[HEVC] x265.EXE: mingw builds

[vhelp]

as for the sorting idea, unless he rebuilds the date column (so that he can sort by that) if sorting by the ver number, then he would have take into account the:

0.5+1
0.5+23
0.5+240

and pad the numbers with zero's to turn it into this and put into a separate (temporary) column when sorting the list initially. then he can add each new version manually.

0.5+001
0.5+023
0.5+240

. . .

i recently added duration to my gui to track the encoding completion times. and after testing it, i could see there is progress in the area of "speed" with some builds going back to { 0.5+243 } to present. tested on 100 frames for each build, nearly 10 second gain from that +243 build. i didn't go that far back.

for others building gui's for distribution, it may benefit to add a duration component (a simple textbox) to it so all can track that aspect of the encoder's progress for their particular OS and SYS build. that may come in handy when discussion speed in different situations and so on.

[LigH.de]

Well, your thoughts are in general right, but ... you chose 3 decimal places for the patch-per-version number. But what if there are more than 999 patches? Then you will need 4 decimal places... Of course, theoretically. But practice proves that such unfortunate things do happen.

[vhelp]

i chose the 3-digit to fit the context of what is currently in use at this time, though 4-digit will probably become the norm later, agreed. you and i know both know that it will go into the 4-digit sooner or later unless they do something different yet again. i.e, they were using 0.4+xxx and then 0.4.1+xxx and now 0.5+xxx. so anything is possible. we just have to wait and see and adjust accordingly and hope that things stay consitantly from that point onward.

[chie]

when i compile with intel c++ in windows server 2008 r2:

NMAKE : fatal error U1073: don't know how to make 'common\CMakeFiles\common.dir\
x86\asm-primitives.cpp.obj'
Stop.
NMAKE : fatal error U1077: '"C:\Program Files (x86)\Microsoft Visual Studio 11.0
\VC\BIN\nmake.exe"' : return code '0x2'
Stop.
NMAKE : fatal error U1077: '"C:\Program Files (x86)\Microsoft Visual Studio 11.0
\VC\BIN\nmake.exe"' : return code '0x2'
Stop.

anyone can help me :\ ??

[El Heggunte]

when i compile with intel c++ in windows server 2008 r2:

NMAKE : fatal error U1073: don't know how to make 'common\CMakeFiles\common.dir\
x86\asm-primitives.cpp.obj'
Stop.
NMAKE : fatal error U1077: '"C:\Program Files (x86)\Microsoft Visual Studio 11.0
\VC\BIN\nmake.exe"' : return code '0x2'
Stop.
NMAKE : fatal error U1077: '"C:\Program Files (x86)\Microsoft Visual Studio 11.0
\VC\BIN\nmake.exe"' : return code '0x2'
Stop.

anyone can help me :\ ??

Yes, Steve Borho can help you, just send a copy of the error message(s) to

steve@borho.org

[-t40-]

Hello, excuse me if it's too far from the topic but today VPlayer for Android added "Support H265/HEVC and VP9 decoding video".
Didn't tested that but seems like the future is coming closer

[vhelp]

for the entire list, the Consolas font, (8 or 9 size) and custom date yyyy-mm-dd, hh:mm A/P looks good.

[x265.cc]

As requested/suggested, i've made some buildbot changes on https://x265.cc:

Webserver:

-fresh nginx recompile
-different security (firewall, etc.) & performance improvements
-changed the directory listing (e.g. https://x265.cc/builds/MSYS-x86/8bpp/)
-added MSYS-x86_64 folder structure

Buildbot:

-minor performance/stability improvements
-changed 3rd party mingw suite to an own one
-added "MSYS/MinGW GCC4.8.2 x64" builds

[El Heggunte]

As requested/suggested, i've made some buildbot changes on https://x265.cc:

Webserver:

-fresh nginx recompile
-different security (firewall, etc.) & performance improvements
-changed the directory listing (e.g. https://x265.cc/builds/MSYS-x86/8bpp/)
-added MSYS-x86_64 folder structure

Buildbot:

-minor performance/stability improvements
-changed 3rd party mingw suite to an own one
-added "MSYS/MinGW GCC4.8.2 x64" builds

and

[vhelp]

- for advanced people -

source_video.avs -> output_hevc.hm10

first, let me go on to say, that prior to encoding the video to hevc format, you will want to create your script similarly like this, using a subtitle string to identify your project and its specs. open notepad.exe and type in the following text and save it as "source_video.avs" file. this is for people who need control of their work and progress. once that is no longer needed, you can remove some of the code and just encode your videos the way your normally do. also, you can remove the comments at anytime.

file: "source_video.avs"

  v = "source_video.ts"
  LoadPlugin(  "C:\PLUGINS\LSMASHSource.dll" )
# 
# the method below is the fastest way i know how to open a video inside virtualdub, or use whatever
# methods you know works for you.
# 
  LWLibavVideoSource(v, cache=true , seek_mode=0 )#.selectevery(2,0) # <-- un-remark this if you get 54.94 fps
# 
# <your favorite ivtc filter goes here>
#
# **
# or, either one of the selectevery way below, starting at the first pppii. i prefere the selectevery method and 
# build many for all my test videos. so to help my eyes and editing, i aline each command and values by padding
# with spaces. but that is how i work. the first trim() is where i start my pppii at the first "p" position. i find this
# position in virtualdub while scaning bw/fw a few times. then, once i have that and run it through virtualdub,
# i will scrub through the video up to the end where i want to stop the encode and place that value in the 
# last trim() section. or skip or snip this section completely, between the **
# 
# trim(0000,0000).separateFields().selectevery(10, 0,1,2,3, 4,5, 6,9).weave( ).trim(0, 0100)
# trim(0000,0000).separateFields().selectevery( 5, 0,1,2,3          ).weave( ).trim(0, 0100)
# ** 
#
# next comes the subtitle text to identify your video and specs. 
# 
  ScriptClip( """ subtitle("hevc video, 0.5+547 build, "  + string(current_frame) + " - " + string(framecount),12,12)""")
  ScriptClip( """ subtitle("params: ultrafast, crf17,  "  + string(current_frame) + " - " + string(framecount),12,12)""")
#
# if you prefer to trim() your videos here, i sometimes do, you can enter your trim(s) here.
# 
  trim( 0,0 )

then, for anyone looking for a way to analize their hevc encoded videos, here are a few methods of doing so. although i did post it in another thread topic, elsewhere, i'll repeat some of it here.

ffmpeg -i g:\1440x1080p.H264.m2ts -ss 00:00:00 -ss 00:00:04 -frames:v 24 -r 23.976 -y im_%4d.png

detail explanation for Method A

-r value is the vidoe's frame rate. make sure you use the correct setting the video(s) you encode.

the -ss value in the 00:00:00 format, is the seek method into the video. it is the timecode of the video. you can get this detail from virtualdub, if you open a video using an avisynth script. in the above example, i use two -ss commands. there are advantanges** of where to use -ss, before the -i video.hm10 and after. this affects the accuracy (the distance between i-frames) and/or speed, depending on the placement. experimentation is your best friend. but for general extracting, it doesn't matter much. if you exclude the -ss (or -t, which is the duration or length you want to extract when using that command parameter) you will only get the first frame and no more.

** consult the ffmpeg documenation for further details.

until you no longer need the following step: i recommend adding the frame number as a subtitle into the video as a gauge to help extract specific frames. once you're satisified with the results/review/analizing, you can remove it. this is how i use it in an avisynth script, be sure to add it as the last line in your script. i usually put it at the top of the video:

ScriptClip( """ subtitle("hevc video, 0.5+547 build, "  + string(current_frame) + " - " + string(framecount),12,12)""")

the %4d is the length of the padding for zeros sequentually, or else the sorting won't work correctly when opening/viewing the images in a folder or video timeline.

the -frames:v value, is how many frames you want to extract from that -ss position and onward, up to the end of the video. in the above example, i am extracting 24 frames, i.e., im_0001 im_0002 im_0003, ... im_0024.png

note: if you add the .png at the end of the filename, ffmpeg will understand you want the .png format and convert to it automatically.

detail explanation for Method B

video-only method for image extract or timeline (virtualdub) review. obviously, you would handle your audio separately.

using notepad.exe, type in and save this to a file, "output_video.avs" and then open it inside virtualdub. depending on the length of the video, it may take some time to parse and finally open for the first time. subsequent re-opening of the same file will be less time. you can change the subtitle text for each video you are encoding.

file: "output_video.avs"

  v = "g:\output_video.hm10"
  LoadPlugin( "C:\PLUGINS\LSMASHSource.dll" )
  LWLibavVideoSource(v, cache=true , seek_mode=0 )#.selectevery(2,0) # <-- un-remark this if you get 54.94 fps
# trim( 0,0 ) # <-- if you want to extract a certain section.

this avisynth script will open inside virtualdub, giving you the ability to review the whole clip more easily. from virtualdub, you can use the Ctrl+1 function to copy the current image to the clipboard and paste it into your analizing application for further detail review/analysis/comparing/etc.

you can also use the script as-is in the method a example for ffmpeg.

thats it. the rest you can chew on and change to however way you work. its just to get you started on your way.

[jb_alvarado]

Hello everybody,
I want to include x265 to my automatic build script:
https://github.com/jb-alvarado/media-autobuild_suite

As compiler I use mingw32 and mingw64 4.8.2. To compile x265 is easy, but the problem is that x265 link to libgcc-*.dll. Is there a way to pass "-static -static-libgcc -static-libstdc++" to the compiler?

[El Heggunte]

......

As compiler I use mingw32 and mingw64 4.8.2. To compile x265 is easy, but the problem is that x265 link to libgcc-*.dll. Is there a way to pass "-static -static-libgcc -static-libstdc++" to the compiler?

Hi there,

that's a VERY non-trivial question, and you have two options,

a) wait until Tom Vaughan chimes in this thread ,

OR

b) send an e-mail to "Steve Borho" <steve@borho.org>

As you probably already know, x265 was designed to orbit around MS Visual C++ and to depend on CMake , therefore, ...

[jb_alvarado]

Hello El Heggunte,
thank you for your answer! I will wait a bit, maybe Tom will read this question. I use cmake to build my msys makefile. Normally I found in makefiles some ldflags, cflags or cxxflags where I can put my -static-libgcc command, but in the makefile from x265 I don't found this flags.

[LigH.de]

Apropos Tom ... from the doom9 forum:

I've posted an updated version of the x265 Evaluator's Guide...

x265 Evaluators Guide Nov 24 2013.pdf

Feedback is welcomed.

Tom

[Selur]

@LigH: "Perhaps it's time to open a fresh thread, not focused specifically on MinGW builds."
-> https://forum.videohelp.com/threads/360069-x265-HEVC-Encoder

[x265]

b) send an e-mail to ...

Not only are you not helping the OP, you are pointing people in the wrong direction. Please stop.

A lead architect is not a technical support person. To help the x265 project develop the capabilities that you are looking for as quickly as possible, we ask everyone to respect the ways in which we are set up to provide support. If you email our lead developer you will either get no answer, or you will get an answer like the one below.

To get support for x265...

• These forums (and Doom9's forums) are a good place to ask questions and get answers. I think that was the premise behind the name VideoHELP, anyhow.
• If you think you discovered a bug, our x265-devel mailing list is the right place to report it. Please follow the bug reporting guidelines in the x265 Evaluator's Guide.

Thanks!
Tom

[easyfab]

Hello El Heggunte,
thank you for your answer! I will wait a bit, maybe Tom will read this question. I use cmake to build my msys makefile. Normally I found in makefiles some ldflags, cflags or cxxflags where I can put my -static-libgcc command, but in the makefile from x265 I don't found this flags.

I don't know if it could help :

create a file toolchain.cmake with those lines (adapt to your system, add more commands ... ) :

SET(CMAKE_SYSTEM_NAME Windows)
SET(CMAKE_C_COMPILER i686-w64-mingw32-gcc -static-libgcc)
SET(CMAKE_CXX_COMPILER i686-w64-mingw32-g++ -static-libgcc)
SET(CMAKE_RC_COMPILER windres)
SET(CMAKE_ASM_YASM_COMPILER yasm)

and then run something like this :

cmake -G "Unix Makefiles" -DCMAKE_TOOLCHAIN_FILE=toolchain.cmake ../../source

[El Heggunte]

b) send an e-mail to steve@borho.org

Not only are you not helping the OP, you are pointing people in the wrong direction. Please stop.

I gladly will "stop", on the day that you start acting more like a technically-skilled person and less like a marketer.

As one can easily verify, most (if not ALL) "solutions" regarding the x265 problems are "set" by Steve Borho himself.

Also, steve@borho.org is the address where the mailing-list bot sends messages to, therefore I don't see your point.

Besides, before joining the mailing list, I did send around ten e-mails to Steve, and his answers always were very-helpful and straight-to-the-point. Fortunately he doesn't seem to care about bureaucratic rituals.

[x265]

b) send an e-mail to steve@borho.org

Not only are you not helping the OP, you are pointing people in the wrong direction. Please stop.

I gladly will "stop", on the day that you start acting more like a technically-skilled person and less like a marketer.

As one can easily verify, most (if not ALL) "solutions" regarding the x265 problems are "set" by Steve Borho himself.

Also, steve@borho.org is the address where the mailing-list bot sends messages to, therefore I don't see your point.

Besides, before joining the mailing list, I did send around ten e-mails to Steve, and his answers always were very-helpful and straight-to-the-point. Fortunately he doesn't seem to care about bureaucratic rituals.

Would you like us to publish your email address on a public forum? That would be a classless thing to do, as you should be the one to decide where and when you want to give out your email address.

When it comes to how the x265 project supports our customers and open source adopters, I would request that people respect our need to organize and work efficiently. I remain hopeful that you will want to work with us, and not fight with us.

Thanks
Tom

[Selur]

@x265: regarding support:
- Since the bug tracker (https://bitbucket.org/multicoreware/x265/issues) doesn't seem to be monitored by anybody it might be a good idea to disable it.
- Do you know if there are any plans to support other Mac OS X versions than Mavericks? (compiled x265 on Mavericks without a problem, compiling on Snow Leopard and Lion fails; thing is stuff compiled on Snow Leopard runs on Mavericks, but not the other way around )

[jb_alvarado]

Hello El Heggunte,
thank you for your answer! I will wait a bit, maybe Tom will read this question. I use cmake to build my msys makefile. Normally I found in makefiles some ldflags, cflags or cxxflags where I can put my -static-libgcc command, but in the makefile from x265 I don't found this flags.

I don't know if it could help :

create a file toolchain.cmake with those lines (adapt to your system, add more commands ... ) :

SET(CMAKE_SYSTEM_NAME Windows)
SET(CMAKE_C_COMPILER i686-w64-mingw32-gcc -static-libgcc)
SET(CMAKE_CXX_COMPILER i686-w64-mingw32-g++ -static-libgcc)
SET(CMAKE_RC_COMPILER windres)
SET(CMAKE_ASM_YASM_COMPILER yasm)

and then run something like this :

cmake -G "Unix Makefiles" -DCMAKE_TOOLCHAIN_FILE=toolchain.cmake ../../source

Thank you easyfab, that was helping allot! I only change i686-w64-mingw32-gcc to gcc and the other to g++ and run cmake with -G "MSYS Makefile" [...] and then it works great!
In the next days I will update my script with that on github, maybe someone is interested in that.

[Selur]

in the next days i will update my script with that on github, maybe someone is interested in that.

[vhelp]

0.5+548 was the last stable build. the newer ones from that would shut down the encoder. the good news... the latest 0.5+600 build is stable again

[Marchand]

0.5+548 was the last stable build. the newer ones from that would shut down the encoder. the good news... the latest 0.5+600 build is stable again

Vhelp.......You're right..... I tested both versions and are stable. 0.5+548 and 0.5+600..... See below the 3 tests:

------------------------------------------------------------------------------------------------------------------------------
[C:\Users\Marchand\Videos\teste 4.mp4]
------------------------------------------------------------------------------------------------------------------------------
auo [info]: converting YUY2 -> i420p, using SSE2
auo [info]: x265 options...
--input-csp i420 --preset ultrafast --crf 28 --rc-lookahead 20 --keyint 240 --bframes 3 --b-adapt 1 --ctu 64 --me 1
--merange 60 --ref 2 --frames 179 --input-res 1920x1080 --fps 24 -o "C:\Users\Marchand\Videos\teste 4.265" "-"
yuv [info]: 1920x1080 24Hz C420, frames 0 - 178 of 0
x265 [info]: using cpu capabilities: MMX2 SSE SSE2Fast SSSE3 SSE4.1 Cache64
x265 [info]: HEVC encoder version 0.5+600-ec904fab863a
x265 [info]: build info [Windows][MSVC 1800][32 bit] 8bpp
x265 [info]: Main profile, Level-4 (Main tier)
x265 [info]: WPP streams / pool / frames : 17 / 4 / 2
x265 [info]: CU size : 64
x265 [info]: Max RQT depth inter / intra : 1 / 1
x265 [info]: ME / range / subpel / merge : hex / 60 / 0 / 2
x265 [info]: Keyframe min / max : 240 / 240
x265 [info]: Rate Control : CRF-28.000000
x265 [info]: Lookahead / bframes / badapt : 20 / 3 / 1
x265 [info]: b-pyramid / weightp / ref : 1 / 0 / 2
x265 [info]: tools: cfm esd rd=0
x265 [info]: I-frame 64..4: 66% 9% 12% 9% 2%
x265 [info]: P64: Split 52% Merge 18% Inter 17% Intra 10%(DC 27% P 12% Ang 59%)
x265 [info]: P32: Split 32% Merge 38% Inter 15% Intra 13%(DC 22% P 16% Ang 60%)
x265 [info]: P16: Split 18% Merge 41% Inter 20% Intra 20%(DC 21% P 14% Ang 64%)
x265 [info]: P8 : Split 0% Merge 38% Inter 30% Intra 31%(DC 20% P 13% Ang 65% 4x4 0%)
x265 [info]: B64: Split 8% Merge 85% Inter 5%
x265 [info]: B32: Split 9% Merge 81% Inter 8%
x265 [info]: B16: Split 9% Merge 74% Inter 16%
x265 [info]: B8 : Split 0% Merge 71% Inter 28%
x265 [info]: frame I: 3 kb/s: 2995.39 PSNR Mean: Y:81.593 U:78.656 V:51.529
x265 [info]: frame P: 112 kb/s: 1928.41 PSNR Mean: Y:44.256 U:44.700 V:45.427
x265 [info]: frame B: 64 kb/s: 199.11 PSNR Mean: Y:43.043 U:43.475 V:44.652
x265 [info]: global : 179 kb/s: 1327.99 PSNR Mean: Y:44.448 U:44.831 V:45.252
auo [info]: x265?G?Ѓg?R?[?h???„U : 0???„U 1?? 7.1?b
auo [info]: NeroAacEnc ?„EЃс‡‚???G?Ѓg?R?[?h?„‚?s???„]?ЃE?B AAC-LC ?r?b?g???[?g?wЃf„y, 128kbps
auo [info]: L-SMASH muxer ?„Emux?„‚?s???„]?ЃE?BЃсfЃe?: on, Ѓс‡‚??ff, tcff, ?gЃf????[?h„I?ѓК
auo [info]: L-SMASH remuxer ?„Emux?„‚?s???„]?ЃE?BЃсfЃe?: on, Ѓс‡‚??: on, tcff, ?gЃf????[?h„I?ѓК
auo [info]: Ѓe??G?Ѓg?R?[?h???„U : 0???„U 1?? 9.3?b

2013”N11ЊЋ27“ъ 22Ћћ49•Є ѓGѓ“ѓRЃ[ѓhЏI—№
-------------------------------------------------------------------------------------

------------------------------------------------------------------------------------------------------------------------------
[C:\Users\Marchand\Videos\teste 5.mp4]
------------------------------------------------------------------------------------------------------------------------------
auo [info]: converting YUY2 -> i420p, using SSE2
auo [info]: x265 options...
--input-csp i420 --preset ultrafast --crf 28 --rc-lookahead 20 --keyint 240 --bframes 3 --b-adapt 1 --ctu 64 --me 1
--merange 60 --ref 2 --frames 179 --input-res 1920x1080 --fps 24 -o "C:\Users\Marchand\Videos\teste 5.265" "-"
yuv [info]: 1920x1080 24Hz C420, frames 0 - 178 of 0
x265 [info]: using cpu capabilities: MMX2 SSE SSE2Fast SSSE3 SSE4.1 Cache64
x265 [info]: HEVC encoder version 0.5+600-ec904fab863a
x265 [info]: build info [Windows][GCC 4.8.2][32 bit] 16bpp
x265 [info]: Main profile, Level-4 (Main tier)
x265 [info]: WPP streams / pool / frames : 17 / 4 / 2
x265 [info]: Input bit depth : 8
x265 [info]: CU size : 64
x265 [info]: Max RQT depth inter / intra : 1 / 1
x265 [info]: ME / range / subpel / merge : hex / 60 / 0 / 2
x265 [info]: Keyframe min / max : 240 / 240
x265 [info]: Rate Control : CRF-28.000000
x265 [info]: Lookahead / bframes / badapt : 20 / 3 / 1
x265 [info]: b-pyramid / weightp / ref : 1 / 0 / 2
x265 [info]: tools: cfm esd rd=0
x265 [info]: I-frame 64..4: 68% 6% 11% 11% 1%
x265 [info]: P64: Split 49% Merge 14% Inter 18% Intra 17%(DC 27% P 12% Ang 59%)
x265 [info]: P32: Split 31% Merge 29% Inter 17% Intra 20%(DC 21% P 18% Ang 60%)
x265 [info]: P16: Split 16% Merge 34% Inter 20% Intra 28%(DC 21% P 14% Ang 63%)
x265 [info]: P8 : Split 0% Merge 32% Inter 29% Intra 38%(DC 20% P 14% Ang 65% 4x4 0%)
x265 [info]: B64: Split 12% Merge 77% Inter 9%
x265 [info]: B32: Split 12% Merge 73% Inter 14%
x265 [info]: B16: Split 9% Merge 69% Inter 21%
x265 [info]: B8 : Split 0% Merge 63% Inter 36%
x265 [info]: frame I: 3 kb/s: 1606.46 PSNR Mean: Y:82.910 U:79.815 V:52.816
x265 [info]: frame P: 70 kb/s: 2200.84 PSNR Mean: Y:44.352 U:44.550 V:45.100
x265 [info]: frame B: 106 kb/s: 315.64 PSNR Mean: Y:42.812 U:43.297 V:44.440
x265 [info]: global : 179 kb/s: 1074.50 PSNR Mean: Y:44.086 U:44.399 V:44.838
auo [info]: x265?G?Ѓg?R?[?h???„U : 0???„U 1??51.5?b
auo [info]: NeroAacEnc ?„EЃс‡‚???G?Ѓg?R?[?h?„‚?s???„]?ЃE?B AAC-LC ?r?b?g???[?g?wЃf„y, 128kbps
auo [info]: L-SMASH muxer ?„Emux?„‚?s???„]?ЃE?BЃсfЃe?: on, Ѓс‡‚??ff, tcff, ?gЃf????[?h„I?ѓК
auo [info]: L-SMASH remuxer ?„Emux?„‚?s???„]?ЃE?BЃсfЃe?: on, Ѓс‡‚??: on, tcff, ?gЃf????[?h„I?ѓК
auo [info]: Ѓe??G?Ѓg?R?[?h???„U : 0???„U 1??53.3?b

2013”N11ЊЋ27“ъ 22Ћћ54•Є ѓGѓ“ѓRЃ[ѓhЏI—№
-------------------------------------------------------------------------------------

------------------------------------------------------------------------------------------------------------------------------
[C:\Users\Marchand\Videos\teste 6.mp4]
------------------------------------------------------------------------------------------------------------------------------
auo [info]: converting YUY2 -> i420p, using SSE2
auo [info]: x265 options...
--input-csp i420 --preset ultrafast --crf 28 --rc-lookahead 20 --keyint 240 --bframes 3 --b-adapt 1 --ctu 64 --me 1
--merange 60 --ref 2 --frames 179 --input-res 1920x1080 --fps 24 -o "C:\Users\Marchand\Videos\teste 6.265" "-"
yuv [info]: 1920x1080 24Hz C420, frames 0 - 178 of 0
x265 [info]: using cpu capabilities: MMX2 SSE SSE2Fast SSSE3 SSE4.1 Cache64
x265 [info]: HEVC encoder version 0.5+600-ec904fab863a
x265 [info]: build info [Windows][MSVC 1800][32 bit] 16bpp
x265 [info]: Main profile, Level-4 (Main tier)
x265 [info]: WPP streams / pool / frames : 17 / 4 / 2
x265 [info]: Input bit depth : 8
x265 [info]: CU size : 64
x265 [info]: Max RQT depth inter / intra : 1 / 1
x265 [info]: ME / range / subpel / merge : hex / 60 / 0 / 2
x265 [info]: Keyframe min / max : 240 / 240
x265 [info]: Rate Control : CRF-28.000000
x265 [info]: Lookahead / bframes / badapt : 20 / 3 / 1
x265 [info]: b-pyramid / weightp / ref : 1 / 0 / 2
x265 [info]: tools: cfm esd rd=0
x265 [info]: I-frame 64..4: 66% 8% 11% 11% 1%
x265 [info]: P64: Split 48% Merge 14% Inter 18% Intra 17%(DC 27% P 12% Ang 60%)
x265 [info]: P32: Split 31% Merge 29% Inter 17% Intra 21%(DC 21% P 18% Ang 60%)
x265 [info]: P16: Split 16% Merge 34% Inter 20% Intra 28%(DC 21% P 15% Ang 63%)
x265 [info]: P8 : Split 0% Merge 31% Inter 28% Intra 39%(DC 20% P 14% Ang 64% 4x4 0%)
x265 [info]: B64: Split 12% Merge 77% Inter 9%
x265 [info]: B32: Split 11% Merge 74% Inter 14%
x265 [info]: B16: Split 9% Merge 69% Inter 21%
x265 [info]: B8 : Split 0% Merge 63% Inter 36%
x265 [info]: frame I: 3 kb/s: 2132.29 PSNR Mean: Y:82.207 U:79.328 V:52.273
x265 [info]: frame P: 71 kb/s: 2176.21 PSNR Mean: Y:44.382 U:44.582 V:45.141
x265 [info]: frame B: 105 kb/s: 308.57 PSNR Mean: Y:42.813 U:43.294 V:44.463
x265 [info]: global : 179 kb/s: 1079.93 PSNR Mean: Y:44.095 U:44.409 V:44.863
auo [info]: x265?G?Ѓg?R?[?h???„U : 0???„U 2??13.2?b
auo [info]: NeroAacEnc ?„EЃс‡‚???G?Ѓg?R?[?h?„‚?s???„]?ЃE?B AAC-LC ?r?b?g???[?g?wЃf„y, 128kbps
auo [info]: L-SMASH muxer ?„Emux?„‚?s???„]?ЃE?BЃсfЃe?: on, Ѓс‡‚??ff, tcff, ?gЃf????[?h„I?ѓК
auo [info]: L-SMASH remuxer ?„Emux?„‚?s???„]?ЃE?BЃсfЃe?: on, Ѓс‡‚??: on, tcff, ?gЃf????[?h„I?ѓК
auo [info]: Ѓe??G?Ѓg?R?[?h???„U : 0???„U 2??14.6?b

2013”N11ЊЋ27“ъ 22Ћћ57•Є ѓGѓ“ѓRЃ[ѓhЏI—№
-------------------------------------------------------------------------------------

[dimokami]

Please, someone can tell. Where or when 10bit version of x265 can to download for tests?

[El Heggunte]

Please, someone can tell. Where or when 10bit version of x265 can to download for tests?

No such thing as "10-bit version" of x265. So far, all we have is 8bpp builds and 16bpp builds:

https://x265.cc/builds/

[x265]

Please, someone can tell. Where or when 10bit version of x265 can to download for tests?

You can build the 16 bit version of x265 simply by checking the HIGH_BIT_DEPTH option in CMAKE. The 16 bit version supports 10 or 12 bit /sample video.

Tom

[dimokami]

Unofficial builds 16bit from x265.cc can encode in 10 or 12 bit. Or I need to build new one?

[x265.cc]

Or I need to build new one?

You do not need to build it yourself.

It does not matter if you use the builds from x265.cc or build it yourself.

Personally, i use the builds from x265.cc too, because i'm to lazy to compile every new version on my pc.

Unofficial builds 16bit from x265.cc can encode in 10 or 12 bit.

--input-depth Bit-depth of input file and internal encoder bit depth. Default 8

Anyway, using the 16bpp build like megui is currently a bad idea.

[dimokami]

I try to do this with this key --input-depth, but output video were cracked. I'am doing something wrong or this function does not work? My command line

--bitrate 2000 --input-depth 10 --subme 7 --rdpenalty 1 --no-strong-intra-smoothing --b-adapt 2 --weightp --aq-mode 1 --rd 2 --sao-lcu-bounds 1 --sao-lcu-opt 0