[vorbis] Video codec

[tom barger]

In film-post-production, years of procedure behind this odd problem of 30 fr
versus 29.97, but the average human eye can instanlty detect a 3 frame
out-of-sync lip flap. And most of our sound editors can spot TWO FRAMES a
mile away. The human eye much muore instantaneous than ear. We never skimp on
this, we will be busted flat and fired from our jobs if trying to do sloppy
sync work.

However, I just watched QuickTime channel broadcast of BBC, and on the
interenet, the idea is that the audience will acept any incredible terrible
quality of what looks to me like 3 frames a second streaming. Just rememeber
that audio proceeds in realtime always, there is no frame-rate per se, and we
have output buttons in our editing programs that say (:PLAY EVERY
FRAME")---meaning,if this button is checked, there is no audio. Audio must
play in realtime, and video frames are dropped to accommodate. Maybe you
could say that audio is the master track and video the slave. (That's a funny
old concept from my years of timecode thinking, I'll have to work on that and
recondition my thinking about ideas of linearity!!) But you see, we are
fighting dinosaurs here in Hollywood broadcast, they have every reason to
prtect old-school means of distribution, anmd in fact are completely
out-of-touch with the idea that a new generation whas been raised on video
games, and insists on user control. Zap the commercials, indeed.

I will chime in with a wish list, and hope you all see my ideas from being an
editor and user-based workhorse. I'll start with audio, this is where the
growth is exploding on internet.

Bob Miller wrote:

> Chrissy and Raul wrote:
>
> > >Likewise, we're not worrying that 30 fps isn't really 30 fps in NTSC
> > >video?
> >
> > The difference is small enough to not be a problem.
>
> It would be a problem if your video is sync'd with audio.  After only
> sixty seconds, a 29.97 Hz video played at 30 Hz will be 60 msec ahead
> of the sound track.  That's enough to make a talking person's lips look
> unsynched, I think.  (A 90 minute movie would be 5 seconds out of
> sync.)
>
> May I sugest using the DDA algorithm (used to draw diagonal lines in
> 2D graphics) to calculate frame repeats (or skips) anytime the source
> frame rate doesn't exactly match the output device's rate?  It's
> extremely fast, and it's exactly accurate if the ratio of frame rates
> is a rational number.
> --
>                                         K<bob>
> kbob at jogger-egg.com, http://www.jogger-egg.com/
>
> --- >8 ----
> List archives:  http://www.xiph.org/archives/
> Ogg project homepage: http://www.xiph.org/ogg/
> To unsubscribe from this list, send a message to 'vorbis-request at xiph.org'
> containing only the word 'unsubscribe' in the body.  No subject is needed.
> Unsubscribe messages sent to the list will be ignored/filtered.

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.