[M4IF Technotes] MVD=16 in MPEG4

[Christoph Lampert]

On Fri, 26 Jul 2002, Luca Piccarreta wrote:
> I just gave a look at the standard and discovered that MVD=16 exists,
> but must not be used when vop_fcode=1.
> MVD=16 is needed when vop_fcode>1

You were not discussing B-frame direct mode delta vectors, were you? 
Because those are limited to fcode=1. 
Regular motion vector differences can of course have a value of 16 or much
more... 
Christoph Lampert 
> Luca.
> ----- Original Message -----
> From: "Chen, Kenny" <kenny.chen   intel.com>
> To: <technotes   lists.m4if.org>
> Sent: Friday, July 26, 2023 2:45 AM
> Subject: [M4IF Technotes] MVD=16 in MPEG4
> 
> 
> > Hi,
> >
> > In H.263, the table for MV difference ranges from -16 to 15.5, in MPEG4
> > MVD=16 is added.
> > But in VM18 there is :
> > if (diff_vector < -range)
> > diff_vector = diff_vector + 2*range;
> > else if (diff_vector > range-1)
> > diff_vector = diff_vector - 2*range;
> >
> > range = 32;
> >
> > from this equation, it seems we can not get MVD=16 from it, so why do
> MPEG4
> > add MVD=16 in the vlc table? Is it useless?
> >
> > thanks
> > Ken
> > _______________________________________________
> > Technotes mailing list
> > Technotes   lists.m4if.org
> > http://lists.m4if.org/mailman/listinfo/technotes
> 
> _______________________________________________
> Technotes mailing list
> Technotes   lists.m4if.org
> http://lists.m4if.org/mailman/listinfo/technotes
> 

-- 
Christoph H. Lampert chl   math.uni-bonn.de | Diese Signature wurde maschi-     
Beringstr. 6, Raum 14 Tel. (0228) 73-2948 | nell erstellt und bedarf
Sprechstunden: keine, aber meistens da    | keiner Unterschrift. AZ 27B-6