[XviD-devel] [x]Optimize for TV output

[int21h]

I think gruel and -h have stated many times that these 'dirty' areas are
infact caused by an over quant of black, and it seems to be a general
problem with all implementations of MPEG-4 at the moment.

int21h

-----Original Message-----
From: xvid-devel-admin@xvid.org [mailto:xvid-devel-admin@xvid.org] On
Behalf Of Dirk Knop
Sent: Sunday, September 29, 2002 8:48 AM
To: xvid-devel@xvid.org
Subject: Re: [XviD-devel] [x]Optimize for TV output


peter ross wrote:

> firstly, what exactly is a lum offset of -2?
> i assume it simply decrease each luminace pixel by 2. 

I think this is it. It's the code from mpeg2dec.dll (dvd2avi), and it 
sounds reasonable to just subtract 2 from the luma value of a pixel. 
Marc may tell us more about this... Marc?

>> What do you think?
>
>
> i think this type of thing is codec independant, and should be
> written as an avisynth filter, or built into the encoding
> application. 

In fact, MarcFD coded this into his mpeg2dec-pp.dll (-> avisynth 
filter), but I think it would be a nice feature to add to XviD for 
people who don't know how to "tweak everything" - it's easy to add to 
XviD (I assume, I could try to hack that in), and we would have a unique

feature more ;) (cheap but effective codec improvement)

Correction for the initial post: from my tests I have to say that using 
lumi masking together with this is a no-no as it is simply too strong 
for these conditions. (Have to redo my contact-rip because of that... 
some areas get quantized too high now).

Best regards,
Dirk

_______________________________________________
XviD-devel mailing list
XviD-devel@xvid.org
http://list.xvid.org/mailman/listinfo/xvid-devel