag whatsoever, and the CPU usage for Snapz hovers around 10-15% when there's not too many changes on the screen (i.e.: you're just typing). If you're moving around windows, Snapz's CPU usage can jump to 50-60%, but that load is easily managed by today's hardware. The only time where the CPU usage gets maxed out is when Snapz is actually encoding the video into a movie file after you've finished capturing the footage, and that's not exactly surprising.&lt;/blockquote&gt;

Interesting.

&lt;blockquote&gt;I'm not sure if your bandwidth argument holds up, because it seems to imply that you're writing the contents directly to disk. But if you're just caching the video in memory (and you have enough to do so), I don't think bandwidth is an issue.&lt;/blockquote&gt;

With iShowU, I've used a RAM disk at times. The problem is that the Raw compressor fills my 64-MiB RAM disk in a few seconds. Over the course of an entire recording, it would probably fill the rest of my RAM too, if it could.

&lt;blockquote&gt;You might want to check out Snapz Pro X again on your Mac Pro, and compare performance to iShowU.&lt;/blockquote&gt;</description>
		<content:encoded><![CDATA[<blockquote cite="http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-"><p>Peter: from what you've written here, it sounds like iShowU is very unoptimized for screencasts (I think I'm using that term correctly -- "screencast" means a movie of your screen, right?).</p>
</blockquote>
<p>Yes, it does.</p>
<p>It's quite possible, as iShowU is the newcomer. Also, Snapz may be using the Quartz Services callback that Ken mentioned. I'll give Snapz a try and report back.</p>
<blockquote cite="http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-"><p>I'm taking a full-screen 30 fps screencast on my Intel Core Duo MacBook right now with no perceivable lag whatsoever, and the CPU usage for Snapz hovers around 10-15% when there's not too many changes on the screen (i.e.: you're just typing). If you're moving around windows, Snapz's CPU usage can jump to 50-60%, but that load is easily managed by today's hardware. The only time where the CPU usage gets maxed out is when Snapz is actually encoding the video into a movie file after you've finished capturing the footage, and that's not exactly surprising.</p>
</blockquote>
<p>Interesting.</p>
<blockquote cite="http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-"><p>I'm not sure if your bandwidth argument holds up, because it seems to imply that you're writing the contents directly to disk. But if you're just caching the video in memory (and you have enough to do so), I don't think bandwidth is an issue.</p>
</blockquote>
<p>With iShowU, I've used a RAM disk at times. The problem is that the Raw compressor fills my 64-MiB RAM disk in a few seconds. Over the course of an entire recording, it would probably fill the rest of my RAM too, if it could.</p>
<blockquote cite="http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-"><p>You might want to check out Snapz Pro X again on your Mac Pro, and compare performance to iShowU.</p>
</blockquote>
]]></content:encoded>
	</item>
	<item>
		<title>By: Simone Manganelli</title>
		<link>http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-12051</link>
		<dc:creator>Simone Manganelli</dc:creator>
		<pubDate>Wed, 14 Mar 2007 19:09:36 +0000</pubDate>
		<guid isPermaLink="false">http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-12051</guid>
		<description>Peter: from what you've written here, it sounds like iShowU is very unoptimized for screencasts (I think I'm using that term correctly -- "screencast" means a movie of your screen, right?).

Snapz Pro X can *easily* handle full screen video capture of 30 fps, and it has been able to do so on most hardware for a while now, including, IIRC, G4s.  (I couldn't do it on my iMac G4 800 MHz, but their developers claimed I should have been able to, and from what they told me, I was running up against a very rare bug that they could never reproduce.)  I'm taking a full-screen 30 fps screencast on my Intel Core Duo MacBook right now with no perceivable lag whatsoever, and the CPU usage for Snapz hovers around 10-15% when there's not too many changes on the screen (i.e.: you're just typing).  If you're moving around windows, Snapz's CPU usage can jump to 50-60%, but that load is easily managed by today's hardware.  The only time where the CPU usage gets maxed out is when Snapz is actually encoding the video into a movie file after you've finished capturing the footage, and that's not exactly surprising.

I'm not sure if your bandwidth argument holds up, because it seems to imply that you're writing the contents directly to disk.  But if you're just caching the video in memory (and you have enough to do so), I don't think bandwidth is an issue.  I don't see why you can't take Ambrosia's approach: put off the disk writing until later when you don't need to worry about continuing to capture footage.

You might want to check out Snapz Pro X again on your Mac Pro, and compare performance to iShowU.</description>
		<content:encoded><![CDATA[<p>Peter: from what you've written here, it sounds like iShowU is very unoptimized for screencasts (I think I'm using that term correctly -- "screencast" means a movie of your screen, right?).</p>
<p>Snapz Pro X can *easily* handle full screen video capture of 30 fps, and it has been able to do so on most hardware for a while now, including, IIRC, G4s.  (I couldn't do it on my iMac G4 800 MHz, but their developers claimed I should have been able to, and from what they told me, I was running up against a very rare bug that they could never reproduce.)  I'm taking a full-screen 30 fps screencast on my Intel Core Duo MacBook right now with no perceivable lag whatsoever, and the CPU usage for Snapz hovers around 10-15% when there's not too many changes on the screen (i.e.: you're just typing).  If you're moving around windows, Snapz's CPU usage can jump to 50-60%, but that load is easily managed by today's hardware.  The only time where the CPU usage gets maxed out is when Snapz is actually encoding the video into a movie file after you've finished capturing the footage, and that's not exactly surprising.</p>
<p>I'm not sure if your bandwidth argument holds up, because it seems to imply that you're writing the contents directly to disk.  But if you're just caching the video in memory (and you have enough to do so), I don't think bandwidth is an issue.  I don't see why you can't take Ambrosia's approach: put off the disk writing until later when you don't need to worry about continuing to capture footage.</p>
<p>You might want to check out Snapz Pro X again on your Mac Pro, and compare performance to iShowU.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Peter Hosey</title>
		<link>http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-12049</link>
		<dc:creator>Peter Hosey</dc:creator>
		<pubDate>Wed, 14 Mar 2007 18:11:48 +0000</pubDate>
		<guid isPermaLink="false">http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-12049</guid>
		<description>Ken: That information is probably better delivered to the iShowU people. ;)</description>
		<content:encoded><![CDATA[<p>Ken: That information is probably better delivered to the iShowU people. ;)</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: ken</title>
		<link>http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-12045</link>
		<dc:creator>ken</dc:creator>
		<pubDate>Wed, 14 Mar 2007 16:07:42 +0000</pubDate>
		<guid isPermaLink="false">http://boredzo.org/blog/archives/2007-03-14/screencast-frame-rate-tips#comment-12045</guid>
		<description>Check out CGRegisterScreenRefreshCallback.  That lets you register a CG callback function for updated screen rects.  If you _know_ that the only rect that has changed is the one containing the mouse, you ought to be able to avoid both compression and disk space costs for the unchanged area.

Not that I've tried yet.</description>
		<content:encoded><![CDATA[<p>Check out CGRegisterScreenRefreshCallback.  That lets you register a CG callback function for updated screen rects.  If you _know_ that the only rect that has changed is the one containing the mouse, you ought to be able to avoid both compression and disk space costs for the unchanged area.</p>
<p>Not that I've tried yet.</p>
]]></content:encoded>
	</item>
</channel>
</rss>

<!-- Dynamic Page Served (once) in 0.202 seconds -->
