[MPlayer-users] About CVCD

[Steffen Barszus]

On Thursday 05 December 2002 13:49, Jukka Tastula wrote:
> [Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
>
> On Thursday 05 December 2002 14:30, Steffen Barszus wrote:
> > - at low bitrates (sometimes blocky)
>
> Its always blocky unless you use yuvscaler to make it all blurry :)

Yes, but there is a new yuvscaler compatible programm ym4scaler, announced 
only a few days ago, have to try it I guess: 
from the anounce:
"...
    o To decouple the scaling engine itself from the front-end.
    o To make the front-end smart enough to automatically handle those weird
       details like padding 704 pixel wide sources to 720 pixels.
    o Pedantic accuracy.
..."
It is slower , but that isn't of any interest, since mpeg2enc is slow as hell 
too ;)

>
> > - sharp edges ( ringing affect (a lot ! at least in my experimrents))
>
> Yup, sharpness might look good in the master but it bites you when you try
> compress it...
>

A nice (loong) document about all that ;) :
http://tns-www.lcs.mit.edu/manuals/mpeg2/FAQ


> > - size prediction ( you can only give the quantesizer , no average
> > bitrate or something like that, maybe it can be steered by lowering the
> > max-bitrate and hopefully does not produce crap at highbitrate-scenes)
>
> My experiments tell me setting the quant at 2 gives you pretty much
> constant bitrate (all my encodings vere ~40 minutes so I'm hitting 2500kbps
> all the time anyway) but for this kind of job you'd really want two passes
> to get it right at first go. 

I've asked if a 2-pass and or average bitrate could be integrated. Got only 
arguments against that back: 
"Not needed, quantesizers ar enough"
"complicate to write it and brings nearly no quality"
trough using tosvcd to convert my vdr-recordings, I usually have 6-9 on 
movies, it tries to compress and if the filesize gets to big it increase the 
q-value ( 60 mins max and hoping there aren't so many fast motions or lot 
details and hoping the TV-Set is bad ;), don't want a normal movie on 3 CDs ) 

>
> > The next problem I see is that the max-time of svcd is AFAIK 79:30 ( near
> > 80:00 min at least) so I dont see how there can be 90 min on that cd.
>
> Eh? And where does this limit come from? You must be thinking vcd which has
> a combined bitrate of exactly the same as audio cd, so 1 minute of data = 1
> minute of video. SVCD doesn't have this restriction: you can have _up to_,
> not only 2500kbps variable bitrate video and audio _up to_, not only
> 384kbps (any encoder actually exist that can do that? toolame only goes to
> 320kbps) so you can certainly have more than 80 minutes on one cd if the
> player can play such low bitrate video+audio. It'll be annoyingly blocky at
> standard svcd resolution though :)  It all really depends on the player in
> the end.

Ok , yeah it was vcd.Just had this restriction in mind.



> _______________________________________________
> RTFM!!!  http://www.MPlayerHQ.hu/DOCS
> Search:  http://www.MPlayerHQ.hu/cgi-bin/htsearch
> http://mplayerhq.hu/mailman/listinfo/mplayer-users

-- 
____________________
counter.li.org : #296567.
machine: 181800
vdr-box : 87
____________________
Please dont CC me, since if I have replied I'll watch the tread. Both mails 
will be filtered to the ML-folder. Thanks