[x264] 2 x Intel Xeon X5670 @ 2.93GHz HT ON vs HT OFF

[Atak_Snajpera]

It looks like that more does not mean faster ;)

HT OFF = 35.4 fps
2 x Intel Xeon X5670 @ 2.93GHz ( 6C / 6T )

x264 0.119.2106kMod 07efeb4
(libswscale 2.1.0)
(libavformat 53.17.0)
(ffmpegsource 2.16.2.1)
built by Komisar on Oct 23 2011, gcc: 4.4.7 20110720 (prerelease) (x86_64.generic.Komisar)
configuration: --bit-depth=8 --chroma-format=all
x264 license: GPL version 2 or later
libswscale/libavformat/ffmpegsource license: GPL version 2 or later

ffms [info]: 1920x1080p 1:1 @ 50/1 fps (vfr)

x264 [info]: using SAR=1/1

x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 FastShuffle SSE4.2

x264 [info]: profile High, level 4.2

x264 [info]: cabac=1 ref=3 deblock=1:0:0 analyse=0x3:0x113 me=hex subme=7 psy=1 fade_compensate=0.00 psy_rd=1.00:0.00 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=0 deadzone=21,11 fast_pskip=1 chroma_qp_offset=-2 threads=18 sliced_threads=0 nr=0 decimate=1 interlaced=0 bluray_compat=0 constrained_intra=0 bframes=3 b_pyramid=2 b_adapt=1 b_bias=0 direct=1 weightb=1 open_gop=0 weightp=2 keyint=250 keyint_min=25 scenecut=40 intra_refresh=0 rc_lookahead=40 rc=crf mbtree=1 crf=23.0000 qcomp=0.60 qpmin=0 qpmax=69 qpstep=4 ip_ratio=1.40 aq=1:1.00

x264 [info]: started at Sun Oct 30 17:41:39 2011

x264 [info]: frame I:10 Avg QP:25.24 size:293385

x264 [info]: frame P:1242 Avg QP:28.44 size: 90346

x264 [info]: frame B:1248 Avg QP:31.32 size: 19904

x264 [info]: consecutive B-frames: 8.7% 66.6% 23.2% 1.6%

x264 [info]: mb I I16..4: 3.6% 76.6% 19.8%

x264 [info]: mb P I16..4: 0.1% 4.7% 0.8% P16..4: 38.2% 25.5% 16.0% 0.0% 0.0% skip:14.7%

x264 [info]: mb B I16..4: 0.0% 0.4% 0.0% B16..8: 48.1% 6.2% 1.9% direct: 4.2% skip:39.1% L0:36.0% L1:49.9% BI:14.1%

x264 [info]: 8x8 transform intra:83.2% inter:58.4%

x264 [info]: coded y,uvDC,uvAC intra: 90.8% 82.8% 56.8% inter: 30.4% 19.6% 4.0%

x264 [info]: i16 v,h,dc,p: 17% 37% 10% 35%

x264 [info]: i8 v,h,dc,ddl,ddr,vr,hd,vl,hu: 11% 25% 16% 5% 7% 6% 12% 6% 12%

x264 [info]: i4 v,h,dc,ddl,ddr,vr,hd,vl,hu: 15% 21% 13% 7% 8% 7% 11% 7% 11%

x264 [info]: i8c dc,h,v,p: 52% 26% 15% 7%

x264 [info]: Weighted P-Frames: Y:1.0% UV:0.1%

x264 [info]: ref P L0: 70.4% 21.8% 5.9% 2.0% 0.0%

x264 [info]: ref B L0: 90.8% 7.8% 1.4%

x264 [info]: ref B L1: 95.7% 4.3%

x264 [info]: kb/s:22397.57



encoded 2500 frames, 35.44 fps, 22397.57 kb/s


x264 [info]: ended at Sun Oct 30 17:42:50 2011

x264 [info]: encoding duration 0:01:11

HT ON = 27.7 fps
2 x Intel Xeon X5670 @ 2.93GHz ( 6C / 12T )

x264 0.119.2106kMod 07efeb4
(libswscale 2.1.0)
(libavformat 53.17.0)
(ffmpegsource 2.16.2.1)
built by Komisar on Oct 23 2011, gcc: 4.4.7 20110720 (prerelease) (x86_64.generic.Komisar)
configuration: --bit-depth=8 --chroma-format=all
x264 license: GPL version 2 or later
libswscale/libavformat/ffmpegsource license: GPL version 2 or later

ffms [info]: 1920x1080p 1:1 @ 50/1 fps (vfr)

x264 [info]: using SAR=1/1

x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 FastShuffle SSE4.2

x264 [info]: profile High, level 4.2

x264 [info]: cabac=1 ref=3 deblock=1:0:0 analyse=0x3:0x113 me=hex subme=7 psy=1 fade_compensate=0.00 psy_rd=1.00:0.00 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=0 deadzone=21,11 fast_pskip=1 chroma_qp_offset=-2 threads=36 sliced_threads=0 nr=0 decimate=1 interlaced=0 bluray_compat=0 constrained_intra=0 bframes=3 b_pyramid=2 b_adapt=1 b_bias=0 direct=1 weightb=1 open_gop=0 weightp=2 keyint=250 keyint_min=25 scenecut=40 intra_refresh=0 rc_lookahead=40 rc=crf mbtree=1 crf=23.0000 qcomp=0.60 qpmin=0 qpmax=69 qpstep=4 ip_ratio=1.40 aq=1:1.00

x264 [info]: started at Sun Oct 30 17:30:59 2011

x264 [info]: frame I:10 Avg QP:25.24 size:293382

x264 [info]: frame P:1242 Avg QP:28.44 size: 90349

x264 [info]: frame B:1248 Avg QP:31.32 size: 19906

x264 [info]: consecutive B-frames: 8.7% 66.6% 23.2% 1.6%

x264 [info]: mb I I16..4: 3.6% 76.6% 19.8%

x264 [info]: mb P I16..4: 0.1% 4.7% 0.8% P16..4: 38.2% 25.5% 16.0% 0.0% 0.0% skip:14.7%

x264 [info]: mb B I16..4: 0.0% 0.4% 0.0% B16..8: 48.1% 6.2% 1.9% direct: 4.2% skip:39.1% L0:36.0% L1:49.9% BI:14.1%

x264 [info]: 8x8 transform intra:83.1% inter:58.4%

x264 [info]: coded y,uvDC,uvAC intra: 90.8% 82.8% 56.8% inter: 30.4% 19.6% 4.0%

x264 [info]: i16 v,h,dc,p: 17% 37% 10% 36%

x264 [info]: i8 v,h,dc,ddl,ddr,vr,hd,vl,hu: 11% 25% 16% 5% 7% 6% 12% 6% 13%

x264 [info]: i4 v,h,dc,ddl,ddr,vr,hd,vl,hu: 15% 21% 13% 7% 8% 7% 11% 7% 11%

x264 [info]: i8c dc,h,v,p: 52% 26% 15% 7%

x264 [info]: Weighted P-Frames: Y:1.0% UV:0.1%

x264 [info]: ref P L0: 70.4% 21.8% 5.9% 2.0% 0.0%

x264 [info]: ref B L0: 90.8% 7.8% 1.4%

x264 [info]: ref B L1: 95.6% 4.4%

x264 [info]: kb/s:22398.38



encoded 2500 frames, 27.74 fps, 22398.38 kb/s


x264 [info]: ended at Sun Oct 30 17:32:29 2011

x264 [info]: encoding duration 0:01:30

[J_Darnley]

DAMN! I would like a box with those in!

Have you tried 19-35 threads with HT on? I recall it being said that the useful limit of threading was around height / 40. So with HT and 36 threads, it might be 9 threads too many. Have you considered expanding the tests done for doc/threads.txt?

[cegy]

its 1.5 thread per cpu core/thread (i.e 4cores/8threads (ht on) = 12 threads in x264, 6core/12 threads (ht on) = 18 threads in x264)
in this case you should use 0 threads, if not try 36 threads

[Dark Shikari]

Are you sure that the CPU isn't just auto-overclocking more when fewer cores are used?

[Atak_Snajpera]

@cegi
i let x264 to set optimal number of threads. check logs.

[Groucho2004]

Maybe the x264 was running out of memory (and the OS dumping stuff to the page file) with 36 threads. Unfortunately, the log (wherever it comes from) doesn't state the total memory installed and the usage during the encoding process.

[Atak_Snajpera]

Are you sure that the CPU isn't just auto-overclocking more when fewer cores are used?
HT OFF
http://i.imgur.com/enf49.jpg

HT ON
http://i.imgur.com/zLdcq.jpg

Maybe the x264 was running out of memory (and the OS dumping stuff to the page file) with 36 threads. Unfortunately, the log (wherever it comes from) doesn't state the total memory installed and the usage during the encoding process.
http://i.imgur.com/0hBFd.jpg

[Dark Shikari]

HT OFF
HT ON

... was that with x264 running? Do you have an average of the CPU speed over the total runtime, given that the auto-overclocking fluctuations dozens of times a second?

[Atak_Snajpera]

... was that with x264 running?
yes.

Do you have an average of the CPU speed over the total runtime, given that the auto-overclocking fluctuations dozens of times a second?
What would be good tool for that?

Also Maximum Turbo-mode for this model is 3.33 GHz

[LoRd_MuldeR]

Maybe TMonitor?
http://www.cpuid.com/softwares/tmonitor.html

(Seems to have a logging functionality. From the log it should be trivial to calculate the average)

[nm]

Maybe it's an input bottleneck in FFMS. How many --demuxer-threads have you set?

[akupenguin]

Can you generate a graph of speed vs number of threads?

[Atak_Snajpera]

Maybe it's an input bottleneck in FFMS. How many --demuxer-threads have you set?
--demuxer-threads 6 (HT OFF)
--demuxer-threads 12 (HT ON)

[Atak_Snajpera]

26 threads
2 x Intel Xeon X5670 @ 2.93GHz ( 6C / 12T )

x264 0.119.2106kMod 07efeb4
(libswscale 2.1.0)
(libavformat 53.17.0)
(ffmpegsource 2.16.2.1)
built by Komisar on Oct 23 2011, gcc: 4.4.7 20110720 (prerelease) (x86_64.generic.Komisar)
configuration: --bit-depth=8 --chroma-format=all
x264 license: GPL version 2 or later
libswscale/libavformat/ffmpegsource license: GPL version 2 or later

ffms [info]: 1920x1080p 1:1 @ 50/1 fps (vfr)

x264 [info]: using SAR=1/1

x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 FastShuffle SSE4.2

x264 [info]: profile High, level 4.2

x264 [info]: cabac=1 ref=3 deblock=1:0:0 analyse=0x3:0x113 me=hex subme=7 psy=1 fade_compensate=0.00 psy_rd=1.00:0.00 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=0 deadzone=21,11 fast_pskip=1 chroma_qp_offset=-2 threads=26 sliced_threads=0 nr=0 decimate=1 interlaced=0 bluray_compat=0 constrained_intra=0 bframes=3 b_pyramid=2 b_adapt=1 b_bias=0 direct=1 weightb=1 open_gop=0 weightp=2 keyint=250 keyint_min=25 scenecut=40 intra_refresh=0 rc_lookahead=40 rc=crf mbtree=1 crf=23.0000 qcomp=0.60 qpmin=0 qpmax=69 qpstep=4 ip_ratio=1.40 aq=1:1.00

x264 [info]: started at Tue Nov 01 18:49:13 2011

x264 [info]: frame I:10 Avg QP:25.24 size:293360

x264 [info]: frame P:1242 Avg QP:28.44 size: 90343

x264 [info]: frame B:1248 Avg QP:31.31 size: 19906

x264 [info]: consecutive B-frames: 8.7% 66.6% 23.2% 1.6%

x264 [info]: mb I I16..4: 3.6% 76.7% 19.7%

x264 [info]: mb P I16..4: 0.1% 4.6% 0.8% P16..4: 38.2% 25.5% 16.0% 0.0% 0.0% skip:14.7%

x264 [info]: mb B I16..4: 0.0% 0.4% 0.0% B16..8: 48.1% 6.2% 1.9% direct: 4.2% skip:39.2% L0:36.0% L1:49.9% BI:14.1%

x264 [info]: 8x8 transform intra:83.2% inter:58.4%

x264 [info]: coded y,uvDC,uvAC intra: 90.8% 82.8% 56.8% inter: 30.4% 19.6% 4.0%

x264 [info]: i16 v,h,dc,p: 17% 37% 10% 35%

x264 [info]: i8 v,h,dc,ddl,ddr,vr,hd,vl,hu: 11% 25% 16% 5% 7% 6% 12% 6% 12%

x264 [info]: i4 v,h,dc,ddl,ddr,vr,hd,vl,hu: 15% 21% 13% 7% 8% 7% 11% 7% 11%

x264 [info]: i8c dc,h,v,p: 52% 26% 15% 7%

x264 [info]: Weighted P-Frames: Y:1.0% UV:0.1%

x264 [info]: ref P L0: 70.4% 21.8% 5.9% 2.0% 0.0%

x264 [info]: ref B L0: 90.8% 7.8% 1.4%

x264 [info]: ref B L1: 95.7% 4.3%

x264 [info]: kb/s:22397.26



encoded 2500 frames, 27.32 fps, 22397.26 kb/s


x264 [info]: ended at Tue Nov 01 18:50:44 2011

x264 [info]: encoding duration 0:01:31

28 threads
2 x Intel Xeon X5670 @ 2.93GHz ( 6C / 12T )

x264 0.119.2106kMod 07efeb4
(libswscale 2.1.0)
(libavformat 53.17.0)
(ffmpegsource 2.16.2.1)
built by Komisar on Oct 23 2011, gcc: 4.4.7 20110720 (prerelease) (x86_64.generic.Komisar)
configuration: --bit-depth=8 --chroma-format=all
x264 license: GPL version 2 or later
libswscale/libavformat/ffmpegsource license: GPL version 2 or later

ffms [info]: 1920x1080p 1:1 @ 50/1 fps (vfr)

x264 [info]: using SAR=1/1

x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 FastShuffle SSE4.2

x264 [info]: profile High, level 4.2

x264 [info]: cabac=1 ref=3 deblock=1:0:0 analyse=0x3:0x113 me=hex subme=7 psy=1 fade_compensate=0.00 psy_rd=1.00:0.00 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=0 deadzone=21,11 fast_pskip=1 chroma_qp_offset=-2 threads=28 sliced_threads=0 nr=0 decimate=1 interlaced=0 bluray_compat=0 constrained_intra=0 bframes=3 b_pyramid=2 b_adapt=1 b_bias=0 direct=1 weightb=1 open_gop=0 weightp=2 keyint=250 keyint_min=25 scenecut=40 intra_refresh=0 rc_lookahead=40 rc=crf mbtree=1 crf=23.0000 qcomp=0.60 qpmin=0 qpmax=69 qpstep=4 ip_ratio=1.40 aq=1:1.00

x264 [info]: started at Tue Nov 01 18:51:29 2011

x264 [info]: frame I:10 Avg QP:25.24 size:293360

x264 [info]: frame P:1242 Avg QP:28.44 size: 90345

x264 [info]: frame B:1248 Avg QP:31.32 size: 19906

x264 [info]: consecutive B-frames: 8.7% 66.6% 23.2% 1.6%

x264 [info]: mb I I16..4: 3.6% 76.7% 19.7%

x264 [info]: mb P I16..4: 0.1% 4.7% 0.8% P16..4: 38.2% 25.6% 16.0% 0.0% 0.0% skip:14.7%

x264 [info]: mb B I16..4: 0.0% 0.4% 0.0% B16..8: 48.1% 6.2% 1.9% direct: 4.2% skip:39.2% L0:36.0% L1:49.9% BI:14.1%

x264 [info]: 8x8 transform intra:83.2% inter:58.4%

x264 [info]: coded y,uvDC,uvAC intra: 90.8% 82.8% 56.8% inter: 30.4% 19.6% 4.0%

x264 [info]: i16 v,h,dc,p: 17% 37% 10% 35%

x264 [info]: i8 v,h,dc,ddl,ddr,vr,hd,vl,hu: 11% 25% 16% 5% 7% 6% 12% 6% 13%

x264 [info]: i4 v,h,dc,ddl,ddr,vr,hd,vl,hu: 15% 21% 13% 7% 8% 7% 11% 7% 11%

x264 [info]: i8c dc,h,v,p: 52% 26% 15% 7%

x264 [info]: Weighted P-Frames: Y:1.0% UV:0.1%

x264 [info]: ref P L0: 70.4% 21.8% 5.9% 2.0% 0.0%

x264 [info]: ref B L0: 90.8% 7.8% 1.4%

x264 [info]: ref B L1: 95.6% 4.4%

x264 [info]: kb/s:22397.75



encoded 2500 frames, 28.36 fps, 22397.75 kb/s


x264 [info]: ended at Tue Nov 01 18:52:57 2011

x264 [info]: encoding duration 0:01:28

30 threads
2 x Intel Xeon X5670 @ 2.93GHz ( 6C / 12T )

x264 0.119.2106kMod 07efeb4
(libswscale 2.1.0)
(libavformat 53.17.0)
(ffmpegsource 2.16.2.1)
built by Komisar on Oct 23 2011, gcc: 4.4.7 20110720 (prerelease) (x86_64.generic.Komisar)
configuration: --bit-depth=8 --chroma-format=all
x264 license: GPL version 2 or later
libswscale/libavformat/ffmpegsource license: GPL version 2 or later

ffms [info]: 1920x1080p 1:1 @ 50/1 fps (vfr)

x264 [info]: using SAR=1/1

x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 FastShuffle SSE4.2

x264 [info]: profile High, level 4.2

x264 [info]: cabac=1 ref=3 deblock=1:0:0 analyse=0x3:0x113 me=hex subme=7 psy=1 fade_compensate=0.00 psy_rd=1.00:0.00 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=0 deadzone=21,11 fast_pskip=1 chroma_qp_offset=-2 threads=30 sliced_threads=0 nr=0 decimate=1 interlaced=0 bluray_compat=0 constrained_intra=0 bframes=3 b_pyramid=2 b_adapt=1 b_bias=0 direct=1 weightb=1 open_gop=0 weightp=2 keyint=250 keyint_min=25 scenecut=40 intra_refresh=0 rc_lookahead=40 rc=crf mbtree=1 crf=23.0000 qcomp=0.60 qpmin=0 qpmax=69 qpstep=4 ip_ratio=1.40 aq=1:1.00

x264 [info]: started at Tue Nov 01 18:54:10 2011

x264 [info]: frame I:10 Avg QP:25.24 size:293366

x264 [info]: frame P:1242 Avg QP:28.44 size: 90345

x264 [info]: frame B:1248 Avg QP:31.32 size: 19909

x264 [info]: consecutive B-frames: 8.7% 66.6% 23.2% 1.6%

x264 [info]: mb I I16..4: 3.6% 76.6% 19.8%

x264 [info]: mb P I16..4: 0.1% 4.7% 0.8% P16..4: 38.2% 25.5% 16.0% 0.0% 0.0% skip:14.7%

x264 [info]: mb B I16..4: 0.0% 0.4% 0.0% B16..8: 48.1% 6.2% 1.9% direct: 4.2% skip:39.1% L0:36.0% L1:49.9% BI:14.1%

x264 [info]: 8x8 transform intra:83.2% inter:58.4%

x264 [info]: coded y,uvDC,uvAC intra: 90.8% 82.8% 56.8% inter: 30.4% 19.6% 4.0%

x264 [info]: i16 v,h,dc,p: 18% 37% 10% 35%

x264 [info]: i8 v,h,dc,ddl,ddr,vr,hd,vl,hu: 11% 25% 16% 5% 7% 6% 12% 6% 13%

x264 [info]: i4 v,h,dc,ddl,ddr,vr,hd,vl,hu: 15% 21% 13% 7% 8% 7% 11% 7% 11%

x264 [info]: i8c dc,h,v,p: 52% 26% 15% 7%

x264 [info]: Weighted P-Frames: Y:1.0% UV:0.1%

x264 [info]: ref P L0: 70.4% 21.8% 5.9% 2.0% 0.0%

x264 [info]: ref B L0: 90.8% 7.8% 1.4%

x264 [info]: ref B L1: 95.6% 4.4%

x264 [info]: kb/s:22398.26



encoded 2500 frames, 28.77 fps, 22398.26 kb/s


x264 [info]: ended at Tue Nov 01 18:55:37 2011

x264 [info]: encoding duration 0:01:27

32 threads
2 x Intel Xeon X5670 @ 2.93GHz ( 6C / 12T )

x264 0.119.2106kMod 07efeb4
(libswscale 2.1.0)
(libavformat 53.17.0)
(ffmpegsource 2.16.2.1)
built by Komisar on Oct 23 2011, gcc: 4.4.7 20110720 (prerelease) (x86_64.generic.Komisar)
configuration: --bit-depth=8 --chroma-format=all
x264 license: GPL version 2 or later
libswscale/libavformat/ffmpegsource license: GPL version 2 or later

ffms [info]: 1920x1080p 1:1 @ 50/1 fps (vfr)

x264 [info]: using SAR=1/1

x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 FastShuffle SSE4.2

x264 [info]: profile High, level 4.2

x264 [info]: cabac=1 ref=3 deblock=1:0:0 analyse=0x3:0x113 me=hex subme=7 psy=1 fade_compensate=0.00 psy_rd=1.00:0.00 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=0 deadzone=21,11 fast_pskip=1 chroma_qp_offset=-2 threads=32 sliced_threads=0 nr=0 decimate=1 interlaced=0 bluray_compat=0 constrained_intra=0 bframes=3 b_pyramid=2 b_adapt=1 b_bias=0 direct=1 weightb=1 open_gop=0 weightp=2 keyint=250 keyint_min=25 scenecut=40 intra_refresh=0 rc_lookahead=40 rc=crf mbtree=1 crf=23.0000 qcomp=0.60 qpmin=0 qpmax=69 qpstep=4 ip_ratio=1.40 aq=1:1.00

x264 [info]: started at Tue Nov 01 18:56:14 2011

x264 [info]: frame I:10 Avg QP:25.24 size:293385

x264 [info]: frame P:1242 Avg QP:28.44 size: 90345

x264 [info]: frame B:1248 Avg QP:31.32 size: 19903

x264 [info]: consecutive B-frames: 8.7% 66.6% 23.2% 1.6%

x264 [info]: mb I I16..4: 3.6% 76.6% 19.8%

x264 [info]: mb P I16..4: 0.1% 4.7% 0.8% P16..4: 38.2% 25.5% 16.0% 0.0% 0.0% skip:14.7%

x264 [info]: mb B I16..4: 0.0% 0.4% 0.0% B16..8: 48.0% 6.2% 1.9% direct: 4.2% skip:39.2% L0:36.0% L1:49.9% BI:14.1%

x264 [info]: 8x8 transform intra:83.2% inter:58.4%

x264 [info]: coded y,uvDC,uvAC intra: 90.8% 82.8% 56.8% inter: 30.4% 19.6% 4.0%

x264 [info]: i16 v,h,dc,p: 17% 37% 10% 35%

x264 [info]: i8 v,h,dc,ddl,ddr,vr,hd,vl,hu: 11% 25% 16% 5% 7% 6% 12% 6% 12%

x264 [info]: i4 v,h,dc,ddl,ddr,vr,hd,vl,hu: 15% 21% 13% 7% 8% 7% 11% 7% 11%

x264 [info]: i8c dc,h,v,p: 52% 26% 15% 7%

x264 [info]: Weighted P-Frames: Y:1.0% UV:0.1%

x264 [info]: ref P L0: 70.4% 21.8% 5.9% 2.0% 0.0%

x264 [info]: ref B L0: 90.9% 7.7% 1.4%

x264 [info]: ref B L1: 95.7% 4.3%

x264 [info]: kb/s:22397.13



encoded 2500 frames, 26.18 fps, 22397.12 kb/s


x264 [info]: ended at Tue Nov 01 18:57:50 2011

x264 [info]: encoding duration 0:01:36

34 threads
2 x Intel Xeon X5670 @ 2.93GHz ( 6C / 12T )

x264 0.119.2106kMod 07efeb4
(libswscale 2.1.0)
(libavformat 53.17.0)
(ffmpegsource 2.16.2.1)
built by Komisar on Oct 23 2011, gcc: 4.4.7 20110720 (prerelease) (x86_64.generic.Komisar)
configuration: --bit-depth=8 --chroma-format=all
x264 license: GPL version 2 or later
libswscale/libavformat/ffmpegsource license: GPL version 2 or later

ffms [info]: 1920x1080p 1:1 @ 50/1 fps (vfr)

x264 [info]: using SAR=1/1

x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 FastShuffle SSE4.2

x264 [info]: profile High, level 4.2

x264 [info]: cabac=1 ref=3 deblock=1:0:0 analyse=0x3:0x113 me=hex subme=7 psy=1 fade_compensate=0.00 psy_rd=1.00:0.00 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=0 deadzone=21,11 fast_pskip=1 chroma_qp_offset=-2 threads=34 sliced_threads=0 nr=0 decimate=1 interlaced=0 bluray_compat=0 constrained_intra=0 bframes=3 b_pyramid=2 b_adapt=1 b_bias=0 direct=1 weightb=1 open_gop=0 weightp=2 keyint=250 keyint_min=25 scenecut=40 intra_refresh=0 rc_lookahead=40 rc=crf mbtree=1 crf=23.0000 qcomp=0.60 qpmin=0 qpmax=69 qpstep=4 ip_ratio=1.40 aq=1:1.00

x264 [info]: started at Tue Nov 01 18:58:18 2011

x264 [info]: frame I:10 Avg QP:25.24 size:293382

x264 [info]: frame P:1242 Avg QP:28.44 size: 90348

x264 [info]: frame B:1248 Avg QP:31.32 size: 19913

x264 [info]: consecutive B-frames: 8.7% 66.6% 23.2% 1.6%

x264 [info]: mb I I16..4: 3.6% 76.6% 19.8%

x264 [info]: mb P I16..4: 0.1% 4.7% 0.8% P16..4: 38.2% 25.5% 16.0% 0.0% 0.0% skip:14.7%

x264 [info]: mb B I16..4: 0.0% 0.4% 0.0% B16..8: 48.1% 6.2% 1.9% direct: 4.2% skip:39.1% L0:36.0% L1:49.9% BI:14.1%

x264 [info]: 8x8 transform intra:83.2% inter:58.4%

x264 [info]: coded y,uvDC,uvAC intra: 90.8% 82.8% 56.8% inter: 30.4% 19.6% 4.0%

x264 [info]: i16 v,h,dc,p: 17% 38% 10% 35%

x264 [info]: i8 v,h,dc,ddl,ddr,vr,hd,vl,hu: 11% 25% 16% 5% 7% 6% 12% 6% 12%

x264 [info]: i4 v,h,dc,ddl,ddr,vr,hd,vl,hu: 15% 21% 13% 7% 8% 7% 11% 7% 11%

x264 [info]: i8c dc,h,v,p: 52% 26% 15% 7%

x264 [info]: Weighted P-Frames: Y:1.0% UV:0.1%

x264 [info]: ref P L0: 70.4% 21.7% 5.9% 2.0% 0.0%

x264 [info]: ref B L0: 90.8% 7.8% 1.4%

x264 [info]: ref B L1: 95.6% 4.4%

x264 [info]: kb/s:22399.56



encoded 2500 frames, 27.86 fps, 22399.56 kb/s


x264 [info]: ended at Tue Nov 01 18:59:48 2011

x264 [info]: encoding duration 0:01:30

[Laraul]

More threads you run in parallel results in greater loss of quality... so you shouldn't run more than 8-12 threads at once (using x264). This means you shouldn't use --threads auto since x264 will use 18 threads.

[Groucho2004]

More threads you run in parallel results in greater loss of quality

Isn't that true only for slice based threads?

[Laraul]

From the documentation...

threads
Default: auto (frame based threads: 1.5 * logical processors, rounded down; slice based threads: 1 * logical processors)
Enables parallel encoding by using more than 1 thread to increase speed on multi-core systems. The quality loss from multiple threads is mostly negligible unless using very high numbers of threads (say, above 16). The speed gain should be slightly less than linear until you start using more than 1 thread per 40px of vertical video, at which point the gain from additional threads sharply decreases.
x264 currently has an internal limit on the number of threads set at 128, realistically you should never set it this high.
Recommendation: Default
See also: --sliced-threads (http://mewiki.project357.com/wiki/X264_Settings#sliced-threads" title="X264 Settings") --thread-input (http://mewiki.project357.com/wiki/X264_Settings#thread-input" title="X264 Settings")

[Blue_MiSfit]

I have a lot of dual X5650 servers (12 core, 24 thread), and chose to simply run multiple encodes in parallel with 12 threads each.

[TheRyuu]

I can only guess that 18 threads is simply too much given the source (IIRC you don't want to be exceeding 16 threads). HT should yield about a 25% performance increase over non-HT on the same CPU (at least on cpus for the less fortunate (4C/8T) :p).