Name Description Size
alloccommon.c our internal buffers are always multiples of 16 5238
alloccommon.h 880
arm
blockd.c 877
blockd.h #define DCPRED 1 8581
coefupdateprobs.h Update probabilities for the nodes in the token entropy tree. Generated file included by entropy.c 8725
common.h Interface header for common constant data structures and lookup tables 1368
context.c GENERATED FILE: DO NOT EDIT *** 17620
debugmodes.c print out the macroblock Y modes 3211
default_coef_probs.h Generated file, included by entropy.c 8518
dequantize.c 972
entropy.c vp8_default_zig_zag_mask generated with: void vp8_init_scan_order_mask() { int i; for (i = 0; i < 16; ++i) { vp8_default_zig_zag_mask[vp8_default_zig_zag1d[i]] = 1 << i; } } 5575
entropy.h Coefficient token alphabet 3990
entropymode.c Array indices are identical to previously-existing INTRAMODECONTEXTNODES. 3666
entropymode.h # of subsets 2974
entropymv.c clang-format off 1443
entropymv.h max absolute value of a MV component 1573
extend.c source 5550
extend.h 1097
filter.c note that 1/8 pel positions are just as per alpha -0.5 bicubic 14534
filter.h 852
findnearmv.c Predict motion vectors using those from already-decoded nearby blocks. Note that we only consider one 4x4 subblock from each candidate 16x16 macroblock. 5484
findnearmv.h On L edge, get from MB to left of us 5198
generic
header.h 24 bits total 1120
idct_blk.c 1902
idctllm.c Notes: This implementation makes use of 16 bit fixed point verio of two multiply constants: 1. sqrt(2) * cos (pi/8) 2. sqrt(2) * sin (pi/8) Becuase the first constant is bigger than 1, to maintain the same 16 bit fixed point precision as the second one, we use a trick of x * a = x + x*(a-1) so x * sqrt(2) * cos (pi/8) = x + x * (sqrt(2) *cos(pi/8)-1). ************************************************************************ 4257
invtrans.h eob adjust.... the idct can only skip if both the dc and eob are zero 1590
loongarch
loopfilter.h fraction of total macroblock rows to be used in fast filter level picking 3610
loopfilter_filters.c should we apply any filter at all ( 11111111 yes, 00000000 no) 14200
mbpitch.c y blocks 1535
mfqe.c MFQE: Multiframe Quality Enhancement In rate limited situations keyframes may cause significant visual artifacts commonly referred to as "popping." This file implements a postproccesing algorithm which blends data from the preceeding frame when there is no motion and the q from the previous frame is lower which indicates that it is higher quality. 11946
mips
modecont.c 0 686
modecont.h 648
mv.h facilitates faster equality tests and copies 782
onyx.h Create/destroy static data structures. 9037
onyxc_int.h #ifdef PACKET_TESTING 4457
onyxd.h Create/destroy static data structures. 1838
postproc.c clang-format off 9776
postproc.h 1259
ppflags.h 948
quant_common.c For all x in [0..284], x*155/100 is bitwise equal to (x*101581) >> 16. The smallest precision for that is '(x*6349) >> 12' but 16 is a good word size. 3411
quant_common.h 966
reconinter.c encoder only 17473
reconinter.h 1353
reconintra.c Power PC implementation uses "vec_vsx_ld" to read 16 bytes from uleft_col and vleft_col. Play it safe by reserving enough stack space here. 3382
reconintra.h 1174
reconintra4x4.c Power PC implementation uses "vec_vsx_ld" to read 16 bytes from Above (aka, Aboveb + 4). Play it safe by reserving enough stack space here. Similary for "Left". 2508
reconintra4x4.h 1580
rtcd.c 560
rtcd_defs.pl 11307
setupintrarecon.c set up frame new frame for intra coded blocks 1400
setupintrarecon.h 1321
swapyv12buffer.c 1031
swapyv12buffer.h 775
systemdependent.h 724
threading.h Win32 2910
treecoder.c n = size of alphabet 2796
treecoder.h probability index 3033
vp8_entropymodedata.h Generated file, included by entropymode.c 7404
vp8_loopfilter.c For each possible value for the loop filter fill out limits 18940
vp8_skin_detection.c 3770
vp8_skin_detection.h 1435
x86