[MPlayer-users] Re: Problem Encoding Solid Color Objects With LAVC MPEG4

D Richard Felker III dalias at aerifal.cx
Wed Sep 24 03:07:57 CEST 2003


On Tue, Sep 23, 2003 at 02:44:25PM -0500, Angel wrote:
> [Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
> Rich wrote:
> > I agree. There are no blocking artifacts in either. Perhaps Angel was
> > confusing them with some different sort of artifact -- just plain
> > noise -- but as far as I can tell that has nothing to do with the
> > encode but rather it's just noise in the source. Using the hqdn3d
> > filter might help that.
> 
> As I stated in my last post... the problem is far less noticeable on a 
> computer monitor (which I assume is how you viewed it). The problem I'm 
> talking about is the fact that the noisy background on the wall looks like 
> large blocks moving around in the lavc version. The divx / xvid version does 
> not do this. The source consists of a noisy background wall... that is how it 
> should look in the encoded version (especially with a bitrate of 1500 and 
> little to no motion). The lavc version makes the background look noisy by 
> shifting large solid color blocks around. This is VERY noticeable on a T.V.

OK, maybe I'll dl again and take another look.

> Fabio wrote:
> > I've just seen your samples. Are the blocking artifacts more noticeable in 
> > some places than others? Because, honestly, I cannot spot much difference 
> > from divx to lavc, and I didn't try the xvid one because obviously it is 
> > twice the size so I think a comparison won't be fair.
> > For what I can see, both are good quality. I can't see any blocking artifact 
> > in any of the two.
> 
> Yes the artifacts are worse in some places then others. If you look at the 
> background wall in the lavc version... there are parts of it that look like 
> they consist of large blocks shifting around. As for the xvid version being 
> larger... I used the same bitrate, cropping, and scaling values as the lavc 
> version. On both encodings, I left the min/max quantizer and rate control at 
> the defaults.

Then xvid did not honor your bitrate. Perhaps the rate tolerance
settings are too high for encoding such a short clip... In any case
it's not fair to compare unless both encoders output (approx.) the
same number of bits.

Rich



More information about the MPlayer-users mailing list