<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
<title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
<blockquote>
<blockquote type="cite">
<pre wrap=""><span class="moz-txt-citetags">> </span>I'm trying to implement a simple image grabber, either as a (user)
<span class="moz-txt-citetags">> </span>filter or perhaps differently. I'd like to be able to:
<span class="moz-txt-citetags">></span>
<span class="moz-txt-citetags">> </span>(1) specify a specific time (or frame) in a .nuv file and grab that
<span class="moz-txt-citetags">> </span>frame as a jpeg; I know I don't know enough about the frame process to
<span class="moz-txt-citetags">> </span>know the best way to do it, but I'd like it to be as complete a frame
<span class="moz-txt-citetags">> </span>as possible (without pixelation, incomplete blocks, etc).
<span class="moz-txt-citetags">></span>
<span class="moz-txt-citetags">> </span>The current method of using mplayer and the "-vo jpeg:..." works in
<span class="moz-txt-citetags">> </span>that it'll create a lot of jpeg images for me. However, I really
<span class="moz-txt-citetags">> </span>don't want to create that many images and don't want to go through
<span class="moz-txt-citetags">> </span>them all. When I specify an offset into the show (with "-sb x" or
<span class="moz-txt-citetags">> </span>"-ss x") and limit it with "-frames 10", it starts giving me partial
<span class="moz-txt-citetags">> </span>blocks. I have to use 10 for frames since it does not tend to give me
<span class="moz-txt-citetags">> </span>an image if I specify only 1 frame.
<span class="moz-txt-citetags">></span>
<span class="moz-txt-citetags">> </span>I'm doing this wrong. Even if I'm close to doing it correctly with
<span class="moz-txt-citetags">> </span>this tool, it doesn't seem to be the most efficient. When asking
<span class="moz-txt-citetags">> </span>mplayer to go "deep" into a show, it appears to take a while
<span class="moz-txt-citetags">> </span>processing everything ahead of it. I'm sure that this is possible at
<span class="moz-txt-citetags">> </span>least on a small level, but is there a better way?
<span class="moz-txt-citetags">></span>
<span class="moz-txt-citetags">> </span>(2) Anyway, once this works consistently and accurately, I'd like to
<span class="moz-txt-citetags">> </span>automate it into a script so that I can call it as a user filter. I'd
<span class="moz-txt-citetags">> </span>like to be able to look in the mysql database and find either the
<span class="moz-txt-citetags">> </span>bookmark's location or use editing marks to determine which frame(s)
<span class="moz-txt-citetags">> </span>to grab as images. (Where to store them is secondary at this point
<span class="moz-txt-citetags">> </span>and, I suspect, easy to safely configure.)
<span class="moz-txt-citetags">></span>
<span class="moz-txt-citetags">> </span>I don't really care at this point if the images are useful to current
<span class="moz-txt-citetags">> </span>mythtv programs; though that is perhaps my #3 goal, the first two
<span class="moz-txt-citetags">> </span>would be a 90% solution for my needs.
</pre>
</blockquote>
<br>
This is very similar to what Risto Treksler's patch for the MythTV <br>
preview images does. It creates a PNG thumbnail image by taking into <br>
account the user-specified offset, the commercial cutlist, and--if set <br>
to generate a preview pixmap from the bookmark location--the bookmark <br>
location.<br>
<br>
<a class="moz-txt-link-freetext"
href="http://www.gossamer-threads.com/lists/mythtv/commits/129006#129006">http://www.gossamer-threads.com/lists/mythtv/commits/129006#129006</a><br>
<br>
So, what problem are you attempting to fix here? Are you just trying to
<br>
create a way to force creation of thumbnails or planning to use the <br>
JPEG's for DVD menu creation or ... ?<br>
</blockquote>
<pre wrap=""><!---->In fact, Mike, that is one thing I was hoping to do. (The other is convince my boss -- with some humor -- that he reminds many of us of Magnum P.I. (Tom Selleck), so I'm trying to get as much "incriminating evidence" as possible. Since the presentation of the evidence will likely be in printed form, pictures are the ideal candidate to transfer the proof.)
I'm not (yet) running an svn version of mythtv, so I wasn't immediately aware of that patch going in there (whether it's a part of 0.18.1 or not, though I didn't immediately see it in the docs). The fact that it was conceived with the same premise, functionality and interface tells me I was on the right track, even if I hadn't come up with any code to make it happen.
Thanks for the response.
-bill
</pre>
</body>
</html>