<?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: How To: Implement DirectShow Filter using DirectX Media Object DMO (Part 3: Persistence, Automation and Property Pages)</title>
	<atom:link href="http://alax.info/blog/433/feed" rel="self" type="application/rss+xml" />
	<link>http://alax.info/blog/433</link>
	<description>Software Production Line</description>
	<pubDate>Tue, 06 Jan 2009 09:10:14 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.6.3</generator>
		<item>
		<title>By: jackbowman</title>
		<link>http://alax.info/blog/433#comment-5646</link>
		<dc:creator>jackbowman</dc:creator>
		<pubDate>Tue, 16 Sep 2008 05:41:44 +0000</pubDate>
		<guid isPermaLink="false">http://alax.info/blog/?p=433#comment-5646</guid>
		<description>Roman,

Just FYI.  My problems with no image width or height (Extent in your code) disappeared when I dumped ffdshow (which had been inserted by GraphEdt) and replaced it with AVIDecoder. You were correct about the problem being a filter problem and not a GraphEdt problem.

Thanks again.

Jack</description>
		<content:encoded><![CDATA[<p>Roman,</p>
<p>Just FYI.  My problems with no image width or height (Extent in your code) disappeared when I dumped ffdshow (which had been inserted by GraphEdt) and replaced it with AVIDecoder. You were correct about the problem being a filter problem and not a GraphEdt problem.</p>
<p>Thanks again.</p>
<p>Jack</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: alax</title>
		<link>http://alax.info/blog/433#comment-5645</link>
		<dc:creator>alax</dc:creator>
		<pubDate>Mon, 15 Sep 2008 07:06:37 +0000</pubDate>
		<guid isPermaLink="false">http://alax.info/blog/?p=433#comment-5645</guid>
		<description>I would expect the mentioned scenario working without issues. There is probably a problem somewhere in your code, e.g. media type initialization.</description>
		<content:encoded><![CDATA[<p>I would expect the mentioned scenario working without issues. There is probably a problem somewhere in your code, e.g. media type initialization.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: jackbowman</title>
		<link>http://alax.info/blog/433#comment-5644</link>
		<dc:creator>jackbowman</dc:creator>
		<pubDate>Mon, 15 Sep 2008 07:01:55 +0000</pubDate>
		<guid isPermaLink="false">http://alax.info/blog/?p=433#comment-5644</guid>
		<description>Roman,
My input pin is connected to the output pin (in my test) of ffdshow Video Decoder.  I needed 24bpp RGB frames to encode, so I added a .avi file and GraphEdt always adds in AVI Splitter, ffdshow Video Decoder, and a video renderer.  I replaced the renderer with my encoder filter and a null renderer.  I would actually like to find a way to write the data to a .avi file but I do not seem to have a filter to do that so I may have to modify my filter to write the data to disk itself.

I did try changing the code to not specify the frame size but for some reason the result kept comming back as 3840 x 720.  I have no idea why as the video data I was using was usually 320 x 240 and never had a width over 500.

Jack</description>
		<content:encoded><![CDATA[<p>Roman,<br />
My input pin is connected to the output pin (in my test) of ffdshow Video Decoder.  I needed 24bpp RGB frames to encode, so I added a .avi file and GraphEdt always adds in AVI Splitter, ffdshow Video Decoder, and a video renderer.  I replaced the renderer with my encoder filter and a null renderer.  I would actually like to find a way to write the data to a .avi file but I do not seem to have a filter to do that so I may have to modify my filter to write the data to disk itself.</p>
<p>I did try changing the code to not specify the frame size but for some reason the result kept comming back as 3840 x 720.  I have no idea why as the video data I was using was usually 320 x 240 and never had a width over 500.</p>
<p>Jack</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: alax</title>
		<link>http://alax.info/blog/433#comment-5643</link>
		<dc:creator>alax</dc:creator>
		<pubDate>Mon, 15 Sep 2008 05:37:41 +0000</pubDate>
		<guid isPermaLink="false">http://alax.info/blog/?p=433#comment-5643</guid>
		<description>Jack,

This is not a GraphEdit issue, but the way how the filter connects to its upstream peer, on the input pin. That is it depends on what exactly kind of filter you are connecting your DMO wrapper filter with.

m_LastKnownInputExtent looks more like a workaround but always worked for me a number of times. An alternate method instead of enumerating 320,240 by default is to provide a media type with major type and subtype initialized, but the rest of the fields empty. This may be a better way (which I never needed to even try though) and I suggest that you give it a try. And also mention what is the filter being connected to the input pin.</description>
		<content:encoded><![CDATA[<p>Jack,</p>
<p>This is not a GraphEdit issue, but the way how the filter connects to its upstream peer, on the input pin. That is it depends on what exactly kind of filter you are connecting your DMO wrapper filter with.</p>
<p>m_LastKnownInputExtent looks more like a workaround but always worked for me a number of times. An alternate method instead of enumerating 320,240 by default is to provide a media type with major type and subtype initialized, but the rest of the fields empty. This may be a better way (which I never needed to even try though) and I suggest that you give it a try. And also mention what is the filter being connected to the input pin.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: jackbowman</title>
		<link>http://alax.info/blog/433#comment-5642</link>
		<dc:creator>jackbowman</dc:creator>
		<pubDate>Sun, 14 Sep 2008 22:07:54 +0000</pubDate>
		<guid isPermaLink="false">http://alax.info/blog/?p=433#comment-5642</guid>
		<description>Roman,

I am getting something that I cannot account for, and I am hoping you can help.

I never seem to get a valid input image extent (width and height) on any of the calls from GraphEdt.exe.

For some reason, I always seem to get 0,0 for the extents, and I notice that you seem to get around the problem by using a default of 320,240 which replaces the 0,0 in GetOutputType() and by using m_LastKnownInputExtent.

This works fine for your filter until a video with a larger frame size is used, and then the video displayed when the graph is run is truncated.  I get only the 320 by 240 upper left hand area of a video with frame size 496 x 368.  I assume that because GraphEdt sees the 320 x 240 as a cropping window, it avoids reading and writing off the end of the buffer.

Do you know how to get around this problem?

Thanks,
Jack</description>
		<content:encoded><![CDATA[<p>Roman,</p>
<p>I am getting something that I cannot account for, and I am hoping you can help.</p>
<p>I never seem to get a valid input image extent (width and height) on any of the calls from GraphEdt.exe.</p>
<p>For some reason, I always seem to get 0,0 for the extents, and I notice that you seem to get around the problem by using a default of 320,240 which replaces the 0,0 in GetOutputType() and by using m_LastKnownInputExtent.</p>
<p>This works fine for your filter until a video with a larger frame size is used, and then the video displayed when the graph is run is truncated.  I get only the 320 by 240 upper left hand area of a video with frame size 496 x 368.  I assume that because GraphEdt sees the 320 x 240 as a cropping window, it avoids reading and writing off the end of the buffer.</p>
<p>Do you know how to get around this problem?</p>
<p>Thanks,<br />
Jack</p>
]]></content:encoded>
	</item>
</channel>
</rss>
