AVC in AVI, MP4, MKV and MPG PS filesize comparison

[bond]

as i made a lot of sample files for my decoder comparison (http://forum.doom9.org/showthread.php?t=99402) i dont want to hide the filesizes from you, maybe useable as an "indirect overhead comparison":

the sum of 23 files (~7000 frames and 700kbps) is:
mp4 mkv mpg avi
raw: 616636.69 618279.47 618935.84 620428.00 620564.30
+ 1642.77 + 2299.15 + 3791.31 + 3927.61
+ 0.27% + 0.37% + 0.61% + 0.64%mp4 has been created with x264
avi created from raw with avc2avi
mpg created from mp4 with ffmpeg
mkv created from mp4 with mkvmerge

raw results:
25.08.2005 21:59 27.762.759 nero_2pass-777kbps_Cabac_Deblock-5-adapt_B2_Ref3_noWPred_Qpel_p8x8_cartoon_psy2_extra.mp4
29.08.2005 21:13 27.686.898 nero_2pass-777kbps_Cabac_Deblock-5-adapt_B2_Ref3_noWPred_Qpel_p8x8_cartoon_psy2_extra_track3.h264
15.10.2005 16:34 27.861.470 nero_2pass-777kbps_Cabac_Deblock-5-adapt_B2_Ref3_noWPred_Qpel_p8x8_cartoon_psy2_extra_track3.h264.avi
29.08.2005 21:14 27.844.608 nero_2pass-777kbps_Cabac_Deblock-5-adapt_B2_Ref3_noWPred_Qpel_p8x8_cartoon_psy2_extra_track3.h264.mpg
19.02.2005 22:30 27.761.693 nero_2pass-777kbps_Cabac_Deblock-5-adapt_B2_Ref3_WPred+wbp_Qpel_p8x8_cartoon_psy2_extra.mp4
29.08.2005 21:13 27.686.110 nero_2pass-777kbps_Cabac_Deblock-5-adapt_B2_Ref3_WPred+wbp_Qpel_p8x8_cartoon_psy2_extra_track3.h264
15.10.2005 16:37 27.860.584 nero_2pass-777kbps_Cabac_Deblock-5-adapt_B2_Ref3_WPred+wbp_Qpel_p8x8_cartoon_psy2_extra_track3.h264.avi
29.08.2005 21:14 27.844.608 nero_2pass-777kbps_Cabac_Deblock-5-adapt_B2_Ref3_WPred+wbp_Qpel_p8x8_cartoon_psy2_extra_track3.h264.mpg
23.08.2005 23:06 27.820.132 x264-r103_2pass-constqp21-780kbps_Cabac_Deblock-5-5_noB_Ref3_Subq5_p8x8_mp4box.mp4
29.08.2005 17:29 27.802.602 x264-r103_2pass-constqp21-780kbps_Cabac_Deblock-5-5_noB_Ref3_Subq5_p8x8_mp4box_track1.h264
29.08.2005 19:08 27.977.142 x264-r103_2pass-constqp21-780kbps_Cabac_Deblock-5-5_noB_Ref3_Subq5_p8x8_mp4box_track1.h264.avi
29.08.2005 17:59 27.981.824 x264-r103_2pass-constqp21-780kbps_Cabac_Deblock-5-5_noB_Ref3_Subq5_p8x8_mp4box_track1.h264.mpg
23.08.2005 23:05 27.737.967 x264-r139_2pass-777kbps_Cabac_Deblock-5-5_B2-temporal-adapt_Ref5_Subq5_p4x4-b8x8_mp4box.mp4
29.08.2005 17:29 27.677.915 x264-r139_2pass-777kbps_Cabac_Deblock-5-5_B2-temporal-adapt_Ref5_Subq5_p4x4-b8x8_mp4box_track1.h264
29.08.2005 19:10 27.852.506 x264-r139_2pass-777kbps_Cabac_Deblock-5-5_B2-temporal-adapt_Ref5_Subq5_p4x4-b8x8_mp4box_track1.h264.avi
29.08.2005 17:59 27.848.704 x264-r139_2pass-777kbps_Cabac_Deblock-5-5_B2-temporal-adapt_Ref5_Subq5_p4x4-b8x8_mp4box_track1.h264.mpg
30.08.2005 13:59 27.580.230 x264-r285_bp_720x288_B0_Ref5_p8x8-i4x4_loop-5_wbp_cabac.avi
30.08.2005 14:40 27.580.416 x264-r285_bp_720x288_B0_Ref5_p8x8-i4x4_loop-5_wbp_cabac.avi.mpg
01.09.2005 21:08 27.398.966 x264-r285_bp_720x288_B0_Ref5_p8x8-i4x4_loop-5_wbp_cabac_mp4box_track1.h264
30.08.2005 14:39 27.416.648 x264-r285_bp_720x288_B0_Ref5_p8x8-i4x4_loop-5_wbp_mp4box.mp4
06.06.2005 16:03 27.484.877 x264_hp_2pass_640x256_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac.mp4
29.08.2005 17:27 27.399.499 x264_hp_2pass_640x256_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac_track1.h264
29.08.2005 18:47 27.574.122 x264_hp_2pass_640x256_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac_track1.h264.avi
29.08.2005 17:58 27.568.128 x264_hp_2pass_640x256_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac_track1.h264.mpg
24.08.2005 02:29 27.432.844 x264_hp_2pass_720x288_B0_Ref5_p4x4-i8x8_loop-5_WBP_cabac.mp4
29.08.2005 17:27 27.400.450 x264_hp_2pass_720x288_B0_Ref5_p4x4-i8x8_loop-5_WBP_cabac_track1.h264
29.08.2005 18:25 27.575.038 x264_hp_2pass_720x288_B0_Ref5_p4x4-i8x8_loop-5_WBP_cabac_track1.h264.avi
29.08.2005 17:58 27.580.416 x264_hp_2pass_720x288_B0_Ref5_p4x4-i8x8_loop-5_WBP_cabac_track1.h264.mpg
25.08.2005 15:07 27.483.562 x264_hp_2pass_720x288_B1_Ref5_p4x4-i8x8_loop_WBP_cabac.mp4
29.08.2005 17:28 27.398.416 x264_hp_2pass_720x288_B1_Ref5_p4x4-i8x8_loop_WBP_cabac_track1.h264
29.08.2005 18:26 27.572.966 x264_hp_2pass_720x288_B1_Ref5_p4x4-i8x8_loop_WBP_cabac_track1.h264.avi
29.08.2005 17:58 27.564.032 x264_hp_2pass_720x288_B1_Ref5_p4x4-i8x8_loop_WBP_cabac_track1.h264.mpg
23.08.2005 22:41 27.485.367 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP.mp4
01.08.2005 13:07 27.485.476 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac.mp4
29.08.2005 17:04 27.490.421 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac_cqm-qmatrix.mp4
29.08.2005 17:29 27.405.259 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac_cqm-qmatrix_track1.h264
29.08.2005 18:28 27.579.816 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac_cqm-qmatrix_track1.h264.avi
29.08.2005 17:58 27.574.272 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac_cqm-qmatrix_track1.h264.mpg
29.08.2005 17:28 27.400.134 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac_track1.h264
29.08.2005 18:30 27.574.690 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac_track1.h264.avi
29.08.2005 17:58 27.568.128 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac_track1.h264.mpg
29.08.2005 17:28 27.400.169 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_track1.h264
29.08.2005 18:32 27.574.740 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_track1.h264.avi
29.08.2005 17:58 27.568.128 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_track1.h264.mpg
25.08.2005 18:20 27.486.671 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop6_WBP_cabac.mp4
29.08.2005 17:28 27.401.425 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop6_WBP_cabac_track1.h264
29.08.2005 18:34 27.575.936 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop6_WBP_cabac_track1.h264.avi
29.08.2005 17:58 27.572.224 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop6_WBP_cabac_track1.h264.mpg
23.08.2005 19:23 27.486.106 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_WBP_cabac.mp4
29.08.2005 17:28 27.400.916 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_WBP_cabac_track1.h264
29.08.2005 18:36 27.575.538 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_WBP_cabac_track1.h264.avi
29.08.2005 17:58 27.570.176 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_WBP_cabac_track1.h264.mpg
23.08.2005 23:53 27.471.309 x264_hp_2pass_720x288_B3_Ref5_p4x4-i8x8_loop-5_WBP_cabac.mp4
29.08.2005 17:28 27.400.471 x264_hp_2pass_720x288_B3_Ref5_p4x4-i8x8_loop-5_WBP_cabac_track1.h264
29.08.2005 18:26 27.575.044 x264_hp_2pass_720x288_B3_Ref5_p4x4-i8x8_loop-5_WBP_cabac_track1.h264.avi
29.08.2005 17:58 27.568.128 x264_hp_2pass_720x288_B3_Ref5_p4x4-i8x8_loop-5_WBP_cabac_track1.h264.mpg
05.06.2005 18:37 27.485.673 x264_mp_2pass_640x256_B3-Ref_Ref5_p4x4-i4x4_loop-5_WBP_cabac.mp4
29.08.2005 17:28 27.400.279 x264_mp_2pass_640x256_B3-Ref_Ref5_p4x4-i4x4_loop-5_WBP_cabac_track1.h264
29.08.2005 18:38 27.574.876 x264_mp_2pass_640x256_B3-Ref_Ref5_p4x4-i4x4_loop-5_WBP_cabac_track1.h264.avi
29.08.2005 17:58 27.568.128 x264_mp_2pass_640x256_B3-Ref_Ref5_p4x4-i4x4_loop-5_WBP_cabac_track1.h264.mpg
26.08.2005 00:56 27.479.307 x264_mp_2pass_720x288_B2-Ref_Ref3_p4x4-i4x4_cabac.mp4
29.08.2005 17:28 27.394.149 x264_mp_2pass_720x288_B2-Ref_Ref3_p4x4-i4x4_cabac_track1.h264
29.08.2005 18:56 27.568.730 x264_mp_2pass_720x288_B2-Ref_Ref3_p4x4-i4x4_cabac_track1.h264.avi
29.08.2005 17:59 27.564.032 x264_mp_2pass_720x288_B2-Ref_Ref3_p4x4-i4x4_cabac_track1.h264.mpg
23.08.2005 18:01 27.471.146 x264_mp_2pass_720x288_B2_Ref3_p4x4-i4x4_cabac.mp4
29.08.2005 17:28 27.396.436 x264_mp_2pass_720x288_B2_Ref3_p4x4-i4x4_cabac_track1.h264
29.08.2005 18:49 27.571.082 x264_mp_2pass_720x288_B2_Ref3_p4x4-i4x4_cabac_track1.h264.avi
29.08.2005 17:58 27.561.984 x264_mp_2pass_720x288_B2_Ref3_p4x4-i4x4_cabac_track1.h264.mpg
23.08.2005 13:45 27.471.882 x264_mp_2pass_720x288_B2_Ref3_p8x8-i4x4_cabac.mp4
29.08.2005 17:28 27.397.108 x264_mp_2pass_720x288_B2_Ref3_p8x8-i4x4_cabac_track1.h264
29.08.2005 18:51 27.571.678 x264_mp_2pass_720x288_B2_Ref3_p8x8-i4x4_cabac_track1.h264.avi
29.08.2005 17:59 27.561.984 x264_mp_2pass_720x288_B2_Ref3_p8x8-i4x4_cabac_track1.h264.mpg
23.08.2005 15:04 27.471.139 x264_mp_2pass_720x288_B2_Ref3_p8x8-i4x4_loop-5_cabac.mp4
29.08.2005 17:28 27.396.365 x264_mp_2pass_720x288_B2_Ref3_p8x8-i4x4_loop-5_cabac_track1.h264
29.08.2005 18:54 27.570.960 x264_mp_2pass_720x288_B2_Ref3_p8x8-i4x4_loop-5_cabac_track1.h264.avi
29.08.2005 17:59 27.561.984 x264_mp_2pass_720x288_B2_Ref3_p8x8-i4x4_loop-5_cabac_track1.h264.mpg
25.08.2005 17:05 27.482.851 x264_mp_2pass_720x288_B3-Ref_Ref1_p4x4-i4x4_loop_WBP_cabac.mp4
29.08.2005 17:28 27.397.685 x264_mp_2pass_720x288_B3-Ref_Ref1_p4x4-i4x4_loop_WBP_cabac_track1.h264
29.08.2005 18:58 27.572.308 x264_mp_2pass_720x288_B3-Ref_Ref1_p4x4-i4x4_loop_WBP_cabac_track1.h264.avi
29.08.2005 17:59 27.568.128 x264_mp_2pass_720x288_B3-Ref_Ref1_p4x4-i4x4_loop_WBP_cabac_track1.h264.mpg
23.08.2005 01:16 27.483.294 x264_mp_2pass_720x288_B3-Ref_Ref3_p4x4-i4x4_loop-5_cabac.mp4
29.08.2005 17:29 27.398.084 x264_mp_2pass_720x288_B3-Ref_Ref3_p4x4-i4x4_loop-5_cabac_track1.h264
29.08.2005 19:03 27.572.684 x264_mp_2pass_720x288_B3-Ref_Ref3_p4x4-i4x4_loop-5_cabac_track1.h264.avi
29.08.2005 17:59 27.566.080 x264_mp_2pass_720x288_B3-Ref_Ref3_p4x4-i4x4_loop-5_cabac_track1.h264.mpg
25.08.2005 16:08 27.483.417 x264_mp_2pass_720x288_B3-Ref_Ref3_p4x4-i4x4_loop_WBP_cabac.mp4
29.08.2005 17:29 27.398.215 x264_mp_2pass_720x288_B3-Ref_Ref3_p4x4-i4x4_loop_WBP_cabac_track1.h264
29.08.2005 19:01 27.572.746 x264_mp_2pass_720x288_B3-Ref_Ref3_p4x4-i4x4_loop_WBP_cabac_track1.h264.avi
29.08.2005 17:59 27.566.080 x264_mp_2pass_720x288_B3-Ref_Ref3_p4x4-i4x4_loop_WBP_cabac_track1.h264.mpg
23.08.2005 00:11 27.483.636 x264_mp_2pass_720x288_B3-Ref_Ref5_p4x4-i4x4_loop-5_WBP_cabac.mp4
29.08.2005 17:29 27.398.422 x264_mp_2pass_720x288_B3-Ref_Ref5_p4x4-i4x4_loop-5_WBP_cabac_track1.h264
29.08.2005 19:06 27.572.958 x264_mp_2pass_720x288_B3-Ref_Ref5_p4x4-i4x4_loop-5_WBP_cabac_track1.h264.avi
29.08.2005 17:59 27.566.080 x264_mp_2pass_720x288_B3-Ref_Ref5_p4x4-i4x4_loop-5_WBP_cabac_track1.h264.mpg
16.10.2005 21:13 27.791.427 nero_2pass-777kbps_Cabac_Deblock-5-adapt_B2_Ref3_noWPred_Qpel_p8x8_cartoon_psy2_extra.mp4.mkv
16.10.2005 21:14 27.790.637 nero_2pass-777kbps_Cabac_Deblock-5-adapt_B2_Ref3_WPred+wbp_Qpel_p8x8_cartoon_psy2_extra.mp4.mkv
16.10.2005 21:12 27.891.004 x264-r103_2pass-constqp21-780kbps_Cabac_Deblock-5-5_noB_Ref3_Subq5_p8x8_mp4box.mp4.mkv
16.10.2005 21:12 27.768.452 x264-r139_2pass-777kbps_Cabac_Deblock-5-5_B2-temporal-adapt_Ref5_Subq5_p4x4-b8x8_mp4box.mp4.mkv
16.10.2005 21:12 27.487.960 x264-r285_bp_720x288_B0_Ref5_p8x8-i4x4_loop-5_wbp_mp4box.mp4.mkv
16.10.2005 21:10 27.503.876 x264_hp_2pass_640x256_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac.mp4.mkv
16.10.2005 21:10 27.503.697 x264_hp_2pass_720x288_B0_Ref5_p4x4-i8x8_loop-5_WBP_cabac.mp4.mkv
16.10.2005 21:11 27.501.666 x264_hp_2pass_720x288_B1_Ref5_p4x4-i8x8_loop_WBP_cabac.mp4.mkv
16.10.2005 21:11 27.504.559 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP.mp4.mkv
16.10.2005 21:11 27.504.528 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac.mp4.mkv
16.10.2005 21:11 27.509.616 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop-5_WBP_cabac_cqm-qmatrix.mp4.mkv
16.10.2005 21:11 27.505.795 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_loop6_WBP_cabac.mp4.mkv
16.10.2005 21:11 27.505.306 x264_hp_2pass_720x288_B3-Ref_Ref5_p4x4-i8x8_WBP_cabac.mp4.mkv
16.10.2005 21:11 27.504.804 x264_hp_2pass_720x288_B3_Ref5_p4x4-i8x8_loop-5_WBP_cabac.mp4.mkv
16.10.2005 21:11 27.504.670 x264_mp_2pass_640x256_B3-Ref_Ref5_p4x4-i4x4_loop-5_WBP_cabac.mp4.mkv
16.10.2005 21:11 27.498.770 x264_mp_2pass_720x288_B2-Ref_Ref3_p4x4-i4x4_cabac.mp4.mkv
16.10.2005 21:11 27.500.968 x264_mp_2pass_720x288_B2_Ref3_p4x4-i4x4_cabac.mp4.mkv
16.10.2005 21:11 27.501.665 x264_mp_2pass_720x288_B2_Ref3_p8x8-i4x4_cabac.mp4.mkv
16.10.2005 21:11 27.500.922 x264_mp_2pass_720x288_B2_Ref3_p8x8-i4x4_loop-5_cabac.mp4.mkv
16.10.2005 21:12 27.502.084 x264_mp_2pass_720x288_B3-Ref_Ref1_p4x4-i4x4_loop_WBP_cabac.mp4.mkv
16.10.2005 21:12 27.502.468 x264_mp_2pass_720x288_B3-Ref_Ref3_p4x4-i4x4_loop-5_cabac.mp4.mkv
16.10.2005 21:12 27.502.600 x264_mp_2pass_720x288_B3-Ref_Ref3_p4x4-i4x4_loop_WBP_cabac.mp4.mkv
16.10.2005 21:12 27.502.826 x264_mp_2pass_720x288_B3-Ref_Ref5_p4x4-i4x4_loop-5_WBP_cabac.mp4.mkv

[Dmitry Vergheles]

And what about AVC in ASF format?

[bond]

hm actually i dont really care about asf :(

[Dmitry Vergheles]

I see, but I suppose that for format enumerating to be more full, ASF tests could be fairly usefull.

[multicone]

Do MP4 created with x264 contain something similar to the MKV seekhead for fast seeking ? What about a seeking speed comparison for all of them :D ??

[iapir]

What are the 0.30, 0.47, 0.84 sizes ? Can you cut bytes in pieces ?

[stephanV]

overhead comparisons are useless, because a. its a rather small value and b. you can do a lot of manipulation by choosing the "right" tools.

[Zenitram]

overhead comparisons are useless, because a. its a rather small value and b. you can do a lot of manipulation by choosing the "right" tools.
a. Maybe for you, not for me : I prefer having a the overhead kept for video quality rather than for stupid container things. 0.5 to 1% more for video can give better visual quality.
b. Let's give us the "right" tools, I am interested!

@bond : It would more interesting to compare containers with one or two audio streams, and few text (subtitles) streams, to compare interlacing capacities of the container in a real situation.

[foxyshadis]

What are the 0.30, 0.47, 0.84 sizes ? Can you cut bytes in pieces ?
It's measured in kilobytes, not bytes. 23 files of 27 megs per set. :D

And yes, I'd be interested in a full matchup of containers with many different streams, but I could just as easily do that myself since I have all the tools to.

[stephanV]

a. Maybe for you, not for me : I prefer having a the overhead kept for video quality rather than for stupid container things. 0.5 to 1% more for video can give better visual quality.
Can it? Show me samples using 704 kbps that look remarkably better than ones encoded with 700 kbps.


b. Let's give us the "right" tools, I am interested!

raw h264 size: 10,784 kb
x264 mkvoutput: 10,829 kb (0.42%)
mtx remux: 10,833 kb (0.45%)
haali ds mux: 10,825 kb (0.38%)

other example: the difference between a standard AVI and an openDML avi is 8 byte per frame, and with the new block type and compression method introduced in Matroska you can let sizes vary even more.

[bond]

interested in overhead with more streams? make a comparison yourself and post the results, i would be interested too :D

yeah some tools write smaller files than others for the same container. eg x264's mp4 is bigger than mp4box' output in most cases, because mp4box can write a smaller header when muxing, what you cant do when encoding directly into .mp4
still i would say it would be fair if the container size is measured with files produced with the muxer which outputs the smallest filesizes (some muxer competition is not bad)

[multicone]

Again, comparing size alone makes no sense. If you set mkvtoolnix to not write a seekhead, the MKV will be smallest with a big difference. However, seeking in the file will then suffer. You can also write different cluster sizes, etc.

If you do compare this, you need to compare everything relevant for a container and validate the product of all this properties.

[robU*4]

alexnoe has posted a new overhead comparison between AVI, OGM, Matroska v1 & v2, MP4 using XviD. Matroska v2 (uses SimpleBlock) wins in all cases except for video only where MP4 wins.

[bond]

alexnoe has posted a new overhead comparison between AVI, OGM, Matroska v1 & v2, MP4 using XviD. Matroska v2 (uses SimpleBlock) wins in all cases except for video only where MP4 wins.edit: ah found it:
http://www-user.tu-chemnitz.de/~noe/Video-Zeug/AVIMux%20GUI/en_overhead_comparison.html

so for me this means .mp4 kicks normal mkv (v1), which till now is used, ass clearly in all cases!

about matroska v2:
which parsers support simpleblock? cause afaik its not backwards compatible and needs specific support (a new container? ;) :p )

[SeeMoreDigital]

Does anybody remember what Gej said about AVI's overhead compared to MP4's?


Cheers

[bond]

Does anybody remember what Gej said about AVI's overhead compared to MP4's?no, but it was propably something about that avi rocks :D

[stephanV]

So for video only, between standard AVI and MP4 there is a 3.9 kbps difference and between Matroska v1 and MP4 a 1.3 kbps difference. Yes, this is clearly a thing where a lot of "ass kicking" is done...

[SeeMoreDigital]

no, but it was propably something about that avi rocks :DYes... that was it....

Yep.... AVI rocks so much we decided to create our own version of it, which we will call .divx...... Coz you don't hear our brand name often enough.

Maybe the should make a range of DivX chocolate bars too!

[robU*4]

about matroska v2:
which parsers support simpleblock? cause afaik its not backwards compatible and needs specific support (a new container? ;) :p )

It was probably never mentioned on Doom9, but it's already supported by Haali's splitter, our 1.1.2 pack (150,000 ones already in the wild), VLC 0.8.4-test2a (test2 was missing matroska support), mkvmerge 1.6.0 (use an --engage hack) and AVI-Mux GUI. I might submit a patch for ffmpeg soon.

It's totally backward compatible, as files will not crash your player. It will just tell you your player does not support Matroska v2. But yes, it was called v2 to avoid user confusion. As for using a new codec, using a new feature usually requires an upgrade.

[robU*4]

Does anybody remember what Gej said about AVI's overhead compared to MP4's?


Cheers

Hey, don't bash my boss ;)

Gej is actually a cool guy.

[bond]

So for video only, between standard AVI and MP4 there is a 3.9 kbps difference and between Matroska v1 and MP4 a 1.3 kbps difference. Yes, this is clearly a thing where a lot of "ass kicking" is done...well you can also see it that way:
the overhead of .mp4 for video-only is ~50% smaller than the one from normal mkv

edit:
It's totally backward compatible, as files will not crash your player. It will just tell you your player does not support Matroska v2. But yes, it was called v2 to avoid user confusion. As for using a new codec, using a new feature usually requires an upgrade.ok, but if a file doesnt play its not backwards compatible imho ;)
anyways, the mkv support in tools is structured in a way its not hard to upgrade it to support new features, so this is not really a problem

[SeeMoreDigital]

Hey, don't bash my boss ;)

Gej is actually a cool guy.Yes.... there's nothing wrong with being cool....

I wonder how many DivX guys we have as Doom9 forum members now?

[Caroliano]

There are some things that I don't understand in this comparision:
The difference between (2) and (1) is the overhead caused by the audio stream, and the overhead you have to add if you want to add another stream of this type (this is not 100% accurate, but certainly 99.9%).
Where is the (2) and (1)? And only to check if I understand correctly: if I put another audio stream of the same type the audio overhead will simple double?

And how you can have an negative overhead?!?!?

[foxyshadis]

DTS has a great deal of overhead (error-correction?) inherent in the format already; by placing it in matroska it's possible to completely strip that out, saving a lot of space.

By "(2) and (1)" he means the two items in the list right before; ie, video-only and video+audio.

[Caroliano]

DTS has a great deal of overhead (error-correction?) inherent in the format already; by placing it in matroska it's possible to completely strip that out, saving a lot of space.
Cool. Is it mkv v1 compatible? Or it was implemented now?
By "(2) and (1)" he means the two items in the list right before; ie, video-only and video+audio
Oops. And if I put another audio stream of the same type the audio overhead will simple double or I don't understanded it also?
:thanks:

[stephanV]

well you can also see it that way:
the overhead of .mp4 for video-only is ~50% smaller than the one from normal mkv
Yes, but that would be doing marketing instead of showing the relevance of differences in overhead between containers.

[Elias]

And what about AVC in ASF format?Can AVC be placed in asf? I didn't know that.Gej is actually a cool guy.How can he be cool if his pronounced Gay in Swedish? :D

[robU*4]

Cool. Is it mkv v1 compatible? Or it was implemented now?

Oops. And if I put another audio stream of the same type the audio overhead will simple double or I don't understanded it also?
:thanks:

It's a new feature that is backward compatible. Although if your player doesn't support this "compression" method, that track won't play. Unless SimpleBlock which is used on all Blocks so all the data in Matroska v2 are unusable by a Matroska v1 parser.

The overhead is proportional to the number of streams, yes. But that means if you add tracks that would produce negative overhead, the overhead will decrease accordingly.