Current Patches, Where to get them, How they affect speed/output

[Lux Delux]

That wasn't my problem though... but it seems my own stupidity as I put FFTW3.DLL in system32 instead of sysWOW64 DOH!:o Seems to be running fine now.

Anyways sorry for going off topic since it turned out to be a an avs prob. Cheers again to all for the hard work with x264 :)

[skystrife]

x264.920.modified.02.exe (http://www.mediafire.com/?lxgztx3yg3y) - Alternate Download (http://skystrife.com/x264/x264.920.modified.02.exe)
(Source) (http://www.mediafire.com/?wxll0xwe0yy) - Alternate Download (http://skystrife.com/x264/x264.920.modified.02.source.7z)

Patches used:

x264_psyRDO_0.5.diff
x264_hrd_pulldown.09_interlace.diff
x264.progress.indication.01.diff
x264_new_bframe_decision_02.diff <-- This patch is highly experimental and has caused crashes on my end. It is enabled by adding --b-adapt 2 to the command line. I would recommend only doing this for testing.

All patches are included with the source.

gcc 3.4.5 fprofiled build.

EDIT: To be fair, it might be a confliction with the patches. I have no idea, to be honest. =P

[kemuri-_9]

If anyone wanted that 2160p50fps crowdrun sample but didn't want the hassle of going through all the downloading and converting for it, I converted it to xvid q2 which drastically brought the size down from the original 23 GB to about 625 MB.

I couldn't manage to grab the last few frames though, they kept giving me permission denied....
http://kemuri9.net/dev/x264/samples/crowdrun_xvid_q2.avi
if you want it.
if you just want a couple frames in png, then can also get them from
http://kemuri9.net/dev/x264/samples/crowdrun/

[LoRd_MuldeR]

If anyone wanted that 2160p50fps crowdrun sample but didn't want the hassle of going through all the downloading and converting for it, I converted it to xvid q2 which drastically brought the size down from the original 23 GB to about 625 MB.

I couldn't manage to grab the last few frames though, they kept giving me permission denied....
http://kemuri9.net/dev/x264/samples/crowdrun_xvid_q2.avi
if you want it.
if you just want a couple frames in png, then can also get them from
http://kemuri9.net/dev/x264/samples/crowdrun/

:thanks:

[bob0r]

x264.922.modified.01.exe (http://files.x264.nl/x264.922.modified.01.exe)
libx264-60.922.modified.01.dll (http://files.x264.nl/libx264-60.922.modified.01.dll)

x264-psyrd-0.5.diff
x264_hrd_pulldown.09_interlace.diff
x264.progress.indication.01.diff

[TL0]

Can someone compile a version with x264_new_bframe_decision patch?

:thanks:

[gigah72]

Can someone compile a version with x264_new_bframe_decision patch?

:thanks:

http://www.mediafire.com/?zb0dgxly1gc

with:
bframe patch: http://pastebin.com/m2a557e16
x264-psyrd-0.5.diff
x264_hrd_pulldown.09_interlace.diff
x264.progress.indication.01.diff

[gav1577]

Hi the above build ^ does not work for me can anyone confirm ? thanks

[martino]

Works fine here.

[gav1577]

Works fine here.
Thats odd does not work for me with --b-adapt 2 However skystrife's 920 Build does :confused:

EDIT: I found the problem the build for some reason will not work with a custom cqm where as skystrife's build will. I know its not recommended to use a cqm with psyrd but thought i should let it be known just in case its a bug :-)

[bob0r]

x264.928.modified.01.exe (http://files.x264.nl/x264.928.modified.01.exe)
libx264-60.928.modified.01.dll (http://files.x264.nl/libx264-60.928.modified.01.dll)

x264-psyrd-0.5.diff
x264_hrd_pulldown.09_interlace.diff
x264.progress.indication.01.diff

[le_canz]

I apologise if I'm asking a dumb question, but....

I try to keep following x264 development, and I'm wondering why the HRD patch isn't commited into git, as it is required to generate blu-ray compliant streams ?

Or maybe I have completely misunderstood how blu-ray compliant streams are outputed ?

Anyway, thanks to all the people involved in x264 ; developpers as well as providers of experimental / patched builds :thanks:

[bkman]

Just a note that I also get a crash, at the very start of pass 2 with this build http://forum.doom9.org/showthread.php?p=1164702#post1164702 and --b-adapt 2 enabled.

commandline:
"C:\Apps\megui\tools\x264\x264.exe" --pass 2 --bitrate 1918 --stats "H:\Encoding\Commando\hfyu_Commando3.stats" --level 4.1 --min-keyint 24 --ref 5 --mixed-refs --no-fast-pskip --bframes 3 --b-pyramid --b-rdo --bime --weightb --direct auto --filter -2,-1 --subme 7 --trellis 2 --partitions p8x8,b8x8,i4x4,i8x8 --8x8dct --ratetol 2 --me umh --merange 24 --threads auto --thread-input --sar 629:450 --progress --no-dct-decimate --output "H:\Encoding\Commando\hfyu_Commando301.mkv" "H:\Encoding\Commando\hfyu_Commando3.avs" --keyint 1500 --b-adapt 2

[gigah72]

removed,
build crashes on 2-pass encode

[skystrife]

x264.928.modified.exe (http://www.mediafire.com/?yxjg93s1lys) - Alternate Download (http://skystrife.com/x264/x264.928.modified.exe)
libx264-61.928.modified.dll (http://www.mediafire.com/?9xscjgmn2cj) - Alternate Download (http://skystrife.com/x264/libx264-61.928.modified.dll)
(Source) (http://www.mediafire.com/?ud9gdgvdctg) - Alternate Download (http://skystrife.com/x264/x264.928.modified.source.7z)

Patches used:

x264_psyRDO_0.5.diff
x264_hrd_pulldown.09_interlace.diff
x264.progress.indication.01.diff
x264_new_bframe_decision_03.diff <-- This patch is highly experimental, but no longer crashes on my end. Thank you Dark Shikari!

All patches are included with the source.

gcc 3.4.5 fprofiled build.

[gav1577]

Thanks for the ^build^ skystrife ;)

[elguaxo]

thanks skystrife!

[MythCreator]

@Skystrife:

Where could I get x264_new_bframe_decision_03.diff ? Thanks~

[Yoshiyuki Blade]

@Skystrife:

Where could I get x264_new_bframe_decision_03.diff ? Thanks~

It's in the 7zip file along with the source and other patches (3rd link) :p

[MythCreator]

There is something strange..I Set bitrate=4640Kbps ，Two passes，but the final bitrate is 4822.92Kbps..

Settings:

1st pass:
--cqmfile "M4G_High_Detail_V3.1.cfg" --no-fast-pskip --no-dct-decimate --mixed-refs --scenecut 40 -I 90 -i 25 --me umh --merange 32 --sar 1:1 --threads 3 -r 6 --nr 0 -f -2:-1 -b 16 --direct "auto" --bime --b-bias 0 --b-pyramid -w --deadzone-inter 4 --deadzone-intra 6 --aq-mode 0 --chroma-qp-offset 0 --direct-8x8 1 -p 1 -B 4640 --qpmin 10 --qpmax 51 --8x8dct -A p8x8,p4x4,b8x8,i8x8,i4x4 --ipratio 1.400000 --pbratio 1.300000 -m 7 --b-rdo --ratetol 1.000000 --vbv-init 0.9

2nd pass:
--cqmfile "M4G_High_Detail_V3.1.cfg" --no-fast-pskip --no-dct-decimate --mixed-refs --scenecut 40 -I 90 -i 25 --me umh --merange 32 --sar 1:1 --threads 3 -r 6 --nr 0 -f -2:-1 -b 16 --direct "auto" --bime --b-bias 0 --b-pyramid -w --deadzone-inter 4 --deadzone-intra 6 --aq-mode 0 --chroma-qp-offset 0 --direct-8x8 1 -p 3 -B 4640 --qpmin 10 --qpmax 51 --qpstep 4 --qcomp 0.900000 --ipratio 1.400000 --pbratio 1.300000 --8x8dct -A p8x8,p4x4,b8x8,i8x8,i4x4 -m 7 --b-rdo --ratetol 1.000000 --vbv-init 0.9


Log：


x264 [info]: slice I:9 Avg QP:18.00 size: 98327 PSNR Mean Y:53.52 U:52.63 V:51.76 Avg:52.89 Global:51.94
x264 [info]: slice P:144 Avg QP:18.23 size: 52380 PSNR Mean Y:51.69 U:52.74 V:52.10 Avg:51.87 Global:51.31
x264 [info]: slice B:348 Avg QP:19.81 size: 14616 PSNR Mean Y:50.94 U:52.43 V:51.92 Avg:51.28 Global:50.59
x264 [info]: consecutive B-frames: 6.1% 14.2% 6.7% 16.3% 16.3% 35.4% 2.8% 0.0% 0.0% 0.0% 2.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
x264 [info]: mb I I16..4: 22.1% 72.2% 5.8%
x264 [info]: mb P I16..4: 6.8% 14.1% 0.8% P16..4: 48.9% 10.9% 5.9% 0.1% 0.0% skip:12.5%
x264 [info]: mb B I16..4: 0.6% 0.9% 0.1% B16..8: 34.0% 1.0% 1.2% direct: 1.9% skip:60.4% L0:46.9% L1:43.6% BI: 9.5%
x264 [info]: final ratefactor: 20.15
x264 [info]: 8x8 transform intra:65.8% inter:30.6%
x264 [info]: direct mvs spatial:93.4% temporal:6.6%
x264 [info]: ref P L0 73.8% 15.9% 5.7% 2.2% 1.3% 1.0%
x264 [info]: ref B L0 82.9% 11.5% 3.6% 1.4% 0.6%
x264 [info]: ref B L1 95.8% 4.2%
x264 [info]: SSIM Mean Y:0.9947734
x264 [info]: PSNR Mean Y:51.201 U:52.521 V:51.966 Avg:51.479 Global:50.809 kb/s:5173.88


x264 [info]: slice I:9 Avg QP:17.11 size:102517 PSNR Mean Y:53.82 U:52.51 V:51.71 Avg:52.99 Global:52.51
x264 [info]: slice P:144 Avg QP:19.10 size: 47634 PSNR Mean Y:51.47 U:52.48 V:51.80 Avg:51.63 Global:51.34
x264 [info]: slice B:348 Avg QP:20.53 size: 13832 PSNR Mean Y:50.75 U:52.30 V:51.75 Avg:51.10 Global:50.79
x264 [info]: consecutive B-frames: 6.1% 14.2% 6.7% 16.3% 16.3% 35.4% 2.8% 0.0% 0.0% 0.0% 2.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
x264 [info]: mb I I16..4: 20.2% 74.5% 5.3%
x264 [info]: mb P I16..4: 7.0% 14.0% 0.6% P16..4: 49.3% 9.0% 5.4% 0.1% 0.0% skip:14.6%
x264 [info]: mb B I16..4: 0.5% 0.8% 0.1% B16..8: 32.8% 0.8% 0.9% direct: 1.6% skip:62.4% L0:47.0% L1:41.8% BI:11.2%
x264 [info]: 8x8 transform intra:66.1% inter:29.0%
x264 [info]: direct mvs spatial:88.5% temporal:11.5%
x264 [info]: ref P L0 72.6% 17.2% 5.7% 2.2% 1.3% 1.1%
x264 [info]: ref B L0 82.7% 12.0% 3.3% 1.3% 0.6%
x264 [info]: ref B L1 96.2% 3.8%
x264 [info]: SSIM Mean Y:0.9947686
x264 [info]: PSNR Mean Y:51.014 U:52.352 V:51.766 Avg:51.286 Global:50.967 kb/s:4822.15


encoded 501 frames, 0.65 fps, 4822.92 kb/s




using skystrife's build

[buzzqw]

@MythCreator

please try without a custom matrix

BHH

[Sharktooth]

501 frames. the clip is too short.

[kemuri-_9]

--deadzone-inter 4 --deadzone-intra 6 --aq-mode 0 --qcomp 0.900000

I would speculate the reason would lie in those settings, though i'm not exactly sure what your cqmfile contains.

[gav1577]

x264.928.modified.exe (http://www.mediafire.com/?yxjg93s1lys) - Alternate Download (http://skystrife.com/x264/x264.928.modified.exe)
libx264-61.928.modified.dll (http://www.mediafire.com/?9xscjgmn2cj) - Alternate Download (http://skystrife.com/x264/libx264-61.928.modified.dll)
(Source) (http://www.mediafire.com/?ud9gdgvdctg) - Alternate Download (http://skystrife.com/x264/x264.928.modified.source.7z)

Patches used:

x264_psyRDO_0.5.diff
x264_hrd_pulldown.09_interlace.diff
x264.progress.indication.01.diff
x264_new_bframe_decision_03.diff <-- This patch is highly experimental, but no longer crashes on my end. Thank you Dark Shikari!

All patches are included with the source.

gcc 3.4.5 fprofiled build.

Hi this build also crashes for me at the beginning of the second pass the odd thing is it does not crash if just doing small clips but will crash if doing a whole movie btw build x264.920.modified.02.exe seems fine

[ajp_anton]

Why does these builds crash on the 2nd pass? The b-frame patch doesn't do anything in the 2nd pass, right?

[LoRd_MuldeR]

Why does these builds crash on the 2nd pass? The b-frame patch doesn't do anything in the 2nd pass, right?

I made a similar experience. With b_adapt=2 it will always crash near the end of the second pass.
But after a comment by Dark Shikari I thought that it's an Avidemux specific problem. Do you get the crash with x264.exe ???

[Dark Shikari]

I've had no issues with --b-adapt 2 on second pass... maybe I haven't tested enough or something? It strikes me as odd...

[elguaxo]

Do you get the crash with x264.exe ???

I ran a couple of small tests and it was fine with --b-adapt 2 (1st and 2nd pass). Now I'm encoding a DVD with some insane settings so it will take some time to see if it crashes on the second pass. :scared: :p

[gav1577]

Does seem odd clips seem fine but whole movies crash must be something to do with changes made after x264.920.modified.02 because that build seems stable so far :) btw i haven't tried dvd only a rip of my bluray movie

[DeeGee]

Same problem here. Build 928 crashes right on the beginning of pass 2, but build 920 seems to be working (only just begun the encoding. ~4 hours to go).

E:\DVD\TGTBTU_D124\VIDEO_TS>x264 --bframes 5 --b-adapt 2 --b-pyramid --ref 5 --d
eblock -2:0 --bitrate 932 --pass 1 --partitions none --direct auto --weightb --m
e dia --subme 1 --b-rdo --psy-rd 1.2 --bime --8x8dct --trellis 0 --sar 64:45 --p
rogress --threads auto --thread-input --output NUL "The Good, The Bad and The Ug
ly.avs"
avis [info]: 704x432 @ 25.00 fps (257037 frames)
x264 [info]: using SAR=64/45
x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 Cache64
x264 [info]: slice I:2060 Avg QP:22.69 size: 28632 PSNR Mean Y:41.53 U:47.83
V:47.99 Avg:42.78 Global:42.00
x264 [info]: slice P:97892 Avg QP:24.83 size: 7824 PSNR Mean Y:39.06 U:46.43
V:46.61 Avg:40.42 Global:39.67
x264 [info]: slice B:157085 Avg QP:26.99 size: 2361 PSNR Mean Y:38.47 U:46.37
V:46.57 Avg:39.88 Global:39.28
x264 [info]: consecutive B-frames: 3.6% 25.6% 55.7% 9.8% 3.6% 1.7%
x264 [info]: mb I I16..4: 11.7% 40.9% 47.4%
x264 [info]: mb P I16..4: 10.8% 0.0% 0.0% P16..4: 71.6% 0.0% 0.0% 0.0% 0
.0% skip:17.7%
x264 [info]: mb B I16..4: 1.6% 0.0% 0.0% B16..8: 19.6% 0.0% 0.0% direct:
15.6% skip:63.3% L0:36.4% L1:46.8% BI:16.8%
x264 [info]: final ratefactor: 23.02
x264 [info]: 8x8 transform intra:5.6% inter:65.2%
x264 [info]: direct mvs spatial:99.9% temporal:0.1%
x264 [info]: ref P L0 56.0% 21.1% 9.5% 6.5% 6.9%
x264 [info]: ref B L0 66.7% 20.0% 8.8% 4.5%
x264 [info]: ref B L1 90.6% 9.4%
x264 [info]: SSIM Mean Y:0.9557316
x264 [info]: PSNR Mean Y:38.723 U:46.406 V:46.598 Avg:40.107 Global:39.442 kb/s:
930.45

encoded 257037 frames, 33.01 fps, 930.52 kb/s

E:\DVD\TGTBTU_D124\VIDEO_TS>x264 --bframes 5 --b-adapt 2 --b-pyramid --ref 5 --d
eblock -2:0 --bitrate 932 --pass 2 --partitions all --direct auto --weightb --me
umh --subme 7 --b-rdo --psy-rd 1.2 --bime --8x8dct --trellis 2 --sar 64:45 --pr
ogress --threads auto --thread-input --output "The Good, The Bad and The Ugly_vi
deo.mp4" "The Good, The Bad and The Ugly.avs"
avis [info]: 704x432 @ 25.00 fps (257037 frames)
x264 [info]: using SAR=64/45
x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 Cache64
--------Crash--------

-----Change to build 920-------
E:\DVD\TGTBTU_D124\VIDEO_TS>x264 --bframes 5 --b-adapt 2 --b-pyramid --ref 5 --d
eblock -2:0 --bitrate 932 --pass 2 --partitions all --direct auto --weightb --me
umh --subme 7 --b-rdo --psy-rd 1.2 --bime --8x8dct --trellis 2 --sar 64:45 --pr
ogress --threads auto --thread-input --output "The Good, The Bad and The Ugly_vi
deo.mp4" "The Good, The Bad and The Ugly.avs"
avis [info]: 704x432 @ 25.00 fps (257037 frames)
x264 [info]: using SAR=64/45
x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 Cache64
mp4 [info]: initial delay 2 (scale 25)
encoded frames: 6426/257037 (2.5%), 18.58 fps, eta 3:44:47

[G_M_C]

My version (skystrife's build) is working fine, with both --b-adapt 2 --weightb in both passes.

It's @ 60% of a 2 hours long film (@1080p) in the second pass. It is quite slow thought, even if i try to get it all @ L4.1 and BR compatabillity (3-bframes, 4 refs etc.), it's only going @ 2,8 ~ 3 fps.

[kemuri-_9]

though i'm using my own build with the same patches, everything is working normally with --b-adapt 2 on both passes.

[Soichiro]

I've never had any crashes either with skystrife's build or my own build.

[elguaxo]

I ran a couple of small tests and it was fine with --b-adapt 2 (1st and 2nd pass). Now I'm encoding a DVD with some insane settings so it will take some time to see if it crashes on the second pass. :scared: :p

no crashes during a DVD encode using --b-adapt 2 on the 1st and 2nd pass! :)

[woah!]

Same problem here. Build 928 crashes right on the beginning of pass 2, but build 920 seems to be working (only just begun the encoding. ~4 hours to go).

E:\DVD\TGTBTU_D124\VIDEO_TS>x264 --bframes 5 --b-adapt 2 --b-pyramid --ref 5 --d
eblock -2:0 --bitrate 932 --pass 1 --partitions none --direct auto --weightb --m
e dia --subme 1 --b-rdo --psy-rd 1.2 --bime --8x8dct --trellis 0 --sar 64:45 --p
rogress --threads auto --thread-input --output NUL "The Good, The Bad and The Ug
ly.avs"
avis [info]: 704x432 @ 25.00 fps (257037 frames)
x264 [info]: using SAR=64/45
x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 Cache64
x264 [info]: slice I:2060 Avg QP:22.69 size: 28632 PSNR Mean Y:41.53 U:47.83
V:47.99 Avg:42.78 Global:42.00
x264 [info]: slice P:97892 Avg QP:24.83 size: 7824 PSNR Mean Y:39.06 U:46.43
V:46.61 Avg:40.42 Global:39.67
x264 [info]: slice B:157085 Avg QP:26.99 size: 2361 PSNR Mean Y:38.47 U:46.37
V:46.57 Avg:39.88 Global:39.28
x264 [info]: consecutive B-frames: 3.6% 25.6% 55.7% 9.8% 3.6% 1.7%
x264 [info]: mb I I16..4: 11.7% 40.9% 47.4%
x264 [info]: mb P I16..4: 10.8% 0.0% 0.0% P16..4: 71.6% 0.0% 0.0% 0.0% 0
.0% skip:17.7%
x264 [info]: mb B I16..4: 1.6% 0.0% 0.0% B16..8: 19.6% 0.0% 0.0% direct:
15.6% skip:63.3% L0:36.4% L1:46.8% BI:16.8%
x264 [info]: final ratefactor: 23.02
x264 [info]: 8x8 transform intra:5.6% inter:65.2%
x264 [info]: direct mvs spatial:99.9% temporal:0.1%
x264 [info]: ref P L0 56.0% 21.1% 9.5% 6.5% 6.9%
x264 [info]: ref B L0 66.7% 20.0% 8.8% 4.5%
x264 [info]: ref B L1 90.6% 9.4%
x264 [info]: SSIM Mean Y:0.9557316
x264 [info]: PSNR Mean Y:38.723 U:46.406 V:46.598 Avg:40.107 Global:39.442 kb/s:
930.45

encoded 257037 frames, 33.01 fps, 930.52 kb/s

E:\DVD\TGTBTU_D124\VIDEO_TS>x264 --bframes 5 --b-adapt 2 --b-pyramid --ref 5 --d
eblock -2:0 --bitrate 932 --pass 2 --partitions all --direct auto --weightb --me
umh --subme 7 --b-rdo --psy-rd 1.2 --bime --8x8dct --trellis 2 --sar 64:45 --pr
ogress --threads auto --thread-input --output "The Good, The Bad and The Ugly_vi
deo.mp4" "The Good, The Bad and The Ugly.avs"
avis [info]: 704x432 @ 25.00 fps (257037 frames)
x264 [info]: using SAR=64/45
x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 Cache64
--------Crash--------

-----Change to build 920-------
E:\DVD\TGTBTU_D124\VIDEO_TS>x264 --bframes 5 --b-adapt 2 --b-pyramid --ref 5 --d
eblock -2:0 --bitrate 932 --pass 2 --partitions all --direct auto --weightb --me
umh --subme 7 --b-rdo --psy-rd 1.2 --bime --8x8dct --trellis 2 --sar 64:45 --pr
ogress --threads auto --thread-input --output "The Good, The Bad and The Ugly_vi
deo.mp4" "The Good, The Bad and The Ugly.avs"
avis [info]: 704x432 @ 25.00 fps (257037 frames)
x264 [info]: using SAR=64/45
x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 Cache64
mp4 [info]: initial delay 2 (scale 25)
encoded frames: 6426/257037 (2.5%), 18.58 fps, eta 3:44:47

long shot but is it a VC1 source? if so are you using ffdshow codec wmv9. if so try using libavcodec instead in the codec list.

[gav1577]

long shot but is it a VC1 source? if so are you using ffdshow codec wmv9. if so try using libavcodec instead in the codec list.

Hi i have the same problem as DeeGee crashes rite at the beginning of the second pass i am using ripbot264 with a modified cmd the source is an bluray 1080p MPEG-2 i will test with megui and ffdshow i think ripbot uses DGindex for MPEG-2. i cant see dgindex being a problem though as 920 works fine :confused:

EDIT: tried megui and ffdshow same result crashed at the beginning of second pass :(. will stick with 920

[DeeGee]

Yep, using dgindex for dvd source. Build 920 finished encoding without problems.

[Sharktooth]

Hi i have the same problem as DeeGee crashes rite at the beginning of the second pass i am using ripbot264 with a modified cmd the source is an bluray 1080p MPEG-2 i will test with megui and ffdshow i think ripbot uses DGindex for MPEG-2. i cant see dgindex being a problem though as 920 works fine :confused:

EDIT: tried megui and ffdshow same result crashed at the beginning of second pass :(. will stick with 920
easy, dont use that patch for "production".
you were already warned.
also if you use megui, keep the x264 build that comes with the autoupdate.

[G_M_C]

My version (skystrife's build) is working fine, with both --b-adapt 2 --weightb in both passes.

It's @ 60% of a 2 hours long film (@1080p) in the second pass. It is quite slow thought, even if i try to get it all @ L4.1 and BR compatabillity (3-bframes, 4 refs etc.), it's only going @ 2,8 ~ 3 fps.


Yep, my encode came out fine too. Quality is very good; I havent been making X264's long, but i'm impressed.

1 small question: Skystrife's build did have --psyRDO on by default didn't it ?

My compliments go out to Dark Shikari, Akupenguin and other developers for all their quality-improving paches !

[martino]

1 small question: Skystrife's build did have --psyRDO on by default didn't it ?
All builds with the patch do, unless someone modifies the source code to do the opposite.

[lexor]

1 small question: Skystrife's build did have --psyRDO on by default didn't it ?

Only if you used --subme 6 or 7.

[LoRd_MuldeR]

Only if you used --subme 6 or 7.

Because Psy RDO cannot work at subme < 6 :p

[lexor]

Because Psy RDO cannot work at subme < 6 :p

Was that irony in attempt to correct, or just a note? I can't tell.

[LoRd_MuldeR]

Was that irony in attempt to correct, or just a note? I can't tell.

You said that Psy RDO is on by default, but only when --subme 6 or 7 is used. Well, that is correct. But it sounds like one could still enforce Psy RDO at subme < 6, it just won't be enabled by default at subme < 6. And that assumption would be wrong. So I wanted to make clear that Psy RDO can never work at subme < 6. That's it...

[G_M_C]

Only if you used --subme 6 or 7.

Well ... i did, so i'm good :p

[kemuri-_9]

x264 --(long)help by default prints out the default options in [] for many parameters for when they are not specified explicitly on the command line, i.e.:

--psy-rd Strength of mode decision psychovisual optimization ["1.000000"]
Does nothing at subme < 6.
so if you ever have questions about if it's on, off, or what the value is, just consult the generated --longhelp

[skystrife]

x264.928.modified.02.exe (http://www.mediafire.com/?b8pgyufbqih) - Alternate Download (http://skystrife.com/x264/x264.928.modified.02.exe)
libx264-61.928.modified.02.dll (http://www.mediafire.com/?babhbhnaaqz) - Alternate Download (http://skystrife.com/x264/libx264-61.928.modified.02.dll)
(Source) (http://www.mediafire.com/?vnxv0iqpnmv) - Alternate Download (http://skystrife.com/x264/x264.928.modified.02.source.7z)

Patches used:

x264_psyRDO_0.5+psytrellis.diff <-- This patch is experimental. See the PsyRDO thread for more details.
x264_new_bframe_decision_03.diff <-- This patch is highly experimental.
x264_hrd_pulldown.09_interlace.diff
x264.progress.indication.01.diff

All patches are included with the source.

gcc 3.4.5 fprofiled build.

[gav1577]

Can someone please compile the new trellis patch in a regular build
without the bframe_decision patch for testing Thanks :)

[Dark Shikari]

Can someone please compile the new trellis patch in a regular build
without the bframe_decision patch for testing Thanks :)You know, you can just use --b-adapt 1 for the regular B-frame decision ;). In fact, I think its default.

[shon3i]

@DS, i realy confused how to turn on psytrellis? what switch?