File x265.changes of Package x265 (Revision 16)

Currently displaying revision 16 , Show latest

311
 
1
-------------------------------------------------------------------
2
Tue Dec 20 12:02:33 UTC 2016 - scarabeus@opensuse.org
3
4
- Switch to use cmake macros
5
6
-------------------------------------------------------------------
7
Thu Sep 29 12:26:59 UTC 2016 - idonmez@suse.com
8
9
- Update to version 2.1
10
  Encoder enhancements
11
  * Support for qg-size of 8
12
  * Support for inserting non-IDR I-frames at scenecuts and when
13
    running with settings for fixed-GOP (min-keyint = max-keyint)
14
  * Experimental support for slice-parallelism.
15
  API changes
16
  * Encode user-define SEI messages passed in through x265_picture
17
    object.
18
  * Disable SEI and VUI messages from the bitstream
19
  * Specify qpmin and qpmax
20
  * Control number of bits to encode POC.
21
  Bug fixes
22
  * QP fluctuation fix for first B-frame in mini-GOP for 2-pass
23
    encoding with tune-grain.
24
  * Assembly fix for crashes in 32-bit from dct_sse4.
25
  * Threadpool creation fix in windows platform.
26
27
-------------------------------------------------------------------
28
Sun Aug 28 11:51:23 UTC 2016 - joerg.lorenzen@ki.tng.de
29
30
- Update to version 2.0
31
  API and Key Behavior Changes
32
  * x265_rc_stats added to x265_picture, containing all RC decision
33
    points for that frame.
34
  * PTL: high tier is now allowed by default, chosen only if
35
    necessary.
36
  * multi-pass: First pass now uses slow-firstpass by default,
37
    enabling better RC decisions in future passes.
38
  * pools: fix behaviour on multi-socketed Windows systems, provide
39
    more flexibility in determining thread and pool counts.
40
  * ABR: improve bits allocation in the first few frames, abr reset,
41
    vbv and cutree improved.
42
  New Features
43
  * uhd-bd: Enforce Ultra-HD Blu-ray Disc parameters
44
    (overrides any other settings).
45
  * rskip: Enables skipping recursion to analyze lower CU sizes
46
    using heuristics at different rd-levels. Provides good visual
47
    quality gains at the highest quality presets.
48
  * rc-grain: Enables a new rate control mode specifically for
49
    grainy content. Strictly prevents QP oscillations within and
50
    between frames to avoid grain fluctuations.
51
  * tune grain: A fully refactored and improved option to encode
52
    film grain content including QP control as well as analysis
53
    options.
54
  * asm: ARM assembly is now enabled by default, native or cross
55
    compiled builds supported on armv6 and later systems.
56
  Misc
57
  * An SSIM calculation bug was corrected
58
- soname bump to 87.
59
- Fixed arm.patch.
60
- Added libnuma-devel as buildrequires for arch x86_64 (except
61
  for openSUSE 13.1 because libnuma-devel >= 2.0.9 is required).
62
63
-------------------------------------------------------------------
64
Wed Feb  3 13:22:42 UTC 2016 - idonmez@suse.com
65
66
- Update to version 1.9
67
  API Changes:
68
  * x265_frame_stats returns many additional fields: maxCLL, maxFALL,
69
    residual energy, scenecut and latency logging
70
  * --qpfile now supports frametype 'K"
71
  * x265 now allows CRF ratecontrol in pass N (N greater than or equal to 2)
72
  * Chroma subsampling format YUV 4:0:0 is now fully supported and tested
73
  New Features:
74
  * Quant offsets: This feature allows block level quantization offsets
75
    to be specified for every frame. An API-only feature.
76
  * --intra-refresh: Keyframes can be replaced by a moving column
77
    of intra blocks in non-keyframes.
78
  * --limit-modes: Intelligently restricts mode analysis.
79
  * --max-luma and --min-luma for luma clipping, optional for HDR use-cases
80
  * Emergency denoising is now enabled by default in very low bitrate, 
81
    VBV encodes
82
  Presets and Performance:
83
  * Recently added features lookahead-slices, limit-modes, limit-refs
84
    have been enabled by default for applicable presets.
85
  * The default psy-rd strength has been increased to 2.0
86
  * Multi-socket machines now use a single pool of threads that can
87
    work cross-socket.
88
89
-------------------------------------------------------------------
90
Fri Nov 27 18:21:04 UTC 2015 - aloisio@gmx.com
91
92
- Update to version 1.8:
93
  API Changes:
94
  * Experimental support for Main12 is now enabled. Partial
95
    assembly support exists.
96
  * Main12 and Intra/Still picture profiles are now supported.
97
    Still picture profile is detected based on
98
    x265_param::totalFrames.
99
  * Three classes of encoding statistics are now available
100
    through the API.
101
    + x265_stats - contains encoding statistics, available
102
      through x265_encoder_get_stats()
103
    + x265_frame_stats and x265_cu_stats - contains frame
104
      encoding statistics, available through recon x265_picture
105
  * --csv
106
  * x265_encoder_log() is now deprecated
107
  * x265_param::csvfn is also deprecated
108
  * --log-level now controls only console logging, frame
109
    level console logging has been removed.
110
  * Support added for new color transfer characteristic ARIB
111
    STD-B67
112
  New Features:
113
  * limit-refs
114
    + This feature limits the references analysed for
115
      individual CUS.
116
    + Provides a nice tradeoff between efficiency and
117
      performance.
118
    + aq-mode 3
119
  * A new aq-mode that provides additional biasing for
120
    low-light conditions.
121
  * An improved scene cut detection logic that allows
122
    ratecontrol to manage visual quality at fade-ins and
123
    fade-outs better.
124
  Preset and Tune Options:
125
  * tune grain
126
    + Increases psyRdoq strength to 10.0, and rdoq-level to 2.
127
    + qg-size
128
  * Default value changed to 32.
129
- soname bump to 68
130
- Reworked arm.patch for 1.8
131
132
-------------------------------------------------------------------
133
Fri May 29 09:11:02 UTC 2015 - aloisio@gmx.com
134
135
- soname bump to 59
136
- Update to version 1.7
137
  * large amount of assembly code optimizations
138
  * some preliminary support for high dynamic range content
139
  * improvements for multi-library support
140
  * some new quality features
141
    (full documentation at: http://x265.readthedocs.org/en/1.7)
142
  * This release simplifies the multi-library support introduced
143
    in version 1.6. Any libx265 can now forward API requests to
144
    other installed libx265 libraries (by name) so applications
145
    like ffmpeg and the x265 CLI can select between 8bit and 10bit
146
    encodes at runtime without the need of a shim library or
147
    library load path hacks. See --output-depth, and
148
    http://x265.readthedocs.org/en/1.7/api.html#multi-library-interface
149
  * For quality, x265 now allows you to configure the quantization
150
    group size smaller than the CTU size (for finer grained AQ
151
    adjustments). See --qg-size.
152
  * x265 now supports limited mid-encode reconfigure via a new public
153
    method: x265_encoder_reconfig()
154
  * For HDR, x265 now supports signaling the SMPTE 2084 color transfer
155
    function, the SMPTE 2086 mastering display color primaries, and the
156
    content light levels. See --master-display, --max-cll
157
  * x265 will no longer emit any non-conformant bitstreams unless
158
    --allow-non-conformance is specified.
159
  * The x265 CLI now supports a simple encode preview feature. See
160
    --recon-y4m-exec.
161
  * The AnnexB NAL headers can now be configured off, via x265_param.bAnnexB
162
    This is not configurable via the CLI because it is a function of the
163
    muxer being used, and the CLI only supports raw output files. See
164
    --annexb
165
  Misc:
166
  * --lossless encodes are now signaled as level 8.5
167
  * --profile now has a -P short option
168
  * The regression scripts used by x265 are now public, and can be found at:
169
    https://bitbucket.org/sborho/test-harness
170
  * x265's cmake scripts now support PGO builds, the test-harness can be
171
    used to drive the profile-guided build process.
172
173
-------------------------------------------------------------------
174
Tue Apr 28 20:08:06 UTC 2015 - aloisio@gmx.com
175
176
- soname bumped to 51
177
- Update to stable version 1.6
178
  Perfomance changes:
179
  * heavy improvements for AVX2 capable platforms
180
    (Haswell and later Intel CPUs) and work efficiency
181
    improvements for multiple-socket machines.
182
  
183
  API changes:
184
  * --threads N replaced by --pools N,N and --lookahead-slices N
185
  * --[no-]rdoq-level N - finer control over RDOQ effort
186
  * --min-cu-size N - trade-off compression for performance
187
  * --max-tu-size N - trade-off compression for performance
188
  * --[no-]temporal-layers - code unreferenced B frames in temporal
189
    layer 1
190
  * --[no-]cip aliases added for --[no-]constrained-intra
191
  * Added support for new color transfer functions "smpte-st-2084"
192
    and "smpte-st-428
193
  * --limit-refs N was added, but not yet implemented
194
  * Deprecated x265_setup_primitives() was removed from the public
195
    API and is no longer exported DLLs
196
  
197
  Threading changes:
198
  * The x265 thread pool has been made NUMA aware.
199
  * The --threads  parameter, which used to specify a global
200
    pool size, has been replaced with a --pools parameter which
201
    allows you to specify a pool size per NUMA node (aka CPU socket
202
    or package). The default is still to allocate one pool worker
203
    thread per logical core on the machine, but with --pools one
204
    can isolate those threads to a given socket.
205
  * Other than socket isolation, the biggest visible change in the
206
    NUMA aware thread pools is the increase in work efficiency.
207
    The total utilization will generally decrease but the performance
208
    will increase since worker threads spend less time context
209
    switching.  Also, the threading of the lookahead was made more
210
    work-efficient. Each lookahead job is a much larger piece of work.
211
    Before (1.5):
212
    disable thread pool: --threads 1
213
    default thread pool: --threads 0
214
    restrict to 4 threads: --threads 4
215
    After (1.6):
216
    disable thread pools: --pools 0
217
    default thread pools: --pools *
218
    restrict to 4 threads: --pools 4
219
    restrict to 4 threads on socket 1: --pools -,4
220
    restrict to all threads on socket 0: --pools +,-
221
  
222
  Multi-lib interface:
223
  * In order to support runtime selection of a libx265
224
    shared library, we have introduced an x265_api structure
225
    and an x265_api_get() function. Applications which use
226
    this interface to acquire the libx265 functional interface
227
    will be able to use shim libraries to bind a particular build
228
    of libx265 at run time. See the API documentation for full
229
    details.
230
231
-------------------------------------------------------------------
232
Sun Feb 22 09:07:11 UTC 2015 - aloisio@gmx.com
233
234
- soname bump
235
- Update to stable version 1.5
236
  * improvements in Main10 compression efficiency and performance
237
    and psycho-visual optimizations now enabled by default
238
  Feature additions:
239
  * analysis re-use features have been completed
240
  * rate control zones have been introduced
241
  * --tune grain introduced
242
  * deblocking tC and Beta offsets are now configurable
243
  * denoise is seperately configurable for inter and intra CUs
244
  * frame based CSV logging has been improved
245
  * New support for VTune task profiles
246
  Presets and defaults:
247
  * ultrafast no longer disables the deblocking loop filter
248
  * psy-rd defaults to 0.3   (was 0, disabled)
249
  * psy-rdoq defaults to 1.0 (was 0, disabled)
250
  * aq-mode defaults to 1    (was 2, auto-variance)
251
  * 4:2:2 and 4:4:4 encodes no longer generate compliance warnings
252
  API changes:
253
  * param.rc.rateTolerance has been removed and replaced with a simpler
254
    param.rc.bStrictCbr flag.
255
  * --log-level debug is now --log-level 4 instead of --log-level 3.
256
    A new 'frame' log level was inserted at level 3 in order to support
257
    frame level CSV logging without also enabling frame level console
258
    logging. Using the string name 'debug' is unambiguous as its
259
    behavior has not changed.
260
  
261
- version 1.4
262
  * large refactoring in the analysis code
263
  Feature additions:
264
  * --pmode (parallel mode decision)
265
  * --pme (parallel motion estimation).
266
  Presets and defaults:
267
  * --amp is now respected in RD levels 2, 3, and 4 (previously only
268
    in 5 and 6).
269
  * --b-intra is now respected in all RD levels.
270
  * --fast-cbf, which has only ever effective at RD levels 5 and 6,
271
    is no longer enabled uselessly in the fastest presets.
272
  * --weightb is now enabled by default at presets slower, veryslow,
273
    and placebo.
274
  * --cu-lossless was changed to only attempt a lossless encode of
275
    the best lossy encode method. This made --cu-lossless a much less
276
    expensive encode option to have enabled, and hopefully made the
277
    feature more robust and maintainable.
278
  * The upper threshold for --psy-rdoq was raised to 50 (from 10)
279
    since the higher values were found to be beneficial for sources
280
    with high frequency noise (film grain).
281
  * The default thread pool size logic was updated to account for the
282
    addition of --pmode and --pme (if WPP is disabled but --pmode or
283
    --pme are enabled, a thread pool is still allocated).
284
  
285
-------------------------------------------------------------------
286
Mon Dec  8 04:53:31 UTC 2014 - crrodriguez@opensuse.org
287
288
- Ensure we use the proper CXXFLAGS, CFLAGS and therefore
289
  debuginfo packages are generated correctly.
290
291
-------------------------------------------------------------------
292
Sat Oct  4 15:24:23 UTC 2014 - aloisio@gmx.com
293
294
- Bumped to version hg20140928
295
296
-------------------------------------------------------------------
297
Thu Jun  5 14:55:45 UTC 2014 - guillaume@opensuse.org
298
299
- Fix ARM build with arm.patch from Arch Linux:
300
  https://github.com/archlinuxarm/PKGBUILDs/blob/master/extra/x265/arm.patch
301
302
-------------------------------------------------------------------
303
Thu May 08 00:00:00 UTC 2014 - Manfred.Tremmel@iiv.de
304
305
- added baselibs.conf
306
307
-------------------------------------------------------------------
308
Mon Mar 24 00:00:00 UTC 2014 - Manfred.Tremmel@iiv.de
309
310
- initial build of todays mercurial checkout
311