>> Progress
Face/Off (1997)
Subtitles.......: English, Swedish.
Audio Bitrate...: 640kbit/s
Audio Codec.....: AC3
Audio Language..: English
Audio untouched?: Yes
Video Bitrate...: 3685kbit/s
Video Codec.....: x264
Resolution......: 1280x544
Framerate.......: 23.976
Video source....: Blu-Ray H.264
Audio source....: Blu-Ray AC3
IMDb............: http://www.imdb.com/title/tt0119094/
Notes...........: To sumarize: the proper is because CDDHD rushed their rip with crappy settings.
We have been doing quite a few proper, its
sad, we dont like doing it, but it has to be
done to improve quality. We are aware that we
are contradicting the rules we signed with this
but read and you will know why.
We have "low" bitrate but you CANT judge quality soley on bitrate
anyone with a little encode experince will tell you this,
the reason is simply that we used the qunatization levels
determine the size of the rip. In easier terms we let the
encoder determine what bitrate was needed.
To CDDHD:
We have been enduring your settings quite too long, cddhd.
First of all 8x8dct=0 WHY? Dont give us the crap
about standalone compabiliy. The players arent out yet. We
dont know what they will support. If you are so worried about
compablity, go for WMWHD, and rape quality.
And your keyframe settings are insane. Min=1 and Max = 150?
Are you high? keyint min = 1, do you need a I-frame for every frame? WHEN?
Do you think it will help jumping in the movie?
Ever noticed problems with jumping in another rip? Its max 250 (~10sec)
and min 25 by default. Those settings are fine. I-frames take alot
more space and you cant work the algorithm magic on them. So there is a
reason to try and keep them as few as possible.
(But kudos you guys have used way worse on other rips).
Motion Estimation Method: Diamond.
To quote
http://www.digital-digest.com/articles/x264_options_page6.html
"Only use when maximum encoding speed is required."
This should NEVER ever be used.
Rate Distortion Optimization
brdo=0, wtf? you are using b-frames right? yeah you are (had to check).
Another rape on quality for no reason except faster encode.
mixed_ref=0, anyone supprised?
All these have been changed from better default
values, to these faster values. So its done on purpose to win
races and rape our quality.
Comparision screenshots are avabile in /screens if you dont trust us.
And as an extra bonus CDDHD rip is slightly overcropped.
The only reason their encode looks as good as ours is that they have
DOUBLE bitrate and still not better than ours!
There is need for other sizes than dvd5 and dvd9 but fortunetly thoose wheels
are already rolling, and hopefully soon an updated set of rules are avabile.
Encoder commandline we used for this:
--qp 20 --ref 5 --mixed-refs --no-fast-pskip --bframes 3 --b-pyramid --b-rdo --bime
--weightb --filter -2,-2 --subme 7 --analyse all --8x8dct --threads 2 --thread-input --progress --output
"faceoff.mkv" "faceoff.avs" --me umh
Trellis is useless on constant quant.
Encode log......: avis [info]: 1280x544 @ 23.98 fps (199863 frames)
x264 [info]: using cpu capabilities MMX MMXEXT SSE SSE2
The NCBENUM return code is: 0x0
The NCBENUM return code is: 0x0
x264 [info]: slice I:3873 Avg QP:17.00 size: 69008 PSNR Mean Y:46.38 U:50.87 V:51.83 Avg:47.46 Global:47.29
x264 [info]: slice P:121234 Avg QP:20.00 size: 23989 PSNR Mean Y:44.16 U:48.83 V:49.86 Avg:45.28 Global:45.00
x264 [info]: slice B:74756 Avg QP:21.81 size: 8858 PSNR Mean Y:43.35 U:48.37 V:49.44 Avg:44.51 Global:44.18
x264 [info]: mb I I16..4: 41.6% 46.2% 12.2%
x264 [info]: mb P I16..4: 9.3% 14.9% 2.2% P16..4: 39.2% 10.6% 4.2% 0.2% 0.1% skip:19.4%
x264 [info]: mb B I16..4: 0.6% 0.9% 0.2% B16..8: 34.4% 1.4% 3.6% direct: 5.1% skip:53.8%
x264 [info]: 8x8 transform intra:55.3% inter:46.4%
x264 [info]: ref P 66.1% 18.0% 9.1% 3.8% 3.0%
x264 [info]: ref B 78.3% 14.3% 3.9% 2.1% 1.4%
x264 [info]: SSIM Mean Y:0.9705157
x264 [info]: PSNR Mean Y:43.899 U:48.701 V:49.744 Avg:45.035 Global:44.706 kb/s:3683.06
This is a CQ job so theres no desired bitrate. Obtained video bitrate: 3685 kbit/s