[M4IF Technotes] question about intra_dc_vlc_thr2

tetsuo kosuge kosuge td.idc.lsi.sanyo.co.jp
Wed Mar 5 17:06:23 EST 2003


Hi Mr.Aggarwal.
That's my previous question,so I have no confidence.
But I understand as follows(Mr.Gautam's answer is helpful me to understand) :
-----------------------------------------------------------------------------
If(Intra AC VLC is turned on) {
	if(pattern_code == 1) {
		this case coeff without distinction of ac or dc.
		MB has 64 coeff. And they are derived from Tbl.B-16 or Tbl.B-17.
		<first(dc) coeff && last = 1  case exist>
	}
	else { //(pattern_code == 0) 
		64 coeff are all 0.
	}
}
else { //Intra AC VLC is turned off
	if(pattern_code == 1) {
		MB has 64 coeff(1 dc_coeff and 63 ac_coeff).
		dc_coeff is derived from Tbl.B13-Tbl.B15.
		ac_coeff are derived from Tbl.B-16 or Tbl.B-17.	
	}
	else { //(pattern_code == 0) 
		dc_coeff is derived from Tbl.B13-Tbl.B15.
		ac_coeff(63 coeff) are all 0.
	}
}
---------------------------------------------------------------------------------
Regards,
 Tetsuo
>For an Intra block with Intra AC VLC turned on, suppose there was
>only 1 DCT coefficient, then is it correct to say that the decode
>of the first VLC DCT coefficient will give out a LAST==1? 
>
>Or lets say that Intra AC VLC is turned on and cbpy[] for an intra
>block is 0. Then it means that there is no DCT coefficient at all
>for this block. Neither DC nor AC.
>
>And to understand this completely, if the Intra AC VLC is turned 
>off for the block and there is only 1 DCT coefficient, it means
>that only DC coefficient is present. Hence, the cbpy[] for this
>block will be '0' since there are no AC coefficients. Correct?
>
>Gaurav



More information about the Mp4-tech mailing list