Re: [x265] Error x265 2.9+1: frame type not compatible with keyframe interval

2019-03-20 Thread Mario *LigH* Rohkrämer

Hi.

Just a guess: Are you possibly reusing a 1st-pass statistics file from a 
different encode, or did you change parameters between 1st and 2nd pass 
which affect the GOP size limits or scene detection sensitivity?


Just to exclude the most obvious first, before digging deeper if there 
was a known GOP type bug for this version specifically.


Posting the whole command lines for both passes would certainly be helpful.


Zero Un schrieb am 17.03.2019 um 13:17:

Hello i have this mistake many times on a encode

x265 [warning]: specified frame type (5) at 77330 is not compatible with 
keyframe interval

x265 [error]: slice=I but 2pass stats say B


___
x265-devel mailing list
x265-devel@videolan.org
https://mailman.videolan.org/listinfo/x265-devel


--

Fun and success!

Mario *LigH* Rohkrämer
maito:cont...@ligh.de
___
x265-devel mailing list
x265-devel@videolan.org
https://mailman.videolan.org/listinfo/x265-devel


[x265] Error x265 2.9+1

2019-03-20 Thread Zero Un
Hello i have this mistake many times on a encode

x265 [warning]: specified frame type (5) at 77330 is not compatible with
keyframe interval
x265 [error]: slice=I but 2pass stats say B
___
x265-devel mailing list
x265-devel@videolan.org
https://mailman.videolan.org/listinfo/x265-devel


[x265] Issue #482: Leak Memory (multicoreware/x265)

2019-03-20 Thread Hữu Quang Linh Lê
New issue 482: Leak Memory
https://bitbucket.org/multicoreware/x265/issues/482/leak-memory

Hữu Quang Linh Lê:

I found some input causing leak memory. This is crash information.
```
INFO: Seed: 3904123746
INFO: Loaded 1 modules   (50397 inline 8-bit counters): 50397 [0x1267560, 
0x1273a3d),
INFO: Loaded 1 PC tables (50397 PCs): 50397 [0x1273a40,0x1338810),
./encoder-fuzzer: Running 1 inputs 1 time(s) each.
Running: leak-034018c6753ae7d385399bf5c3071ba3863a95c8

=
==30238==ERROR: LeakSanitizer: detected memory leaks

Direct leak of 984 byte(s) in 1 object(s) allocated from:
#0 0x4b2f4e in posix_memalign 
/src/llvm/projects/compiler-rt/lib/asan/asan_malloc_linux.cc:218
#1 0xc2f80c in x265::x265_malloc(unsigned long) 
/src/x265/source/common/common.cpp:81:9
#2 0xc302d9 in x265_param_alloc /src/x265/source/common/param.cpp:95:25
#3 0x7f059d in x265_encoder_open_169 /src/x265/source/encoder/api.cpp:97:37
#4 0x6082d0 in x265_encode_image(void*, heif_image const*, 
heif_image_input_class) /src/libheif/libheif/heif_encoder_x265.cc:707:22
#5 0x5d52d4 in 
heif::HeifContext::Image::encode_image_as_hevc(std::__1::shared_ptr,
 heif_encoder*, heif_encoding_options const*, heif_image_input_class) 
/src/libheif/libheif/heif_context.cc:1700:27
#6 0x5d4791 in 
heif::HeifContext::encode_image(std::__1::shared_ptr, 
heif_encoder*, heif_encoding_options const*, heif_image_input_class, 
std::__1::shared_ptr&) 
/src/libheif/libheif/heif_context.cc:1608:28
#7 0x5b8f47 in heif_context_encode_image 
/src/libheif/libheif/heif.cc:1561:25
#8 0x60bfa8 in LLVMFuzzerTestOneInput 
/src/libheif/libheif/encoder_fuzzer.cc:161:9
#9 0x6465f5 in fuzzer::Fuzzer::ExecuteCallback(unsigned char const*, 
unsigned long) /src/libfuzzer/FuzzerLoop.cpp:529:15
#10 0x60e076 in fuzzer::RunOneTest(fuzzer::Fuzzer*, char const*, unsigned 
long) /src/libfuzzer/FuzzerDriver.cpp:286:6
#11 0x619ba3 in fuzzer::FuzzerDriver(int*, char***, int (*)(unsigned char 
const*, unsigned long)) /src/libfuzzer/FuzzerDriver.cpp:715:9
#12 0x60d6ec in main /src/libfuzzer/FuzzerMain.cpp:19:10
#13 0x7f9cd365582f in __libc_start_main 
(/lib/x86_64-linux-gnu/libc.so.6+0x2082f)

SUMMARY: AddressSanitizer: 984 byte(s) leaked in 1 allocation(s).
```
Do you think this is a bug? I want to report to you so that I can fix this soon.


___
x265-devel mailing list
x265-devel@videolan.org
https://mailman.videolan.org/listinfo/x265-devel


[x265] Blocking on edges in most recent x265 commit

2019-03-20 Thread Orth, Blake
I just ran some short encodes with the most recent version of x265 from the 
mercurial repo and I'm seeing some odd blocking happening at hard edges with 
motion.  I reverted to changeset:   12496:d97a256d46a7 (x86 primitive for 
ssim-rd) and the issue does not seem to be present in this version.  I haven't 
had a chance to do a deep analysis of the video yet, but the blocks that are 
erroneous look large - probably an entire 64x64 CTU.

-Blake
___
x265-devel mailing list
x265-devel@videolan.org
https://mailman.videolan.org/listinfo/x265-devel


[x265] Issue #481: Bug reporting version 3.0 not selectable (multicoreware/x265)

2019-03-20 Thread flip -
New issue 481: Bug reporting version 3.0 not selectable
https://bitbucket.org/multicoreware/x265/issues/481/bug-reporting-version-30-not-selectable

flip -:

When filing a bug report there is a dropdown box with some versions to select. 
This box only goes up to version 1.9 thus i can not select the latest 3.0 
version


___
x265-devel mailing list
x265-devel@videolan.org
https://mailman.videolan.org/listinfo/x265-devel


[x265] Issue #480: Compiler warnings on x265 3.0 (multicoreware/x265)

2019-03-20 Thread flip -
New issue 480: Compiler warnings on x265 3.0
https://bitbucket.org/multicoreware/x265/issues/480/compiler-warnings-on-x265-30

flip -:

During compilation of handbrake, the following warnings/hints appear:

```
  CC   src/libbluray/bdnav/libbluray_la-uo_mask.lo
/home/flip/programs/src/HandBrake/build/contrib/x265/x265_3.0/source/encoder/ratecontrol.cpp:
 In member function ‘int 
x265::RateControl::writeRateControlFrameStats(x265::Frame*, 
x265::RateControlEntry*)’:
/home/flip/programs/src/HandBrake/build/contrib/x265/x265_3.0/source/encoder/ratecontrol.cpp:2864:21:
 warning: passing argument 1 to restrict-qualified parameter aliases with 
argument 3 [-Wrestrict]
 sprintf(deltaPOC, "%s%d~", deltaPOC, rpsWriter->deltaPOC[i]);
 ^~~~   
/home/flip/programs/src/HandBrake/build/contrib/x265/x265_3.0/source/encoder/ratecontrol.cpp:2865:21:
 warning: passing argument 1 to restrict-qualified parameter aliases with 
argument 3 [-Wrestrict]
 sprintf(bUsed, "%s%d~", bUsed, rpsWriter->bUsed[i]);
 ^   ~
  CC   src/libbluray/decoders/libbluray_la-graphics_controller.lo
  CXX  libAACdec/src/aacdec_hcr_bit.lo
  CXX  libAACdec/src/aacdec_hcrs.lo
/home/flip/programs/src/HandBrake/build/contrib/x265/x265_3.0/source/encoder/ratecontrol.cpp:2864:31:
 warning: ‘~’ directive writing 1 byte into a region of size between 0 and 127 
[-Wformat-overflow=]
 sprintf(deltaPOC, "%s%d~", deltaPOC, rpsWriter->deltaPOC[i]);
   ^~~
In file included from /usr/include/stdio.h:873,
 from /usr/include/c++/8/cstdio:42,
 from 
/home/flip/programs/src/HandBrake/build/contrib/x265/x265_3.0/source/common/common.h:33,
 from 
/home/flip/programs/src/HandBrake/build/contrib/x265/x265_3.0/source/encoder/ratecontrol.cpp:30:
/usr/include/x86_64-linux-gnu/bits/stdio2.h:36:34: note: 
‘__builtin___sprintf_chk’ output between 3 and 140 bytes into a destination of 
size 128
   return __builtin___sprintf_chk (__s, __USE_FORTIFY_LEVEL - 1,
  ^~
   __bos (__s), __fmt, __va_arg_pack ());
   ~
/home/flip/programs/src/HandBrake/build/contrib/x265/x265_3.0/source/encoder/ratecontrol.cpp:2865:28:
 warning: ‘~’ directive writing 1 byte into a region of size between 0 and 39 
[-Wformat-overflow=]
 sprintf(bUsed, "%s%d~", bUsed, rpsWriter->bUsed[i]);
^~~
In file included from /usr/include/stdio.h:873,
 from /usr/include/c++/8/cstdio:42,
 from 
/home/flip/programs/src/HandBrake/build/contrib/x265/x265_3.0/source/common/common.h:33,
 from 
/home/flip/programs/src/HandBrake/build/contrib/x265/x265_3.0/source/encoder/ratecontrol.cpp:30:
/usr/include/x86_64-linux-gnu/bits/stdio2.h:36:34: note: 
‘__builtin___sprintf_chk’ output between 3 and 42 bytes into a destination of 
size 40
   return __builtin___sprintf_chk (__s, __USE_FORTIFY_LEVEL - 1,
  ^~
   __bos (__s), __fmt, __va_arg_pack ());
   ~
```

Probably related to

https://bitbucket.org/multicoreware/x265/issues/410/compiler-warnings

https://bitbucket.org/multicoreware/x265/issues/423/linux-build-error

But now it's half a year later so these are the "current" results.


___
x265-devel mailing list
x265-devel@videolan.org
https://mailman.videolan.org/listinfo/x265-devel


[x265] Issue #479: x265 2.8: Build script x265_2.8\build\vc15-x86_64\multilib.bat doesn't work with Visual Studio Community 2017 (multicoreware/x265)

2019-03-20 Thread Anonymous
New issue 479: x265 2.8: Build script x265_2.8\build\vc15-x86_64\multilib.bat 
doesn't work with Visual Studio Community 2017
https://bitbucket.org/multicoreware/x265/issues/479/x265-28-build-script-x265_28-build-vc15

Former user:

I have Visual Studio Community 2017 installed but failed to build x265 because 
VS2017 has changed the path of vcvarsall.bat and its default parameters. I have 
to make the following change to make x265_2.8\build\vc15-x86_64\multilib.bat to 
work with Visual Studio Community 2017:
call "%VS150COMNTOOLS%\..\..\VC\Auxiliary\Build\vcvarsall.bat" x86_amd64 
10.0.17763.0
Please update x265_2.8\build\vc15-x86_64\multilib.bat.
Thanx!
Jiong


___
x265-devel mailing list
x265-devel@videolan.org
https://mailman.videolan.org/listinfo/x265-devel