[M4IF Technotes] CBPC and CBPY

[Bhaskar Sherigar]

Hi,
Adding to this, on the decoder  side
an all-zero block  resulted after ac/dc-pred and
IQ, undergoes mis-match control and makes the all-zero block
to non-zero coeff block.
Regards,
Bhaskar
Chandra Sekhar Reddy G wrote:
> hi,
>
> whether dc-coeff (ofcourse diff-dc) to be coded independently or along with
> ac-coeff is dependent on intra-dc-vlc-thr and quant.
> if ac-prediction is used, then the block, with only top-row-coeff or only
> first-column-coef, *may* become all-zero after ac-prediction.
> for that block, cbp will be 0.
>
> but on the decoder side, an all-zero block (cbp is 0 for that block), may
> get some coeff after ac/dc-pred, which will go through IQ and IDCT.
>
> regards,
> chandra
>
> ----- Original Message -----
> From: "Herbert Thoma" <tma   iis.fhg.de>
> To: "Bart Lehane" <lehaneb   eeng.dcu.ie>
> Cc: <technotes   lists.m4if.org>
> Sent: Wednesday, March 26, 2024 8:45 PM
> Subject: Re: [M4IF Technotes] CBPC and CBPY
>
> > Bart Lehane schrieb:
> > >
> > > Hello,
> > >
> > > In a rectangular I frame, is it possible to have values of cbpy (coded
> > > block pattern for luminance) that do not equal '1111', or values of cbpc
> > > (cbp for chrominance) that do not equal '11'??
> > >
> > > If true, would this give a pattern code not equal to '111111' which
> > > would mean that there would be no DCT coefficients for a particular
> > > block?? (i.e if pattern_code[5] = 0, then the second chrominance block
> > > would have no dct coefficients)
> > >
> > > First of all is this scenario possible, where an intra macroblock
> > > contains a block with no (AC) DCT coefficients? If it is possible, what
> > > values would go into that block?
> >
> > In an I frame the DC coefficient is not regarded when computing cbp.
> > The DC coefficient is allways present and transmitted. If the pattern
> > code for a particular block is 0 then only the DC coefficient is present
> > and all pixels of that block get the same value according to the
> > DC coefficient.
> >
> > Regards,
> >  Herbert.
> >
> > > Thanks in advance,
> > >
> > > Bart
> > >
> > > _______________________________________________
> > > Technotes mailing list
> > > Technotes   lists.m4if.org
> > > http://lists.m4if.org/mailman/listinfo/technotes
> >
> > --
> > Herbert Thoma
> > FhG-IIS A, Studio Department
> > Am Wolfsmantel 33, 91058 Erlangen, Germany
> > Phone: +49-9131-776-323
> > Fax:   +49-9131-776-399
> > email: tma   iis.fhg.de
> > www: http://www.iis.fhg.de/
> > _______________________________________________
> > 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