On 19 Dec 2003, Florin Andrei wrote:

> You mean that for whole movies (> 1 hour) or for smaller scenes?

        For the movie as a whole.   Shorter scenes will, as you have seen,
        run up against the limit and there's nothing "wrong" about that.

        The guidline often is given as 20% but I've found 10% to be workable
        too.    

> The results? If i set -q so that the average is 10% lower than peaks for
> the whole movie thing, there will be scenes (not just a second or two,
> but entire scenes) where the bitrate will be pressed hard into the

        Nothing wrong about that - as long as it's short term relative to
        the entire movie.

> I am not exactly sure what's going on with those "bitrate crushed into
> the ceiling" scenes, but there's something weird about them. All bugs

        Lots of motion would be one reasonable cause.   Professional encoding
        folks will apply a selective blur to those types of scenes - not to
        the entire movie.   The human eye would just be seeing a blur of
        motion anyhow so a slight soften/blur by the encoder would lower the
        bitrate to keep the peak rate in bounds.

        Perhaps when Andrew gets the 2 pass encoding going there will be
        hooks for selective filtering of regions where the encoder was running
        up against the rate limit.

> everywhere in the video chain (recording, processing, storing, playing)
> tend to be much easier revealed by them.
> E.g., mplex reports the average being higher than the peak. :-)

        Statistical anomaly - I've seen it too.   

        What I'd really like to have is a bitstream analyzer that could
        take a stream and produce plots/reports about the bitrate (and other
        info).   

        Cheers,
        Steven Schultz



-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click
_______________________________________________
Mjpeg-users mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/mjpeg-users

Reply via email to