[MPlayer-users] lossless codecs make bigger files than raw DV ?!

[Andrzej M. Ostruszka]

First of all thank you for all response.

Now, going back to the beginning:

> > audio -> encoded with mp3lame with abr and br=64 (64 it's enough for me
> >        - the sound recorded by the camcoder is not of top quality)
> 
> This is broken logic. For low-quality input, you need _higher_ bitrate
> or else you'll make it much worse. I would recommend using normal vbr
> (not abr) with q=8.

Bad wording on my side.  I meant to say that the type of sound which
I record -- I'm not going with the camcoder to the philharmony :)) --
does not need large bitrate to be faithfull (at least according to my
"auditory system").

> > video -> lavc with vcodec=mpeg4:vhq:vbitrate=2000:vpass=1/2 -vop pp=lb
> 
> At the very least, replace the outdated vhq with mbd=2 and use trell.

I see that I need to update my linux distribution in a more swift
fashion.  I was waiting with mplayer upgrade until then (still using
0.90rc5).

As to the debate whether to change default or not I opt for leaving them
at the current values but to enhance the documentation.  There's not
much in mplayer man page that would suggest which options to use if you
are interested in quality and which if only in speed, just a list of
them with short description (probably sufficient for those dealing with
video signal processing on a daily basis).  Since there's a lot of them
I'm not able to understand every one thoroughly and it would be
beneficial for me to have some sort of deeper explanation of what given
option stands for.  For example I don't know what lumi_mask does and
I refuse to test every posible value of it to figure out what efect it
might have.  I'm rather experienced Linux user with over 10y bagage and
I always read the manuals (which I qualify as a basic responsibility of
a user).

The other thing that Richard mentioned -- deinterlacing.  I thought that
deinterlacing does not throw away every second frame but merge them with
some sort of smoothing so the resulting picture is better suited for
encoding.  But now I'm told something quite different.  (I used
deinterlacing cause it is suggested on www.bunkus.org/... dvdripping
site).  Could somebody be more elaborate how this really works -- my
main interest is quality of the encoded video (no matter what the cpu
load it causes -- well to some reasonable values e.g. not longer than
one night for a DVD).
					Best regards
					Andrzej Ostruszka