[MPlayer-users] Two-passes encoding - What happens exactly?

The Wanderer inverseparadox at comcast.net
Tue Jun 1 10:32:23 CEST 2004


Fabio S. wrote:

>>> I read though on this mailing-list that two-passes encoding
>>> improve quality. I don't disagree of course but I would like to
>>> know how it happens. On the man page, the only thing I read is
>>> "vpass activates two pass mode etc ...".
> 
> ...
> 
>>> As I don't fully understand what happens during the two-passes
>>> encoding, I don't know if I have to recopy all the parameters. I
>>> explain: -vf is for filtering thus once filtering is done it
>>> should not be filtered a second time except, I suppose, if I
>>> recrop and rescale. But post-processing? And software scaler
>>> (though I read in man page that software scaler had to be used
>>> with the -zoom option)?
> 
> ...
> 
>>> My question is. Do I have to recopy all the options? Sorry if it
>>> is stupid.
> 
> It is the same for me: it would be nice if somebody would explain
> exactly (in a qualitative way) what happens during two pass encoding,
> how to use it for (as you said: do we need to rewrite all options?)
> and when we can use it (is it possible to use it for capturing from
> the tv?)

Well, I'm not sure I quite understand what you're asking in the first of
those three cases, but I can take a shot at the others.

Yes, you need to use all of the same options on both passes; there are
apparently a few which you can vary slightly without hurting anything,
but for the most part changing the options between passes will mess
things up.

No, you cannot use two-pass mode directly when capturing from TV,
because both passes have to be done on the same source; unless the exact
same signal is going to be sent twice and you can catch it identically
both times, you aren't going to get that with TV capture. However, if
you capture to a different (preferably lossless) format from TV, you can
then use two-pass mode on that.

As for the first question... as best I understand it, what happens is
that during the first pass, MEncoder keeps track of which parts of the
video need more bits than they're getting and which don't need as many,
and writes that information to a file. In the second pass, it uses the
information from that file to determine how many bits to allocate for
each part, to achieve the best overall quality within the target bitrate.

If that isn't what you wanted to know, please try to clarify the
question.

-- 
       The Wanderer never thought *he*'d be answering questions here...

Warning: Simply because I argue an issue does not mean I agree with any
side of it.

A government exists to serve its citizens, not to control them.




More information about the MPlayer-users mailing list