1<!DOCTYPE html> 2<html lang="en"> 3 <head> 4 <meta charset="utf-8" /> 5 <title> 6FFmpeg</title> 7 <link rel="stylesheet" href="css/font-awesome.min.css" /> 8 <link rel="stylesheet" href="css/bootstrap.min.css" /> 9 <link rel="stylesheet" href="css/simple-sidebar.css" /> 10 <link rel="stylesheet" href="css/style.min.css" /> 11 <meta name="viewport" content="width=device-width,initial-scale=1.0"> 12 <!--[if lt IE 9]> 13
13<script src="js/html5shiv.min.js"></script>
13 14
14<script src="js/respond.min.js"></script>
14 15 <![endif]--> 16 17 <link href="main.rss" rel="alternate" type="application/rss+xml" title="ffmpeg.org" /> 18 <link rel="shortcut icon" href="favicon.ico" /> 19 </head> 20 <body> 21 22 <div id="wrapper"> 23 24 <nav id="sidebar-wrapper"> 25 <ul class="sidebar-nav"> 26 <li class="sidebar-brand"><a href="."> 27 <img src="img/ffmpeg3d_white_20.png" alt="FFmpeg" /> 28 FFmpeg</a> 29 </li> 30 <li><a href="about.html">About</a></li> 31 <li><a href="index.html#news">News</a></li> 32 <li><a href="download.html">Download</a></li> 33 <li><a href="documentation.html">Documentation</a></li> 34 <li><a href="community.html">Community</a> 35 <ul> 36 <li><a href="community.html#Code-of-Conduct">Code of Conduct</a></li> 37 <li><a href="contact.html#MailingLists">Mailing Lists</a></li> 38 <li><a href="contact.html#IRCChannels">IRC</a></li> 39 <li><a href="contact.html#Forums">Forums</a></li> 40 <li><a href="bugreports.html">Bug Reports</a></li> 41 <li><a href="https://trac.ffmpeg.org">Wiki</a></li> 42 <li><a href="https://trac.ffmpeg.org/wiki/Conferences">Conferences</a></li> 43 </ul> 44 </li> 45 <li><a href="developer.html">Developers</a> 46 <ul> 47 <li><a href="download.html#get-sources">Source Code</a> 48 <li><a href="developer.html#Introduction">Contribute</a></li> 49 <li><a href="http://fate.ffmpeg.org">FATE</a></li> 50 <li><a href="http://coverage.ffmpeg.org">Code Coverage</a></li> 51 <li><a href="spi.html">Funding through SPI</a></li> 52 </ul> 53 </li> 54 <li><a href="#">More</a> 55 <ul> 56 <li><a href="donations.html"> 57 Donate<span style="color: #e55; font-size: 0.8em; margin-left: -10px"><i class="fa fa-heart"></i></span></a></li> 58 <li><a href="consulting.html">Hire Developers</a></li> 59 <li><a href="contact.html">Contact</a></li> 60 <li><a href="security.html">Security</a></li> 61 <li><a href="legal.html">Legal</a></li> 62 </ul> 63 </li> 64 </ul> 65 </nav> 66 67 <div id="page-content-wrapper"> 68 <header class="content-header"> 69 <h1> 70 <a id="menu-toggle" href="#" class="btn btn-success"><i class="fa fa-reorder"></i></a> 71FFmpeg </h1> 72 </header> 73 <div class="page-content inset"> 74 75<div id="index"> 76 77 <div class="row"> 78 <div class="col-md-6"> 79 <h2 class="description"> 80 A complete, cross-platform solution to record, convert and stream audio and video. 81 </h2> 82 </div> <!-- col --> 83 <div class="col-md-6"> 84 <div class="btn-download-wrapper"> 85 <a href="download.html" class="btn btn-success"> 86 <i class="fa fa-cloud-download"></i> 87 Download 88 </a> 89 </div> <!-- btn-download-wrapper --> 90 </div> <!-- col --> 91 </div> <!-- row --> 92 93 <div class="well example"> 94 <h3>Converting <strong>video</strong> and <strong>audio</strong> has never been so easy. 95 </h3> 96 <pre>$ ffmpeg -i input.mp4 output.avi</pre> 97 <div class="text-right"> 98 <a href="about.html" class="btn btn-success btn-lg">Discover more</a> 99 </div> 100 </div> <!-- well --> 101 102 <h1 id="news"> 103 <span class="pull-right"> 104 <a href="main.rss"><strong><i class="fa fa-rss"></i></strong></a> 105 <a href="https://twitter.com/FFmpeg"><strong><i class="fa fa-twitter"></i></strong></a> 106 <a href="https://www.facebook.com/ffmpeg"><strong><i class="fa fa-facebook"></i></strong></a> 107 </span> 108 News 109 </h1> 110 111 <h3 id="ampere2026">June 24th, 2026, Ampere Server Donation</h3> 112 <p> 113 Ampere has generously donated a server to FFmpeg: an AmpereOne® (Mt. Mitchell) 114 2U system with 192 Arm cores, 512 GB of RAM, 24 NVMe bays and 2×25G 115 networking, weighing in at 28 kg. Thank you, Ampere! 116 </p> 117 <div class="row" style="margin-top: 1em; margin-bottom: 1em;"> 118 <div class="col-sm-6"> 119 <img class="img-responsive" src="img/AmpereServer1.avif" alt="The boxed Ampere server on arrival" /> 120 </div> 121 <div class="col-sm-6"> 122 <img class="img-responsive" src="img/AmpereServer2.avif" alt="The Ampere server unboxed in its packaging foam" /> 123 </div> 124 </div> 125 <p> 126 To celebrate, Dascha (daschasara) answered FFmpeg's call for non-AI artwork and 127 drew this piece for us: 128 </p> 129 <img src="img/dascha-ffmpeg-ampere.png" alt="Artwork by Dascha: the FFmpeg logo o
129pening a gift box containing an Ampere server" style="display: block; margin: 1em auto; width: 100%; max-width: 500px;" /> 130 <p style="text-align: center;"> 131 <small>Artwork by Dascha (daschasara)</small> 132 </p> 133 134 <h3 id="pr8.1">March 16th, 2026, FFmpeg 8.1 <span title="Sir Charles Antony Richard Hoare">"Hoare"</span></h3> 135 <p> 136 A new minor release, <a href="download.html#release_8.1">FFmpeg 8.1 <span title="Sir Charles Antony Richard Hoare">"Hoare"</span></a>, 137 is now available for download. Here are some of the highlights: 138 <ul> 139 <li>Decoders: xHE-AAC Mps212 (experimental) MPEG-H decoding via libmpeghdec</li> 140 <li>EXIF Metadata Parsing</li> 141 <li>LCEVC: support for parsing and forwarding metadata</li> 142 <li>Vulkan compute-based codecs: ProRes encoding and decoding, DPX decoding</li> 143 <li>D3D12: D3D12 H.264/AV1 encoding, scale_d3d12, mestimate_d3d12, deinterlace_d3d12 filters</li> 144 <li>Rockchip H.264/HEVC hardware encoding</li> 145 <li>IAMF: Projection mode Ambisonic Audio Elements muxing and demuxing</li> 146 <li>Formats: hxvs demuxer</li> 147 <li>Filters: drawvg, vpp_amf</li> 148 </ul> 149 </p> 150 151 <p> 152 This release features a lot of internal changes and bugfixes. The groundwork for the upcoming swscale rewrite is progressing. 153 The Vulkan compute-based codecs, and a few filters, no longer depend on runtime GLSL compilation, which speeds up their initialization. 154 </p> 155 156 <p> 157 A companion post about the Vulkan Compute-based codec implementations has been published on the 158 <a href="https://www.khronos.org/blog/video-encoding-and-decoding-with-vulkan-compute-shaders-in-ffmpeg">Khronos blog</a>, 159 featuring technical details on the implementations and future plans. 160 </p> 161 162 <p> 163 We recommend users, distributors, and system integrators to upgrade unless they use current git master. 164 </p> 165 166 <h3 id="pr8.0">August 22nd, 2025, FFmpeg 8.0 <span title="David A. Huffman">"Huffman"</span></h3> 167 <p> 168 A new major release, <a href="download.html#release_8.0">FFmpeg 8.0 <span title="David A. Huffman">"Huffman"</span></a>, 169 is now available for download. 170 Thanks to several delays, and modernization of our entire infrastructure, this release ended up 171 being one of our largest releases to date. In short, its new features are: 172 <ul> 173 <li>Native decoders: <span title="Advanced Professional Video">APV</span>, ProRes RAW, RealVideo 6.0, Sanyo LD-ADPCM, G.728</li> 174 <li>VVC decoder improvements: <span title="Intra Block Copy">IBC</span>, 175 <span title="Adaptive Color Transform">ACT</span>, 176 Palette Mode</li> 177 <li>Vulkan compute-based codecs: FFv1 (encode and decode), ProRes RAW (decode only)</li> 178 <li>Hardware accelerated decoding: Vulkan VP9, VAAPI VVC, OpenHarmony H264/5</li> 179 <li>Hardware accelerated encoding: Vulkan AV1, OpenHarmony H264/5</li> 180 <li>Formats: MCC, G.728, Whip, APV</li> 181 <li>Filters: colordetect, pad_cuda, scale_d3d11, Whisper, and others</li> 182 </ul> 183 </p> 184 185 <p> 186 A new class of decoders and encoders based on pure Vulkan compute implementation have been added. 187 Vulkan is a cross-platform, open standard set of APIs that allows programs to use GPU hardware in various ways, 188 from drawing on screen, to doing calculations, to decoding video via custom hardware accelerators. 189 Rather than using a custom hardware accelerator present, these codecs are based on compute shaders, and work 190 on any implementation of Vulkan 1.3.<br> 191 Decoders use the same hwaccel API and commands, so users do not need to do anything special to enable them, 192 as enabling <a href="https://trac.ffmpeg.org/wiki/HWAccelIntro#Vulkan">Vulkan decoding</a> is sufficient to use them.<br> 193 Encoders, like our hardware accelerated encoders, require specifying a new encoder (ffv1_vulkan). 194 Currently, the only codecs supported are: FFv1 (encoding and decoding) and ProRes RAW (decode only). 195 ProRes (encode+decode) and VC-2 (encode+decode) implementations are complete and currently in review, 196 to be merged soon and available with the next minor release.<br>
197 Only codecs specifically designed for parallelized decoding can be implemented in such a way, with 198 more mainstream codecs not being planned for support.<br> 199 Depending on the hardware, these new codecs can provide very significant speedups, and open up 200 possibilities to work with them for situations like non-linear video editors and 201 lossless screen recording/streaming, so we are excited to learn what our downstream users can make with them. 202 </p> 203 204 <p> 205 The project has recently started to modernize its infrastructure. Our mailing list servers have been 206 fully upgraded, and we have recently started to accept contributions via a new forge, available on 207 <a href="https://code.ffmpeg.org/">code.ffmpeg.org</a>, running a Forgejo instance. 208 </p> 209 210 <p> 211 As usual, we recommend that users, distributors, and system integrators to upgrade unless they use current git master. 212 </p> 213 214 <h3 id="pr7.1">September 30th, 2024, FFmpeg 7.1 <span title="Rózsa Péter">"Péter"</span></h3> 215 <p> 216 <a href="download.html#release_7.1">FFmpeg 7.1 "Péter"</a>, a new 217 major release, is now available! A full list of changes can be found in the release 218 <a href="https://git.ffmpeg.org/gitweb/ffmpeg.git/blob/refs/heads/release/7.1:/Changelog">changelog</a>. 219 </p> 220 <p> 221 The more important highlights of the release are that the VVC decoder, merged as experimental in version 7.0, 222 has had enough time to mature and be optimized enough to be declared as stable. The codec is starting to gain 223 traction with broadcast standardization bodies.<br> 224 Support has been added for a native AAC USAC (part of the xHE-AAC coding system) decoder, with the format starting 225 to be adopted by streaming websites, due to its extensive volume normalization metadata.<br> 226 MV-HEVC decoding is now supported. This is a stereoscopic coding tool that begun to be shipped and generated 227 by recent phones and VR headsets.<br> 228 LC-EVC decoding, an enhancement metadata layer to attempt to improve the quality of codecs, is now supported via an 229 external library.<br> 230 </p> 231 <p> 232 Support for Vulkan encoding, with H264 and HEVC was merged. This finally allows fully Vulkan-based decode-filter-encode 233 pipelines, by having a sink for Vulkan frames, other than downloading or displaying them. The encoders have feature-parity 234 with their VAAPI implementation counterparts. Khronos has announced that support for AV1 encoding is also coming soon to Vulkan, 235 and FFmpeg is aiming to have day-one support. 236 </p> 237 <p> 238 In addition to the above, this release has had a lot of important internal work done. By far, the standout internally 239 are the improvements made for full-range images. Previously, color range data had two paths, no negotiation, 240 and was unreliably forwarded to filters, encoders, muxers. Work on cleaning the system up started more than 10 241 years ago, however this stalled due to how fragile the system was, and that breaking behaviour would be unacceptable. 242 The new system fixes this, so now color range is forwarded correctly and consistently everywhere needed, and also 243 laid the path for more advanced forms of negotiation.<br> 244 Cropping metadata is now supported with Matroska and MP4 formats. This metadata is important not only for archival, 245 but also with AV1, as hardware encoders require its signalling due to the codec not natively supporting one. 246 </p> 247 <p> 248 As usual, we recommend that users, distributors, and system integrators to upgrade unless they use current git master. 249 </p> 250 251 <h3 id="coverity">September 11th, 2024, Coverity</h3> 252 <p> 253 The number of issues FFmpeg has in <a href="https://scan.coverity.com/projects/ffmpeg">Coverity (a static analyzer)</a> is now lower than it has been since 2016. 254 Our defect density is less than one 30th of the average in OSS with over a million code 255 lines. All this was possible thanks to a grant from the <a href="https://www.sovereigntechfund.de/">Sovereign Tech Fund</a>. 256 </p> 257 <img src="img/coverity-lifetime-2024-08.PNG" alt="Coverity Lifetime Graph till 2024-08" style="width: 80%; display: block; margin-left: auto; margin-right: auto;" /> 258 259 <h3 id="xheaac">June 2nd, 2024, native xHE-AAC decoder</h3> 260 <p> 261 FFmpeg now implements a native xHE-AAC decoder. Currently, streams without (e)SBR, USAC or MPEG-H Surround 262 are supported, which means the majority of xHE-AAC streams in use should work. Support for USAC and (e)SBR is 263 coming soon. Work is also ongoing to improve its stability and compatibility. 264 During the process we found several specification issues, which were then submitted back to the authors 265 for discussion and potential inclusion in a future errata. 266 </p> 267 268 <h3 id="stf24">May 13th, 2024, Sovereign Tech Fund</h3> 269 <p> 270 The FFmpeg community is excited to announce that Germany's 271 <a href="https://www.sovereigntechfund.de/tech/ffmpeg">Sovereign Tech Fund</a> 272 has become its first governmental sponsor. Their support will help 273 sustain the maintainance of the FFmpeg project, a critical open-source 274 software multimedia component essential to bringing audio and video to 275 billions around the world everyday. 276 </p> 277 278 <h3 id="pr7.0">April 5th, 2024, FFmpeg 7.0 "Dijkstra"</h3> 279 <p> 280 A new major release, <a href="download.html#release_7.0">FFmpeg 7.0 "Dijkstra"</a>, 281 is now available for download. The most noteworthy changes for most users are 282 a <a href="#vvcdec">native VVC decoder</a> (currently experimental, until more 283 fuzzing is done), <a href="#iamf">IAMF support</a>, or a 284 <a href="#cli_threading">multi-threaded <code>ffmpeg</code> CLI tool</a>.
285 </p> 286 287 <p> 288 This release is <em>not</em> backwards compatible, removing APIs deprecated before 6.0. 289 The biggest change for most library callers will be the removal of the old bitmask-based 290 channel layout API, replaced by the <code>AVChannelLayout</code> API allowing such 291 features as custom channel ordering, or Ambisonics. Certain deprecated <code>ffmpeg</code> 292 CLI options were also removed, and a C11-compliant compiler is now required to build 293 the code. 294 </p> 295 296 <p> 297 As usual, there is also a number of new supported formats and codecs, new filters, APIs, 298 and countless smaller features and bugfixes. Compared to 6.1, the <code>git</code> repository 299 contains almost ∼2000 new commits by ∼100 authors, touching >100000 lines in 300 ∼2000 files — thanks to everyone who contributed. See the 301 <a href="https://git.ffmpeg.org/gitweb?p=ffmpeg.git;a=blob_plain;f=Changelog;hb=n7.0">Changelog</a>, 302 <a href="https://git.ffmpeg.org/gitweb?p=ffmpeg.git;a=blob_plain;f=doc/APIchanges;hb=n7.0">APIchanges</a>, 303 and the git log for more comprehensive lists of changes. 304 </p> 305 306 <h3 id="vvcdec">January 3rd, 2024, native VVC decoder</h3> 307 <p> 308 The <code>libavcodec</code> library now contains a native VVC (Versatile Video Coding) 309 decoder, supporting a large subset of the codec's features. Further optimizations and 310 support for more features are coming soon. The code was written by Nuo Mi, Xu Mu, 311 Frank Plowman, Shaun Loo, and Wu Jianhua. 312 </p> 313 314 <h3 id="iamf">December 18th, 2023, IAMF support</h3> 315 <p> 316 The <code>libavformat</code> library can now read and write <a href="https://aomediacodec.github.io/iamf/">IAMF</a> 317 (Immersive Audio) files. The <code>ffmpeg</code> CLI tool can configure IAMF structure with the new 318 <code>-stream_group</code> option. IAMF support was written by James Almer. 319 </p> 320 321 <h3 id="cli_threading">December 12th, 2023, multi-threaded <code>ffmpeg</code> CLI tool</h3> 322 <p> 323 Thanks to a major refactoring of the <code>ffmpeg</code> command-line tool, all the major 324 components of the transcoding pipeline (demuxers, decoders, filters, encodes, muxers) now 325 run in parallel. This should improve throughput and CPU utilization, decrease latency, 326 and open the way to other exciting new features. 327 </p> 328 329 <p> 330 Note that you should <em>not</em> expect significant performance improvements in cases 331 where almost all computational time is spent in a single component (typically video 332 encoding). 333 </p> 334 335 <h3 id="pr6.1">November 10th, 2023, FFmpeg 6.1 "Heaviside"</h3> 336 <p> 337 <a href="download.html#release_6.1">FFmpeg 6.1 "Heaviside"</a>, a new 338 major release, is now available! Some of the highlights: 339 </p> 340 <ul> 341 <li>libaribcaption decoder</li> 342 <li>Playdate video decoder and demuxer</li> 343 <li>Extend VAAPI support for libva-win32 on Windows</li> 344 <li>afireqsrc audio source filter</li> 345 <li>arls filter</li> 346 <li>ffmpeg CLI new option: -readrate_initial_burst</li> 347 <li>zoneplate video source filter</li> 348 <li>command support in the setpts and asetpts filters</li> 349 <li>Vulkan decode hwaccel, supporting H264, HEVC and AV1</li> 350 <li>color_vulkan filter</li> 351 <li>bwdif_vulkan filter</li> 352 <li>nlmeans_vulkan filter</li> 353 <li>RivaTuner video decoder</li> 354 <li>xfade_vulkan filter</li> 355 <li>vMix video decoder</li> 356 <li>Essential Video Coding parser, muxer and demuxer</li> 357 <li>Essential Video Coding frame merge bsf</li> 358 <li>bwdif_cuda filter</li> 359 <li>Microsoft RLE video encoder</li> 360 <li>Raw AC-4 muxer and demuxer</li> 361 <li>Raw VVC bitstream parser, muxer and demuxer</li> 362 <li>Bitstream filter for editing metadata in VVC streams</li> 363 <li>Bitstream filter for converting VVC from MP4 to Annex B</li> 364 <li>scale_vt filter for videotoolbox</li> 365 <li>transpose_vt filter for videotoolbox</li> 366 <li>support for the P_SKIP hinting to speed up libx264 encoding</li> 367 <li>Support HEVC,VP9,AV1 codec in enhanced flv format</li> 368 <li>apsnr and asisdr audio filters</li> 369 <li>OSQ demuxer and decoder</li> 370 <li>Support HEVC,VP9,AV1 codec fourcclist in enhanced rtmp protocol</li> 371 <li>
371CRI USM demuxer</li> 372 <li>ffmpeg CLI '-top' option deprecated in favor of the setfield filter</li> 373 <li>VAAPI AV1 encoder</li> 374 <li>ffprobe XML output schema changed to account for multiple variable-fields elements within the same parent element</li> 375 <li>ffprobe -output_format option added as an alias of -of</li> 376 </ul> 377 <p> 378 This release had been overdue for at least half a year, but due to constant activity in the repository, 379 had to be delayed, and we were finally able to branch off the release recently, before some of the large 380 changes scheduled for 7.0 were merged. 381 </p> 382 <p> 383 Internally, we have had a number of changes too. The FFT, MDCT, DCT and DST implementation used for codecs 384 and filters has been fully replaced with the faster libavutil/tx (full article about it coming soon).<br> 385 This also led to a reduction in the the size of the compiled binary, which can be noticeable in small builds.<br> 386 There was a very large reduction in the total amount of allocations being done on each frame throughout video decoders, 387 reducing overhead.<br> 388 RISC-V optimizations for many parts of our DSP code have been merged, with mainly the large decoders being left.<br> 389 There was an effort to improve the correctness of timestamps and frame durations of each packet, increasing the 390 accurracy of variable frame rate video. 391 </p> 392 <p> 393 Next major release will be version 7.0, scheduled to be released in February. We will attempt to better stick 394 to the new release schedule we announced at the start of this year. 395 </p> 396 <p> 397 We strongly recommend users, distributors, and system integrators to upgrade unless they use current git master. 398 </p> 399 400 <h3 id="vk2023">May 31st, 2023, Vulkan decoding</h3> 401 <p> 402 A few days ago, Vulkan-powered decoding hardware acceleration code was merged into the codebase. 403 This is the first vendor-generic and platform-generic decode acceleration API, enabling the 404 same code to be used on multiple platforms, with very minimal overhead. 405 This is also the first multi-threaded hardware decoding API, and our code makes full use of this, 406 saturating all available decode engines the hardware exposes. 407 </p> 408 <p> 409 Those wishing to test the code can read our 410 <a href="https://trac.ffmpeg.org/wiki/HWAccelIntro#Vulkan">documentation page</a>. 411 For those who would like to integrate FFmpeg's Vulkan code to demux, parse, decode, and receive 412 a VkImage to present or manipulate, documentation and examples are available in our source tree. 413 Currently, using the latest available git checkout of our 414 <a href="https://git.ffmpeg.org/gitweb?p=ffmpeg.git;a=summary">repository</a> is required. 415 The functionality will be included in stable branches with the release of version 6.1, due 416 to be released soon. 417 </p> 418 <p> 419 As this is also the first practical implementation of the specifications, bugs may be present, 420 particularly in drivers, and, although passing verification, the implementation itself. 421 New codecs, and encoding support are also being worked on, by both the Khronos organization 422 for standardizing, and us as implementing it, and giving feedback on improving. 423 </p> 424 425 <h3 id="pr6.0">February 28th, 2023, FFmpeg 6.0 "Von Neumann"</h3> 426 <p> 427 A new major release, <a href="download.html#release_6.0">FFmpeg 6.0 "Von Neumann"</a>, 428 is now available for download. This release has many new encoders and decoders, filters, 429 ffmpeg CLI tool improvements, and also, changes the way releases are done. All major 430 releases will now bump the version of the ABI. We plan to have a new major release each 431 year. Another release-specific change is that deprecated APIs will be removed after 3 432 releases, upon the next major bump. 433 This means that releases will be done more often and will be more organized. 434 </p> 435 <p> 436 New decoders featured are Bonk, RKA, Radiance, SC-4, APAC, VQC, WavArc and a few ADPCM formats. 437 QSV and NVenc now support AV1 encoding. The FFmpeg CLI (we usually reffer to it as ffmpeg.c 438 to avoid confusion) has speed-up improvements due to threading, as well as statistics options, 439 and the ability to pass option values for filters from a file. There are quite a few new audio 440 and video filters, such as adrc, showcwt, backgroundkey and ssim360, with a few hardware ones too. 441 Finally, the release features many behind-the-scenes changes, including a new FFT and MDCT
442 implementation used in codecs (expect a blog post about this soon), numerous bugfixes, better 443 ICC profile handling and colorspace signalling improvement, introduction of a number of RISC-V 444 vector and scalar assembly optimized routines, and a few new improved APIs, which can be viewed 445 in the doc/APIchanges file in our tree. 446 A few submitted features, such as the Vulkan improvements and more FFT optimizations will be in the 447 next minor release, 6.1, which we plan to release soon, in line with our new release schedule. 448 Some highlights are: 449 </p> 450 <ul> 451 <li>Radiance HDR image support</li> 452 <li>ddagrab (Desktop Duplication) video capture filter</li> 453 <li>ffmpeg -shortest_buf_duration option</li> 454 <li>ffmpeg now requires threading to be built</li> 455 <li>ffmpeg now runs every muxer in a separate thread</li> 456 <li>Add new mode to cropdetect filter to detect crop-area based on motion vectors and edges</li> 457 <li>VAAPI decoding and encoding for 10/12bit 422, 10/12bit 444 HEVC and VP9</li> 458 <li>WBMP (Wireless Application Protocol Bitmap) image format</li> 459 <li>a3dscope filter</li> 460 <li>bonk decoder and demuxer</li> 461 <li>Micronas SC-4 audio decoder</li> 462 <li>LAF demuxer</li> 463 <li>APAC decoder and demuxer</li> 464 <li>Media 100i decoders</li> 465 <li>DTS to PTS reorder bsf</li> 466 <li>ViewQuest VQC decoder</li> 467 <li>backgroundkey filter</li> 468 <li>nvenc AV1 encoding support</li> 469 <li>MediaCodec decoder via NDKMediaCodec</li> 470 <li>MediaCodec encoder</li> 471 <li>oneVPL support for QSV</li> 472 <li>QSV AV1 encoder</li> 473 <li>QSV decoding and encoding for 10/12bit 422, 10/12bit 444 HEVC and VP9</li> 474 <li>showcwt multimedia filter</li> 475 <li>corr video filter</li> 476 <li>adrc audio filter</li> 477 <li>afdelaysrc audio filter</li> 478 <li>WADY DPCM decoder and demuxer</li> 479 <li>CBD2 DPCM decoder</li> 480 <li>ssim360 video filter</li> 481 <li>ffmpeg CLI new options: -stats_enc_pre[_fmt], -stats_enc_post[_fmt], -stats_mux_pre[_fmt]</li> 482 <li>hstack_vaapi, vstack_vaapi and xstack_vaapi filters</li> 483 <li>XMD ADPCM decoder and demuxer</li> 484 <li>media100 to mjpegb bsf</li> 485 <li>ffmpeg CLI new option: -fix_sub_duration_heartbeat</li> 486 <li>WavArc decoder and demuxer</li> 487 <li>CrystalHD decoders deprecated</li> 488 <li>SDNS demuxer</li> 489 <li>RKA decoder and demuxer</li> 490 <li>filtergraph syntax in ffmpeg CLI now supports passing file contents as option values</li> 491 <li>hstack_qsv, vstack_qsv and xstack_qsv filters</li> 492 </ul> 493 <p> 494 We strongly recommend users, distributors, and system integrators to 495 upgrade unless they use current git master. 496 </p> 497 498 <h3 id="pr5.1">July 22nd, 2022, FFmpeg 5.1 "Riemann"</h3> 499 <p> 500 <a href="download.html#release_5.1">FFmpeg 5.1 "Riemann"</a>, a new 501 major release, is now available! Some of the highlights: 502 </p> 503 <ul> 504 <li>add ipfs/ipns protocol support</li> 505 <li>dialogue enhance audio filter</li> 506 <li>dropped obsolete XvMC hwaccel</li> 507 <li>pcm-bluray encoder</li> 508 <li>DFPWM audio encoder/decoder and raw muxer/demuxer</li> 509 <li>SITI filter</li> 510 <li>Vizrt Binary Image encoder/decoder</li> 511 <li>avsynctest source filter</li> 512 <li>feedback video filter</li> 513 <li>pixelize video filter</li> 514 <li>colormap video filter</li> 515 <li>colorchart video source filter</li> 516 <li>multiply video filter</li> 517 <li>PGS subtitle frame merge bitstream filter</li> 518 <li>blurdetect filter</li> 519 <li>tiltshelf audio filter</li> 520 <li>QOI image format support</li> 521 <li>ffprobe -o option</li> 522 <li>virtualbass audio filter</li> 523 <li>VDPAU AV1 hwaccel</li> 524 <li>PHM image format support</li> 525 <li>remap_opencl filter</li> 526 <li>added chromakey_cuda filter</li> 527 </ul> 528 <p> 529 We strongly recommend users, distributors, and system integrators to 530 upgrade unless they use current git master. 531 </p> 532 533 <h3 id="pr5.0">January 17th, 2022, FFmpeg 5.0 "Lorentz"</h3> 534 <p> 535 <a href="download.html#release_5.0">FFmpeg 5.0 "Lorentz"</a>, a new 536 major release, is now available! For this long-overdue release, a major effort 537 underwent to remove the old encode/decode APIs and replace them with an 538 N:M-based API, the entire libavresample library was removed, libswscale 539 has a new, easier to use AVframe-based API, the Vulkan code was much improved, 540 many new filters were added, including libplacebo integration, and finally, 541 DoVi support was added, including tonemapping and remuxing. The default 542 AAC encoder settings were also changed to improve quality. 543 Some of the changelog highlights: 544 </p> 545 <ul> 546 <li>ADPCM IMA Westwood encoder</li> 547 <li>Westwood AUD muxer</li> 548 <li>ADPCM IMA Acorn Replay decoder</li> 549 <li>Argonaut Games CVG demuxer</li> 550 <li>Argonaut Games CVG muxer</li> 551 <li>Concatf protocol</li> 552 <li>afwtdn audio filter</li> 553 <li>audio and video segment filters</li> 554 <li>Apple Graphics (SMC) encoder</li> 555 <li>hsvkey and hsvhold video filters</li> 556 <li>adecorrelate audio filter</li> 557 <li>atilt audio filter</li> 558 <li>grayworld video filter</li> 559 <li>AV1 Low overhead bitstream format muxer</li> 560 <li>swscale slice threading</li> 561 <li>MSN Siren decoder</li> 562 <li>scharr video filter</li> 563 <li>apsyclip audio filter</li> 564 <li>morpho video filter</li> 565 <li>amr parser</li> 566 <li>(a)latency filters</li> 567 <li>GEM Raster image decoder</li> 568 <li>asdr audio filter</li> 569 <li>speex decoder</li> 570 <li>limitdiff video filter</li> 571 <li>xcorrelate video filter</li> 572 <li>varblur video filter</li> 573 <li>huesaturation video filter</li> 574 <li>colorspectrum source video filter</li> 575 <li>RTP packetizer for uncompressed video (RFC 4175)</li> 576 <li>bitpacked encoder</li> 577 <li>VideoToolbox VP9 hwaccel</li> 578 <li>VideoToolbox ProRes hwaccel</li> 579 <li>support loongarch.</li> 580 <li>aspectralstats audio filter</li> 581 <li>adynamicsmooth audio filter</li> 582 <li>libplacebo filter</li> 583 <li>vflip_vulkan, hflip_vulkan and flip_vulkan filters</li> 584 <li>adynamicequalizer audio filter</li> 585 <li>yadif_videotoolbox filter</li> 586 <li>VideoToolbox ProRes encoder</li> 587 <li>anlmf audio filter</li> 588 </ul> 589 <p> 590 We strongly recommend users, distributors, and system integrators to 591 upgrade unless they use current git master. 592 </p> 593 594 <h3 id="irc2021">June 19th, 2021, IRC</h3> 595 <p> 596 We have a new IRC home at Libera Chat 597 now! Feel free to join us at #ffmpeg and #ffmpeg-devel. More info at <a href="https://ffmpeg.org/contact.html#IRCChannels">contact#IRCChannels</a> 598 </p> 599 600 <h3 id="pr4.4">April 8th, 2021, FFmpeg 4.4 "Rao"</h3> 601 <p> 602 <a href="download.html#release_4.4">FFmpeg 4.4 "Rao"</a>, a new 603 major release, is now available! Some of the highlights: 604 </p> 605 <ul> 606 <li>AudioToolbox output device</li> 607 <li>MacCaption demuxer</li> 608 <li>PGX decoder</li> 609 <li>chromanr video filter</li> 610 <li>VDPAU accelerated HEVC 10/12bit decoding</li> 611 <li>ADPCM IMA Ubisoft APM encoder</li> 612 <li>Rayman 2 APM muxer</li> 613 <li>AV1 encoding support SVT-AV1</li> 614 <li>Cineform HD encoder</li> 615 <li>ADPCM Argonaut Games encoder</li> 616 <li>Argonaut Games ASF muxer</li> 617 <li>AV1 Low overhead bitstream format demuxer</li> 618 <li>RPZA video encoder</li> 619 <li>ADPCM IMA MOFLEX decoder</li> 620 <li>MobiClip FastAudio decoder</li> 621 <li>MobiClip video decoder</li> 622 <li>MOFLEX demuxer</li> 623 <li>MODS demuxer</li> 624 <li>PhotoCD decoder</li> 625 <li>MCA demuxer</li> 626 <li>AV1 decoder (Hardware acceleration used only)</li> 627 <li>SVS demuxer</li> 628 <li>Argonaut Games BRP demuxer</li> 629 <li>DAT demuxer</li> 630 <li>aax demuxer</li> 631 <li>IPU decoder, parser and demuxer</li> 632 <li>Intel QSV-accelerated AV1 decoding</li> 633 <li>Argonaut Games Video decoder</li> 634 <li>libwavpack encoder removed</li> 635 <li>ACE demuxer</li> 636 <li>AVS3 demuxer</li> 637 <li>AVS3 video decoder via libuavs3d</li> 638 <li>Cintel RAW decoder</li> 639 <li>VDPAU accelerated VP9 10/12bit decoding</li> 640 <li>afreqshift and aphaseshift filters</li> 641 <li>
641High Voltage Software ADPCM encoder</li> 642 <li>LEGO Racers ALP (.tun & .pcm) muxer</li> 643 <li>AV1 VAAPI decoder</li> 644 <li>adenorm filter</li> 645 <li>ADPCM IMA AMV encoder</li> 646 <li>AMV muxer</li> 647 <li>NVDEC AV1 hwaccel</li> 648 <li>DXVA2/D3D11VA hardware accelerated AV1 decoding</li> 649 <li>speechnorm filter</li> 650 <li>SpeedHQ encoder</li> 651 <li>asupercut filter</li> 652 <li>asubcut filter</li> 653 <li>Microsoft Paint (MSP) version 2 decoder</li> 654 <li>Microsoft Paint (MSP) demuxer</li> 655 <li>AV1 monochrome encoding support via libaom >= 2.0.1</li> 656 <li>asuperpass and asuperstop filter</li> 657 <li>shufflepixels filter</li> 658 <li>tmidequalizer filter</li> 659 <li>estdif filter</li> 660 <li>epx filter</li> 661 <li>Dolby E parser</li> 662 <li>shear filter</li> 663 <li>kirsch filter</li> 664 <li>colortemperature filter</li> 665 <li>colorcontrast filter</li> 666 <li>PFM encoder</li> 667 <li>colorcorrect filter</li> 668 <li>binka demuxer</li> 669 <li>XBM parser</li> 670 <li>xbm_pipe demuxer</li> 671 <li>colorize filter</li> 672 <li>CRI parser</li> 673 <li>aexciter audio filter</li> 674 <li>exposure video filter</li> 675 <li>monochrome video filter</li> 676 <li>setts bitstream filter</li> 677 <li>vif video filter</li> 678 <li>OpenEXR image encoder</li> 679 <li>Simbiosis IMX decoder</li> 680 <li>Simbiosis IMX demuxer</li> 681 <li>Digital Pictures SGA demuxer and decoders</li> 682 <li>TTML subtitle encoder and muxer</li> 683 <li>identity video filter</li> 684 <li>msad video filter</li> 685 <li>gophers protocol</li> 686 <li>RIST protocol via librist</li> 687 </ul> 688 <p> 689 We strongly recommend users, distributors, and system integrators to 690 upgrade unless they use current git master. 691 </p> 692 693 <h3 id="pr4.3">June 15th, 2020, FFmpeg 4.3 "4:3"</h3> 694 <p> 695 <a href="download.html#release_4.3">FFmpeg 4.3 "4:3"</a>, a new 696 major release, is now available! Some of the highlights: 697 </p> 698 <ul> 699 <li>v360 filter</li> 700 <li>Intel QSV-accelerated MJPEG decoding</li> 701 <li>Intel QSV-accelerated VP9 decoding</li> 702 <li>Support for TrueHD in mp4</li> 703 <li>Support AMD AMF encoder on Linux (via Vulkan)</li> 704 <li>IMM5 video decoder</li> 705 <li>ZeroMQ protocol</li> 706 <li>support Sipro ACELP.KELVIN decoding</li> 707 <li>streamhash muxer</li> 708 <li>sierpinski video source</li> 709 <li>scroll video filter</li> 710 <li>photosensitivity filter</li> 711 <li>anlms filter</li> 712 <li>arnndn filter</li> 713 <li>bilateral filter</li> 714 <li>maskedmin and maskedmax filters</li> 715 <li>VDPAU VP9 hwaccel</li> 716 <li>median filter</li> 717 <li>QSV-accelerated VP9 encoding</li> 718 <li>AV1 encoding support via librav1e</li> 719 <li>AV1 frame merge bitstream filter</li> 720 <li>AV1 Annex B demuxer</li> 721 <li>axcorrelate filter</li> 722 <li>mvdv decoder</li> 723 <li>mvha decoder</li> 724 <li>MPEG-H 3D Audio support in mp4</li> 725 <li>thistogram filter</li> 726 <li>freezeframes filter</li> 727 <li>Argonaut Games ADPCM decoder</li> 728 <li>Argonaut Games ASF demuxer</li> 729 <li>xfade video filter</li> 730 <li>xfade_opencl filter</li> 731 <li>afirsrc audio filter source</li> 732 <li>pad_opencl filter</li> 733 <li>Simon & Schuster Interactive ADPCM decoder</li> 734 <li>Real War KVAG demuxer</li> 735 <li>CDToons video decoder</li> 736 <li>siren audio decoder</li> 737 <li>Rayman 2 ADPCM decoder</li> 738 <li>Rayman 2 APM demuxer</li> 739 <li>cas video filter</li> 740 <li>High Voltage Software ADPCM decoder</li> 741 <li>LEGO Racers ALP (.tun & .pcm) demuxer</li> 742 <li>AMQP 0-9-1 protocol (RabbitMQ)</li> 743 <li>Vulkan support</li> 744 <li>avgblur_vulkan, overlay_vulkan, scale_vulkan and chromaber_vulkan filters</li> 745 <li>ADPCM IMA MTF decoder</li> 746 <li>FWSE demuxer</li> 747 <li>DERF DPCM decoder</li> 748 <li>DERF demuxer</li> 749 <li>CRI HCA decoder</li> 750 <li>CRI HCA demuxer</li> 751 <li>overlay_cuda filter</li> 752 <li>switch from AvxSynth to AviSynth+ on Linux</li> 753 <li>mv30 decoder</li> 754 <li>Expanded styling support for 3GPP Timed Text Subtitles (movtext)</li> 755 <li>WebP parser</li> 756 <li>tmedian filter</li> 757 <li>maskedthreshold filter</li> 758 <li>Support for muxing pcm and pgs in m2ts</li> 759 <li>Cunning Developments ADPCM decoder</li> 760 <li>asubboost filter</li> 761 <li>Pro Pinball Series Soundbank demuxer</li> 762 <li>pcm_rechunk bitstream filter</li> 763 <li>scdet filter</li> 764 <li>NotchLC decoder</li> 765 <li>gradients source video filter</li> 766 <li>MediaFoundation encoder wrapper</li> 767 <li>untile filter</li> 768 <li>Simon & Schuster Interactive ADPCM encoder</li> 769 <li>PFM decoder</li> 770 <li>dblur video filter</li> 771 <li>Real War KVAG muxer</li> 772 </ul> 773 <p> 774 We strongly recommend users, distributors, and system integrators to 775 upgrade unless they use current git master. 776 </p> 777 778 <h3 id="photosensitivity">October 5th, 2019, Bright Lights</h3> 779 <p> 780 FFmpeg has added a realtime bright flash removal filter to libavfilter. 781 </p> 782 <p> 783 Note that this filter is not FDA approved, nor are we medical professionals. 784 Nor has this filter been tested with anyone who has photosensitive epilepsy. 785 FFmpeg and its photosensitivity filter are not making any medical claims. 786 </p> 787 <p> 788 That said, this is a new video filter that may help photosensitive people 789 watch tv, play video games or even be used with a VR headset to block 790 out epiletic triggers such as filtered sunlight when they are outside. 791 Or you could use it against those annoying white flashes on your tv screen. 792 The filter fails on some input, such as the 793 <a href="https://www.youtube.com/watch?v=8L_9hXnUzRk">Incredibles 2 Screen Slaver</a> 794 scene. It is not perfect. If you have other clips that you want this filter to 795 work better on, please report them to us on our <a href="http://trac.ffmpeg.org">trac</a>.
796 </p> 797 <p> 798 <a href="http://ffmpeg.org/~compn/output20p8.mp4">See for yourself</a>. 799 Example was made with -vf photosensitivity=20:0.8 800 </p> 801 <p> 802 We are not professionals. Please use this in your medical studies to 803 advance epilepsy research. If you decide to use this in a medical 804 setting, or make a hardware hdmi input output realtime tv filter, 805 or find another use for this, <a href="mailto:[email protected]">please let me know</a>. 806 This filter was a feature request of mine 807 <a href="https://trac.ffmpeg.org/ticket/2104">since 2013</a>. 808 </p> 809 810 <h3 id="pr4.2">August 5th, 2019, FFmpeg 4.2 "Ada"</h3> 811 <p> 812 <a href="download.html#release_4.2">FFmpeg 4.2 "Ada"</a>, a new 813 major release, is now available! Some of the highlights: 814 </p> 815 <ul> 816 <li>tpad filter</li> 817 <li>AV1 decoding support through libdav1d</li> 818 <li>dedot filter</li> 819 <li>chromashift and rgbashift filters</li> 820 <li>freezedetect filter</li> 821 <li>truehd_core bitstream filter</li> 822 <li>dhav demuxer</li> 823 <li>PCM-DVD encoder</li> 824 <li>GIF parser</li> 825 <li>vividas demuxer</li> 826 <li>hymt decoder</li> 827 <li>anlmdn filter</li> 828 <li>maskfun filter</li> 829 <li>hcom demuxer and decoder</li> 830 <li>ARBC decoder</li> 831 <li>libaribb24 based ARIB STD-B24 caption support (profiles A and C)</li> 832 <li>Support decoding of HEVC 4:4:4 content in nvdec and cuviddec</li> 833 <li>removed libndi-newtek</li> 834 <li>agm decoder</li> 835 <li>KUX demuxer</li> 836 <li>AV1 frame split bitstream filter</li> 837 <li>lscr decoder</li> 838 <li>lagfun filter</li> 839 <li>asoftclip filter</li> 840 <li>Support decoding of HEVC 4:4:4 content in vdpau</li> 841 <li>colorhold filter</li> 842 <li>xmedian filter</li> 843 <li>asr filter</li> 844 <li>showspatial multimedia filter</li> 845 <li>VP4 video decoder</li> 846 <li>IFV demuxer</li> 847 <li>derain filter</li> 848 <li>deesser filter</li> 849 <li>mov muxer writes tracks with unspecified language instead of English by default</li> 850 <li>added support for using clang to compile CUDA kernels</li> 851 </ul> 852 <p> 853 We strongly recommend users, distributors, and system integrators to 854 upgrade unless they use current git master. 855 </p> 856 857 <h3 id="pr4.1">November 6th, 2018, FFmpeg 4.1 "al-Khwarizmi"</h3> 858 <p> 859 <a href="download.html#release_4.1">FFmpeg 4.1 "al-Khwarizmi"</a>, a new 860 major release, is now available! Some of the highlights: 861 </p> 862 <ul> 863 <li>deblock filter</li> 864 <li>tmix filter</li> 865 <li>amplify filter</li> 866 <li>fftdnoiz filter</li> 867 <li>aderivative and aintegral audio filters</li> 868 <li>pal75bars and pal100bars video filter sources</li> 869 <li>mbedTLS based TLS support</li> 870 <li>adeclick and adeclip filters</li> 871 <li>libtensorflow backend for DNN based filters like srcnn</li> 872 <li>VC1 decoder is now bit-exact</li> 873 <li>ATRAC9 decoder</li> 874 <li>lensfun wrapper filter</li> 875 <li>colorconstancy filter</li> 876 <li>AVS2 video decoder via libdavs2</li> 877 <li>IMM4 video decoder</li> 878 <li>Brooktree ProSumer video decoder</li> 879 <li>MatchWare Screen Capture Codec decoder</li> 880 <li>WinCam Motion Video decoder</li> 881 <li>1D LUT filter (lut1d)</li> 882 <li>RemotelyAnywhere Screen Capture decoder</li> 883 <li>cue and acue filters</li> 884 <li>Support for AV1 in MP4 and Matroska/WebM</li> 885 <li>transpose_npp filter</li> 886 <li>AVS2 video encoder via libxavs2</li> 887 <li>amultiply filter</li> 888 <li>Block-Matching 3d (bm3d) denoising filter</li> 889 <li>acrossover filter</li> 890 <li>ilbc decoder</li> 891 <li>audio denoiser as afftdn filter</li> 892 <li>AV1 parser</li> 893 <li>sinc audio filter source</li> 894 <li>chromahold filter</li> 895 <li>setparams filter</li> 896 <li>vibrance filter</li> 897 <li>S12M timecode decoding in h264</li> 898 <li>xstack filter</li> 899 <li>(a)graphmonitor filter</li> 900 <li>yadif_cuda filter</li> 901 </ul> 902 <p> 903 We strongly recommend users, distributors, and system integrators to 904 upgrade unless they use current git master. 905 </p> 906 907 <h3 id="pr4.0">April 20th, 2018, FFmpeg 4.0 "Wu"</h3> 908 <p> 909 <a href="download.html#release_4.0">FFmpeg 4.0 "Wu"</a>, a new 910 major release, is now available! Some of the highlights: 911 </p> 912 <ul> 913 <li>Bitstream filters for editing metadata in H.264, HEVC and MPEG-2 streams</li> 914 <li>Experimental MagicYUV encoder</li> 915 <li>TiVo ty/ty+ demuxer</li> 916 <li>Intel QSV-accelerated MJPEG encoding</li> 917 <li>native aptX and aptX HD encoder and decoder</li> 918 <li>NVIDIA NVDEC-accelerated H.264, HEVC, MJPEG, MPEG-1/2/4, VC1, VP8/9 hwaccel decoding</li> 919 <li>Intel QSV-accelerated overlay filter</li> 920 <li>mcompand audio filter</li> 921 <li>acontrast audio filter</li> 922 <li>OpenCL overlay filter</li> 923 <li>video mix filter</li> 924 <li>video normalize filter</li> 925 <li>audio lv2 wrapper filter</li> 926 <li>VAAPI MJPEG and VP8 decoding</li> 927 <li>AMD AMF H.264 and HEVC encoders</li> 928 <li>video fillborders filter</li> 929 <li>video setrange filter</li> 930 <li>support LibreSSL (via libtls)</li> 931 <li>
931Dropped support for building for Windows XP. The minimum supported Windows version is Windows Vista.</li> 932 <li>deconvolve video filter</li> 933 <li>entropy video filter</li> 934 <li>hilbert audio filter source</li> 935 <li>aiir audio filter</li> 936 <li>Removed the ffserver program</li> 937 <li>Removed the ffmenc and ffmdec muxer and demuxer</li> 938 <li>VideoToolbox HEVC encoder and hwaccel</li> 939 <li>VAAPI-accelerated ProcAmp (color balance), denoise and sharpness filters</li> 940 <li>Add android_camera indev</li> 941 <li>codec2 en/decoding via libcodec2</li> 942 <li>native SBC encoder and decoder</li> 943 <li>drmeter audio filter</li> 944 <li>hapqa_extract bitstream filter</li> 945 <li>filter_units bitstream filter</li> 946 <li>AV1 Support through libaom</li> 947 <li>E-AC-3 dependent frames support</li> 948 <li>bitstream filter for extracting E-AC-3 core</li> 949 <li>Haivision SRT protocol via libsrt</li> 950 <li>vfrdet filter</li> 951 </ul> 952 <p> 953 We strongly recommend users, distributors, and system integrators to 954 upgrade unless they use current git master. 955 </p> 956 957 <h3 id="pr3.4">October 15th, 2017, FFmpeg 3.4 "Cantor"</h3> 958 <p> 959 <a href="download.html#release_3.4">FFmpeg 3.4 "Cantor"</a>, a new 960 major release, is now available! Some of the highlights: 961 </p> 962 <ul> 963 <li>deflicker video filter</li> 964 <li>doubleweave video filter</li> 965 <li>lumakey video filter</li> 966 <li>pixscope video filter</li> 967 <li>oscilloscope video filter</li> 968 <li>update cuvid/nvenc headers to Video Codec SDK 8.0.14</li> 969 <li>afir audio filter</li> 970 <li>scale_cuda CUDA based video scale filter</li> 971 <li>librsvg support for svg rasterization</li> 972 <li>crossfeed audio filter</li> 973 <li>spec compliant VP9 muxing support in MP4</li> 974 <li>surround audio filter</li> 975 <li>sofalizer filter switched to libmysofa</li> 976 <li>Gremlin Digital Video demuxer and decoder</li> 977 <li>headphone audio filter</li> 978 <li>superequalizer audio filter</li> 979 <li>roberts video filter</li> 980 <li>additional frame format support for Interplay MVE movies</li> 981 <li>support for decoding through D3D11VA in ffmpeg</li> 982 <li>limiter video filter</li> 983 <li>libvmaf video filter</li> 984 <li>Dolby E decoder and SMPTE 337M demuxer</li> 985 <li>unpremultiply video filter</li> 986 <li>tlut2 video filter</li> 987 <li>floodfill video filter</li> 988 <li>pseudocolor video filter</li> 989 <li>raw G.726 muxer and demuxer, left- and right-justified</li> 990 <li>NewTek NDI input/output device</li> 991 <li>FITS demuxer and decoder</li> 992 <li>FITS muxer and encoder</li> 993 <li>despill video filter</li> 994 <li>haas audio filter</li> 995 <li>SUP/PGS subtitle muxer</li> 996 <li>convolve video filter</li> 997 <li>VP9 tile threading support</li> 998 <li>KMS screen grabber</li> 999 <li>CUDA thumbnail filter</li> 1000 <li>V4L2 mem2mem HW assisted codecs</li> 1001 <li>Rockchip MPP hardware decoding</li> 1002 <li>vmafmotion video filter</li> 1003 </ul> 1004 <p> 1005 We strongly recommend users, distributors, and system integrators to 1006 upgrade unless they use current git master. 1007 </p> 1008 1009 <h3 id="pr3.3">April 13th, 2017, FFmpeg 3.3 "Hilbert"</h3> 1010 <p> 1011 <a href="download.html#release_3.3">FFmpeg 3.3 "Hilbert"</a>, a new 1012 major release, is now available! Some of the highlights: 1013 </p> 1014 <ul> 1015 <li>Apple Pixlet decoder</li> 1016 <li>NewTek SpeedHQ decoder</li> 1017 <li>QDMC audio decoder</li> 1018 <li>PSD (Photoshop Document) decoder</li> 1019 <li>FM Screen Capture decoder</li> 1020 <li>ScreenPressor decoder</li> 1021 <li>XPM decoder</li> 1022 <li>DNxHR decoder fixes for HQX and high resolution videos</li> 1023 <li>ClearVideo decoder (partial)</li> 1024 <li>16.8 and 24.0 floating point PCM decoder</li> 1025 <li>Intel QSV-accelerated VP8 video decoding</li> 1026 <li>native Opus encoder</li> 1027 <li>DNxHR 444 and HQX encoding</li> 1028 <li>Quality improvements for the (M)JPEG encoder</li> 1029 <li>VAAPI-accelerated MPEG-2 and VP8 encoding</li> 1030 <li>premultiply video filter</li> 1031 <li>abitscope multimedia filter</li> 1032 <li>readeia608 filter</li> 1033 <li>threshold filter</li> 1034 <li>midequalizer filter</li> 1035 <li>MPEG-7 Video Signature filter</li> 1036 <li>add internal ebur128 library, remove external libebur128 dependency</li> 1037 <li>Intel QSV video scaling and deinterlacing filters</li> 1038 <li>Sample Dump eXchange demuxer</li> 1039 <li>MIDI Sample Dump Standard demuxer</li> 1040 <li>Scenarist Closed Captions demuxer and muxer</li> 1041 <li>Support MOV with multiple sample description tables</li> 1042 <li>Pro-MPEG CoP #3-R2 FEC protocol</li> 1043 <li>Support for spherical videos</li> 1044 <li>CrystalHD decoder moved to new decode API</li> 1045 <li>configure now fails if autodetect-libraries are requested but not found</li> 1046 </ul> 1047 <p> 1048 We strongly recommend users, distributors, and system integrators to 1049 upgrade unless they use current git master. 1050 </p> 1051 1052 <h3 id="gsoc2016finalreport">October 30th, 2016, Results: Summer Of Code 2016.</h3> 1053 <p> 1054 This has been a long time coming but we wanted to give a proper closure to our participation in this run of the program and it takes time. Sometimes it's just to get the final report for each project trimmed down, others, is finalizing whatever was still in progress when the program finished: final patches need to be merged, TODO lists stabilized, future plans agreed; you name it. 1055 </p> 1056 <p> 1057 Without further ado, here's the silver-lining for each one of the projects we sought to complete during this Summer of Code season: 1058 </p> 1059 <h4>FFv1 (Mentor: Michael Niedermayer)</h4> 1060 <p> 1061 Stanislav Dolganov designed and implemented experimental support for motion estimation and compensation in the lossless FFV1 codec. The design and implementation is based on the snow video codec, which uses OBMC. Stanislav's work proved that significant compression gains can be achieved with inter frame c
1061ompression. FFmpeg welcomes Stanislav to continue working beyond this proof of concept and bring its advances into the official FFV1 specification within the IETF. 1062 </p> 1063 <h4>Self test coverage (Mentor: Michael Niedermayer)</h4> 1064 <p> 1065 Petru Rares Sincraian added several self-tests to FFmpeg and successfully went through the in-some-cases tedious process of fine tuning tests parameters to avoid known and hard to avoid problems, like checksum mismatches due to rounding errors on the myriad of platforms we support. His work has improved the code coverage of our self tests considerably. 1066 </p> 1067 <h4>MPEG-4 ALS encoder implementation (Mentor: Thilo Borgmann)</h4> 1068 <p> 1069 Umair Khan updated and integrated the ALS encoder to fit in the current FFmpeg codebase. He also implemented a missing feature for the ALS decoder that enables floating-point sample decoding. FFmpeg support for MPEG-4 ALS has been improved significantly by Umair's work. We welcome him to keep maintaining his improvements and hope for great contributions to come. 1070 </p> 1071 <h4>Tee muxer improvements (Mentor: Marton Balint)</h4> 1072 <p> 1073 Ján Sebechlebský's generic goal was to improve the tee muxer so it tolerated blocking IO and allowed transparent error recovery. During the design phase it turned out that this functionality called for a separate muxer, so Ján spent his summer working on the so-called FIFO muxer, gradually fixing issues all over the codebase. He succeeded in his task, and the FIFO muxer is now part of the main repository, alongside several other improvements he made in the process. 1074 </p> 1075 <h4>TrueHD encoder (Mentor: Rostislav Pehlivanov)</h4> 1076 <p> 1077 Jai Luthra's objective was to update the out-of-tree and pretty much abandoned MLP (Meridian Lossless Packing) encoder for libavcodec and improve it to enable encoding to the TrueHD format. For the qualification period the encoder was updated such that it was usable and throughout the summer, successfully improved adding support for multi-channel audio and TrueHD encoding. Jai's code has been merged into the main repository now. While a few problems remain with respect to LFE channel and 32 bit sample handling, these are in the process of being fixed such that effort can be finally put in improving the encoder's speed and efficiency. 1078 </p> 1079 <h4>Motion interpolation filter (Mentor: Paul B Mahol)</h4> 1080 <p> 1081 Davinder Singh investigated existing motion estimation and interpolation approaches from the available literature and previous work by our own: Michael Niedermayer, and implemented filters based on this research. These filters allow motion interpolating frame rate conversion to be applied to a video, for example, to create a slow motion effect or change the frame rate while smoothly interpolating the video along the motion vectors. There's still work to be done to call these filters 'finished', which is rather hard all things considered, but we are looking optimistically at their future. 1082 </p> 1083 <p> 1084 And that's it. We are happy with the results of the program and immensely thankful for the opportunity of working with such an amazing set of students. We can be a tough crowd but our mentors did an amazing job at hand holding our interns through their journey. Thanks also to Google for this wonderful program and to everyone that made room in their busy lives to help making GSoC2016 a success. See you in 2017! 1085 </p> 1086 <h3 id="sdl1">September 24th, 2016, SDL1 support dropped.</h3> 1087 <p> 1088 Support for the SDL1 library has been dropped, due to it no longer being maintained (as of 1089 January, 2012) and it being superseded by the SDL2 library. As a result, the SDL1 output device 1090 has also been removed and replaced by an SDL2 implementation. Both the ffplay and opengl output 1091 devices have been updated to support SDL2. 1092 </p> 1093 <h3 id="pr3.1.2">August 9th, 2016, FFmpeg 3.1.2 "Laplace"</h3> 1094 <p> 1095 <a href="download.html#release_3.1">FFmpeg 3.1.2</a>, a new point release from the 3.1 release branch, is now available! 1096 It fixes several bugs. 1097 </p> 1098 <p> 1099 We recommend users, distributors, and system integrators, to upgrade unless they use current git master. 1100 </p> 1101 <h3 id="ffserv">July 10th, 2016, ffserver program being dropped</h3> 1102 <p> 1103 After thorough deliberation, we're announcing that we're about to drop the ffserver program from the project starting with the next release. 1104 ffserver has been a problematic program to maintain due to its use of internal APIs, which complicated the recent cleanups to the libavformat 1105 library, and block further cleanups and improvements which are desired by API users and will be easier to maintain. Furthermore the program has 1106 been hard for users to deploy and run due to reliability issues, lack of knowledgable people to help and confusing configuration file syntax. 1107 Current users and members of the community are invited to write a replacement program to fill the same niche that ffserver did using the new APIs 1108 and to contact us so we may point users to test and contribute to its development. 1109 </p> 1110 <h3 id="pr3.1.1">July 1st, 2016, FFmpeg 3.1.1 "Laplace"</h3> 1111 <p> 1112 <a href="download.html#release_3.1">FFmpeg 3.1.1</a>, a new point release from the 3.1 release branch, is now available! 1113 It mainly deals with a few ABI issues introduced in the previous release. 1114 </p> 1115 <p> 1116 We strongly recommend users, distributors, and system integrators, especially those who experienced issues upgrading from 3.0, to 1117 upgrade unless they use current git master. 1118 </p> 1119 1120 <h3 id="pr3.1">June 27th, 2016, FFmpeg 3.1 "Laplace"</h3> 1121 <p> 1122 <a href="download.html#release_3.1">FFmpeg 3.1 "Laplace"</a>, a new 1123 major release, is now available! Some of the highlights: 1124 </p> 1125 <ul> 1126 <li>DXVA2-accelerated HEVC Main10 decoding</li> 1127 <li>fieldhint filter</li> 1128 <li>loop video filter and aloop audio filter</li> 1129 <li>Bob Weaver deinterlacing filter</li> 1130 <li>firequalizer filter</li> 1131 <li>datascope filter</li> 1132 <li>bench and abench filters</li> 1133 <li>ciescope filter</li> 1134 <li>
1134protocol blacklisting API</li> 1135 <li>MediaCodec H264 decoding</li> 1136 <li>VC-2 HQ RTP payload format (draft v1) depacketizer and packetizer</li> 1137 <li>VP9 RTP payload format (draft v2) packetizer</li> 1138 <li>AudioToolbox audio decoders</li> 1139 <li>AudioToolbox audio encoders</li> 1140 <li>coreimage filter (GPU based image filtering on OSX)</li> 1141 <li>libdcadec removed</li> 1142 <li>bitstream filter for extracting DTS core</li> 1143 <li>ADPCM IMA DAT4 decoder</li> 1144 <li>musx demuxer</li> 1145 <li>aix demuxer</li> 1146 <li>remap filter</li> 1147 <li>hash and framehash muxers</li> 1148 <li>colorspace filter</li> 1149 <li>hdcd filter</li> 1150 <li>readvitc filter</li> 1151 <li>VAAPI-accelerated format conversion and scaling</li> 1152 <li>libnpp/CUDA-accelerated format conversion and scaling</li> 1153 <li>Duck TrueMotion 2.0 Real Time decoder</li> 1154 <li>Wideband Single-bit Data (WSD) demuxer</li> 1155 <li>VAAPI-accelerated H.264/HEVC/MJPEG encoding</li> 1156 <li>DTS Express (LBR) decoder</li> 1157 <li>Generic OpenMAX IL encoder with support for Raspberry Pi</li> 1158 <li>IFF ANIM demuxer & decoder</li> 1159 <li>Direct Stream Transfer (DST) decoder</li> 1160 <li>loudnorm filter</li> 1161 <li>MTAF demuxer and decoder</li> 1162 <li>MagicYUV decoder</li> 1163 <li>OpenExr improvements (tile data and B44/B44A support)</li> 1164 <li>BitJazz SheerVideo decoder</li> 1165 <li>CUDA CUVID H264/HEVC decoder</li> 1166 <li>10-bit depth support in native utvideo decoder</li> 1167 <li>libutvideo wrapper removed</li> 1168 <li>YUY2 Lossless Codec decoder</li> 1169 <li>VideoToolbox H.264 encoder</li> 1170 </ul> 1171 <p> 1172 We strongly recommend users, distributors, and system integrators to 1173 upgrade unless they use current git master. 1174 </p> 1175 1176 <h3 id="gsoc2016">March 16th, 2016, Google Summer of Code</h3> 1177 <p> 1178 FFmpeg has been accepted as a <a href="https://summerofcode.withgoogle.com/">Google Summer of Code</a> open source organization. If you wish to 1179 participate as a student see our <a href="https://trac.ffmpeg.org/wiki/SponsoringPrograms/GSoC/2016">project ideas page</a>. 1180 You can already get in contact with mentors and start working on qualification tasks as well as register at google and submit your project proposal draft. 1181 Good luck! 1182 </p> 1183 1184 <h3 id="pr3.0">February 15th, 2016, FFmpeg 3.0 "Einstein"</h3> 1185 <p> 1186 <a href="download.html#release_3.0">FFmpeg 3.0 "Einstein"</a>, a new 1187 major release, is now available! Some of the highlights: 1188 </p> 1189 <ul> 1190 <li><a href="#aac_encoder_stable">The native FFmpeg AAC encoder has seen extensive improvements and is no longer considered experimental</a></li> 1191 <li><a href="#removing_external_aac_encoders">Removed support for libvo-aacenc and libaacplus</a></li> 1192 <li>Over 30 new filters have been added</li> 1193 <li>Many ASM optimizations</li> 1194 <li>VP9 Hardware Acceleration (DXVA2 and VA-API)</li> 1195 <li>Cineform HD decoder</li> 1196 <li>New DCA decoder based on libdcadec with full support for DTS-HD extensions</li> 1197 <li>As with all major releases expect major backward incompatible API/ABI changes</li> 1198 <li>See the <a href="https://git.ffmpeg.org/gitweb?p=ffmpeg.git;a=blob_plain;f=Changelog;hb=n3.0">Changelog</a> for a list of more updates</li> 1199 </ul> 1200 <p> 1201 We strongly recommend users, distributors, and system integrators to 1202 upgrade unless they use current git master. 1203 </p> 1204 1205 <h3 id="removing_external_aac_encoders">January 30, 2016, Removing support for two external AAC encoders</h3> 1206 <p> 1207 We have just removed support for VisualOn AAC encoder (libvo-aacenc) and 1208 libaacplus in FFmpeg master. 1209 </p> 1210 <p> 1211 Even before marking our internal AAC encoder as 1212 <a href="#aac_encoder_stable">stable</a>, it was known that libvo-aacenc 1213 was of an inferior quality compared to our native one for most samples. 1214 However, the VisualOn encoder was used extensively by the Android Open 1215 Source Project, and we would like to have a tested-and-true stable option 1216 in our code base. 1217 </p> 1218 <p> 1219 When first committed in 2011, libaacplus filled in the gap of encoding 1220 High Efficiency AAC formats (HE-AAC and HE-AACv2), which was not supported 1221 by any of the encoders in FFmpeg at that time. 1222 </p> 1223 <p> 1224 The circumstances for both have changed. After the work spearheaded by 1225 Rostislav Pehlivanov and Claudio Freire, the now-stable FFmpeg native AAC 1226 encoder is ready to compete with much more mature encoders. The Fraunhofer 1227 FDK AAC Codec Library for Android was added in 2012 as the fourth 1228 supported external AAC encoder, and the one with the best quality and the 1229 most features supported, including HE-AAC and HE-AACv2. 1230 </p> 1231 <p> 1232 Therefore, we have decided that it is time to remove libvo-aacenc and 1233 libaacplus. If you are currently using libvo-aacenc, prepare to transition 1234 to the native encoder (<code>aac</code>) when updating to the next version 1235 of FFmpeg. In most cases it is as simple as merely swapping the encoder 1236 name. If you are currently using libaacplus, start using FDK AAC 1237 (<code>libfdk_aac</code>) with an appropriate <code>profile</code> option 1238 to select the exact AAC profile that fits your needs. In both cases, you 1239 will enjoy an audible quality improvement and as well as fewer licensing 1240 headaches. 1241 </p> 1242 <p> 1243 Enjoy! 1244 </p> 1245 1246 <h3 id="pr2.8.5">January 16, 2016, FFmpeg 2.8.5, 2.7.5, 2.6.7, 2.5.10</h3> 1247 <p> 1248 We have made several new point releases (<b><a href="download.html#release_2.8">2.8.5</a>, 1249 <a href="download.html#release_2.7">2.7.5</a>, 1250 <a href="download.html#release_2.6">2.6.7</a>, 1251 <a href="download.html#release_2.5">2.5.10</a></b>). 1252 They fix various bugs, as well as CVE-2016-1897 and CVE-2016-1898. 1253 Please see the changelog for each release for more details. 1254 </p> 1255 <p> 1256 We recommend users, distributors and system integrators to upgrade unless they use 1257 current git master. 1258 </p> 1259 1260 <h3 id="aac_encoder_stable">December 5th, 2015, The native FFmpeg AAC encoder is now stable!</h3> 1261 <p> 1262 After seven years the native FFmpeg AAC encoder has had its experimental flag 1263 removed and declared as ready for general use. The encoder is transparent 1264 at 128kbps for most samples tested with artifacts only appearing in extreme 1265 cases. Subjective quality tests put the encoder to be of equal or greater 1266 quality than most of the other encoders available to the public. 1267 </p> 1268 <p> 1269 Licensing has always been an issue with encoding AAC audio as most of the 1270 encoders have had a license making FFmpeg unredistributable if compiled with 1271 support for them. The fact that there now exists a fully open and truly 1272 free AAC encoder integrated directly within the project means a lot to those 1273 who wish to use accepted and widespread standards. 1274 </p> 1275 <p> 1276 The majority of the work done to bring the encoder up to quality was started 1277 during this year's GSoC by developer Claudio Freire and Rostislav Pehlivanov. 1278 Both continued to work on the encoder with the latter joining as a developer 1279 and mainainer, working on other parts of the project as well. Also, thanks 1280 to <a href="http://d.hatena.ne.jp/kamedo2/">Kamedo2</a> who does comparisons 1281 and tests, the original authors and all past and current contributors to the 1282 encoder. Users are suggested and encouraged to use the encoder and provide 1283 feedback or breakage reports through our <a href="https://trac.ffmpeg.org/">bug tracker</a>.
1284 </p> 1285 1286 <h3 id="thanks_sponsor_0001">October 13th, 2015, Telepoint & MediaHub are now supporting our project</h3> 1287 <p> 1288 A big thank you note goes to our newest supporters: MediaHub and Telepoint. 1289 Both companies have donated a dedicated server with free of charge internet 1290 connectivity. Here is a little bit about them in their own words: 1291 </p> 1292 1293 <ul> 1294 <li> 1295 <p> 1296 <a href="http://www.telepoint.bg/en/">Telepoint</a> is the biggest 1297 carrier-neutral data center in Bulgaria. Located in the heart of Sofia 1298 on a cross-road of many Bulgarian and International networks, the 1299 facility is a fully featured Tier 3 data center that provides flexible 1300 customer-oriented colocation solutions (ranging from a server to a 1301 private collocation hall) and a high level of security. 1302 </p> 1303 </li> 1304 1305 <li> 1306 <p> 1307 MediaHub Ltd. is a Bulgarian IPTV platform and services provider which 1308 uses FFmpeg heavily since it started operating a year ago. <i>"Donating 1309 to help keep FFmpeg online is our way of giving back to the community" 1310 </i>. 1311 </p> 1312 </li> 1313 </ul> 1314 1315 <p> 1316 Thanks Telepoint and MediaHub for their support! 1317 </p> 1318 1319 <h3 id="gsoc2015_result">September 29th, 2015, GSoC 2015 results</h3> 1320 1321 <p> 1322 FFmpeg participated to the latest edition of 1323 the <a href="http://www.google-melange.com/gsoc/homepage/google/gsoc2015">Google 1324 Summer of Code</a> Project. FFmpeg got a total of 8 assigned 1325 projects, and 7 of them were successful. 1326 </p> 1327 1328 <p>We want to thank <a href="https://www.google.com">Google</a>, the 1329 participating students, and especially the mentors who joined this 1330 effort. We're looking forward to participating in the next GSoC 1331 edition! 1332 </p> 1333 1334 <p> 1335 Below you can find a brief description of the final outcome of 1336 each single project. 1337 </p> 1338 1339 <h4>Basic servers for network protocols, mentee: Stephan Holljes, mentor: Nicolas George</h4> 1340 1341 <p> 1342 Stephan Holljes's project for this session of Google Summer of Code was to 1343 implement basic HTTP server features for libavformat, to complement the 1344 already present HTTP client and RTMP and RTSP server code. 1345 </p> 1346 1347 <p> 1348 The first part of the project was to make the HTTP code capable of accepting 1349 a single client; it was completed partly during the qualification period and 1350 partly during the first week of the summer. Thanks to this work, it is now 1351 possible to make a simple HTTP stream using the following commands: 1352 </p> 1353 1354 <pre> 1355 ffmpeg -i /dev/video0 -listen 1 -f matroska \ 1356 -c:v libx264 -preset fast -tune zerolatency http://:8080 1357 ffplay http://localhost:8080/ 1358 </pre> 1359 1360 <p> 1361 The next part of the project was to extend the code to be able to accept 1362 several clients, simultaneously or consecutively. Since libavformat did not 1363 have an API for that kind of task, it was necessary to design one. This part 1364 was mostly completed before the midterm and applied shortly afterwards. 1365 Since the ffmpeg command-line tool is not ready to serve several clients, 1366 the test ground for that new API is an example program serving hard-coded 1367 content. 1368 </p> 1369 1370 <p> 1371 The last and most ambitious part of the project was to update ffserver to 1372 make use of the new API. It would prove that the API is usable to implement 1373 real HTTP servers, and expose the points where more control was needed. By 1374 the end of the summer, a first working patch series was undergoing code 1375 review. 1376 </p> 1377 1378 <h4>Browsing content on the server, mentee: Mariusz SzczepaÅczyk, mentor: Lukasz Marek</h4> 1379 1380 <p> 1381 Mariusz finished an API prepared by the FFmpeg community and implemented 1382 Samba directory listing as qualification task. 1383 </p> 1384 1385 <p> 1386 During the program he extended the API with the possibility to 1387 remove and rename files on remote servers. He completed the 1388 implementation of these features for file, Samba, SFTP, and FTP 1389 protocols. 1390 </p> 1391 1392 <p> 1393 At the end of the program, Mariusz provided a sketch of an 1394 implementation for HTTP directory listening. 1395 </p> 1396 1397 <h4>Directshow digital video capture, mentee: Mate Sebok, mentor: Roger Pack</h4> 1398 1399 <p> 1400 Mate was working on directshow input from digital video sources. He 1401 got working input from ATSC input sources, with specifiable tuner. 1402 </p> 1403 1404 <p> 1405 The code has not been committed, but a patch of it was sent to the 1406 ffmpeg-devel mailing list for future use. 1407 </p> 1408 1409 <p> 1410 The mentor plans on cleaning it up and committing it, at least for the 1411 ATSC side of things. Mate and the mentor are still working trying to 1412 finally figure out how to get DVB working. 1413 </p> 1414 1415 <h4>Implementing full support for 3GPP Timed Text Subtitles, mentee: Niklesh Lalwani, mentor: Philip Langdale</h4> 1416 1417 <p> 1418 Niklesh's project was to expand our support for 3GPP Timed Text 1419 subtitles. This is the native subtitle format for mp4 containers, and 1420 is interesting because it's usually the only subtitle format supported 1421 by the stock playback applications on iOS and Android devices. 1422 </p> 1423 1424 <p> 1425 ffmpeg already had basic support for these subtitles which ignored all 1426 formatting information - it just provided basic plain-text support. 1427 </p> 1428 1429 <p> 1430 Niklesh did work to add support on both the encode and decode side for 1431 text formatting capabilities, such as font size/colour and effects like 1432 bold/italics, highlighting, etc.
1433 </p> 1434 1435 <p> 1436 The main challenge here is that Timed Text handles formatting in a very 1437 different way from most common subtitle formats. It uses a binary 1438 encoding (based on mp4 boxes, naturally) and stores information 1439 separately from the text itself. This requires additional work to track 1440 which parts of the text formatting applies to, and explicitly dealing 1441 with overlapping formatting (which other formats support but Timed 1442 Text does not) so it requires breaking the overlapping sections into 1443 separate non-overlapping ones with different formatting. 1444 </p> 1445 1446 <p> 1447 Finally, Niklesh had to be careful about not trusting any size 1448 information in the subtitles - and that's no joke: the now infamous 1449 Android stagefright bug was in code for parsing Timed Text subtitles. 1450 </p> 1451 1452 <p> 1453 All of Niklesh's work is committed and was released in ffmpeg 2.8. 1454 </p> 1455 1456<h4>libswscale refactoring, mentee: Pedro Arthur, mentors: Michael Niedermayer, Ramiro Polla</h4> 1457 1458 <p> 1459 Pedro Arthur has modularized the vertical and horizontal scalers. 1460 To do this he designed and implemented a generic filter framework
1461 and moved the existing scaler code into it. These changes now allow 1462 easily adding removing, splitting or merging processing steps. 1463 The implementation was benchmarked and several alternatives were 1464 tried to avoid speed loss. 1465 </p> 1466 1467 <p> 1468 He also added gamma corrected scaling support. 1469 An example to use gamma corrected scaling would be: 1470 </p> 1471 1472 <pre> 1473 ffmpeg -i input -vf scale=512:384:gamma=1 output 1474 </pre> 1475 1476 <p> 1477 Pedro has done impressive work considering the short time available, 1478 and he is a FFmpeg committer now. He continues to contribute to 1479 FFmpeg, and has fixed some bugs in libswscale after GSoC has 1480 ended. 1481 </p> 1482 1483 <h4>AAC Encoder Improvements, mentee: Rostislav Pehlivanov, mentor: Claudio Freire</h4> 1484 1485 <p> 1486 Rostislav Pehlivanov has implemented PNS, TNS, I/S coding and main 1487 prediction on the native AAC encoder. Of all those extensions, only 1488 TNS was left in a less-than-usable state, but the implementation has 1489 been pushed (disabled) anyway since it's a good basis for further 1490 improvements. 1491 </p> 1492 1493 <p> 1494 PNS replaces noisy bands with a single scalefactor representing the 1495 energy of that band, gaining in coding efficiency considerably, and 1496 the quality improvements on low bitrates are impressive for such a 1497 simple feature. 1498 </p> 1499 1500 <p> 1501 TNS still needs some polishing, but has the potential to reduce coding 1502 artifacts by applying noise shaping in the temporal domain (something 1503 that is a source of annoying, notable distortion on low-entropy 1504 bands). 1505 </p> 1506 1507 <p> 1508 Intensity Stereo coding (I/S) can double coding efficiency by 1509 exploiting strong correlation between stereo channels, most effective 1510 on pop-style tracks that employ panned mixing. The technique is not as 1511 effective on classic X-Y recordings though. 1512 </p> 1513 1514 <p> 1515 Finally, main prediction improves coding efficiency by exploiting 1516 correlation among successive frames. While the gains have not been 1517 huge at this point, Rostislav has remained active even after the GSoC, 1518 and is polishing both TNS and main prediction, as well as looking for 1519 further improvements to make. 1520 </p> 1521 1522 <p> 1523 In the process, the MIPS port of the encoder was broken a few times, 1524 something he's also working to fix. 1525 </p> 1526 1527 <h4>Animated Portable Network Graphics (APNG), mentee: Donny Yang, mentor: Paul B Mahol</h4> 1528 1529 <p> 1530 Donny Yang implemented basic keyframe only APNG encoder as the 1531 qualification task. Later he wrote interframe compression via 1532 various blend modes. The current implementation tries all blend 1533 modes and picks one which takes the smallest amount of memory. 1534 </p> 1535 1536 <p> 1537 Special care was taken to make sure that the decoder plays 1538 correctly all files found in the wild and that the encoder 1539 produces files that can be played in browsers that support APNG. 1540 </p> 1541 1542 <p> 1543 During his work he was tasked to fix any encountered bug in the 1544 decoder due to the fact that it doesn't match APNG 1545 specifications. Thanks to this work, a long standing bug in the 1546 PNG decoder has been fixed. 1547 </p> 1548 1549 <p> 1550 For latter work he plans to continue working on the encoder, 1551 making it possible to select which blend modes will be used in the 1552 encoding process. This could speed up encoding of APNG files. 1553 </p> 1554 1555 <h3 id="pr2.8">September 9th, 2015, FFmpeg 2.8</h3> 1556 <p> 1557 We published release <b><a href="download.html#release_2.8">2.8</a></b> as new major version. 1558 It contains all features and bug fixes of the git master branch from September 8th. Please see 1559 the <b><a href="https://raw.githubusercontent.com/FFmpeg/FFmpeg/release/2.8/Changelog">changelog</a></b> 1560 for a list of the most important changes. 1561 </p> 1562 <p> 1563 We recommend users, distributors and system integrators to upgrade unless they use current git master. 1564 </p> 1565 1566 <h3 id="message">August 1st, 2015, A message from the FFmpeg project</h3> 1567 <p> 1568 Dear multimedia community, 1569 </p> 1570 <p> 1571 The resignation of Michael Niedermayer as leader of FFmpeg yesterday has 1572 come by surprise. He has worked tirelessly on the FFmpeg project for many 1573 years and we must thank him for the work that he has done. We hope that in
1574 the future he will continue to contribute to the project. In the coming 1575 weeks, the FFmpeg project will be managed by the active contributors. 1576 </p> 1577 <p> 1578 The last four years have not been easy for our multimedia community - both 1579 contributors and users. We should now look to the future, try to find 1580 solutions to these issues, and to have reconciliation between the forks, 1581 which have split the community for so long. 1582 </p> 1583 <p> 1584 Unfortunately, much of the disagreement has taken place in inappropriate 1585 venues so far, which has made finding common ground and solutions 1586 difficult. We aim to discuss this in our communities online over the coming 1587 weeks, and in person at the <a 1588 href="https://www.videolan.org/videolan/events/vdd15/">VideoLAN Developer 1589 Days</a> in Paris in September: a neutral venue for the entire open source 1590 multimedia community. 1591 </p> 1592 <p> 1593 The FFmpeg project. 1594 </p> 1595 1596 <h3 id="needhost">July 4th, 2015, FFmpeg needs a new host</h3> 1597 <p><b>UPDATE:</b> We have received more than 7 offers for hosting and servers, thanks a lot to everyone!</p> 1598 <p> 1599 After graciously hosting our projects (<a href="http://www.ffmpeg.org">FFmpeg</a>, <a href="http://www.mplayerhq.hu">MPlayer</a> 1600 and <a href="http://rtmpdump.mplayerhq.hu">rtmpdump</a>) for 4 years, Arpi (our hoster) has informed us that we have to secure a new host somewhere else immediately. 1601 </p> 1602 <p> 1603 If you want to host an open source project, please let us know, either on <a href="http://ffmpeg.org/mailman/listinfo/ffmpeg-devel">ffmpeg-devel</a> 1604 mailing list or irc.freenode.net #ffmpeg-devel. 1605 </p> 1606 <p> 1607 We use about 4TB of storage and at least 4TB of bandwidth / month for various mailing lists, <a href="http://trac.ffmpeg.org">trac</a>, <a href="http://samples.ffmpeg.org">samples repo</a>, svn, etc. 1608 </p> 1609 1610 <h3 id="pr2.6.1">March 16, 2015, FFmpeg 2.6.1</h3> 1611 <p> 1612 We have made a new major release (<b><a href="download.html#release_2.6">2.6</a></b>) 1613 and now one week afterward 2.6.1. It contains all features and bugfixes of the git master branch from the 6th March. 1614 Please see the <b><a href="https://git.ffmpeg.org/gitweb?p=ffmpeg.git;a=blob;f=RELEASE_NOTES;hb=release/2.6">Release Notes</a></b> for a 1615 list of note-worthy changes. 1616 </p> 1617 <p> 1618 We recommend users, distributors and system integrators to upgrade unless they use 1619 current git master. 1620 </p> 1621 1622 <h3 id="gsoc2015">March 4, 2015, Google Summer of Code</h3> 1623 <p> 1624 FFmpeg has been accepted as a <a href="http://www.google-melange.com/gsoc/homepage/google/gsoc2015">Google Summer of Code</a> Project. If you wish to 1625 participate as a student see our <a href="https://trac.ffmpeg.org/wiki/SponsoringPrograms/GSoC/2015">project ideas page</a>. 1626 You can already get in contact with mentors and start working on qualification tasks. Registration 1627 at Google for students will open March 16th. Good luck! 1628 </p> 1629 1630 <h3 id="clt2015">March 1, 2015, Chemnitzer Linux-Tage</h3> 1631 <p> 1632 We happily announce that FFmpeg will be represented at Chemnitzer Linux-Tage 1633 (CLT) in Chemnitz, Germany. The event will take place on 21st and 22nd of March. 1634 </p> 1635 1636 <p> 1637 More information can be found <a href="https://chemnitzer.linux-tage.de/2015/en/">here</a> 1638 </p> 1639 1640 <p> 1641 We demonstrate usage of FFmpeg, answer your questions and listen to 1642 your problems and wishes. <strong>If you have media files that cannot be 1643 processed correctly with FFmpeg, be sure to have a sample with you 1644 so we can have a look!</strong> 1645 </p> 1646 <p> 1647 For the first time in our CLT history, there will be an <strong>FFmpeg workshop</strong>! 1648 You can read the details <a href="https://chemnitzer.linux-tage.de/2015/de/programm/beitrag/209">here</a>. 1649 The workshop is targeted at FFmpeg beginners. First the basics of 1650 multimedia will be covered. Thereafter you will learn how to use 1651 that knowledge and the FFmpeg CLI tools to analyse and process media 1652 files. The workshop is in German language only and prior registration 1653 is necessary. The workshop will be on Saturday starting at 10 o'clock. 1654 </p> 1655 <p> 1656 We are looking forward to meet you (again)! 1657 </p> 1658 1659 <h3 id="pr2.5">December 5, 2014, FFmpeg 2.5</h3> 1660 <p> 1661 We have made a new major release (<b><a href="download.html#release_2.5">2.5</a></b>) 1662 It contains all features and bugfixes of the git master branch from the 4th December. 1663 Please see the <b><a href="https://git.ffmpeg.org/gitweb?p=ffmpeg.git;a=blob;f=RELEASE_NOTES;hb=release/2.5">Release Notes</a></b> for a 1664 list of note-worthy changes. 1665 </p> 1666 <p> 1667 We recommend users, distributors and system integrators to upgrade unless they use 1668 current git master. 1669 </p> 1670 1671 <h3 id="ffmpeg_back_in_sid">October 10, 2014, FFmpeg is in Debian unstable again</h3> 1672 <p>
1673 We wanted you to know there are 1674 <a href="https://packages.debian.org/search?keywords=ffmpeg&searchon=sourcenames&suite=unstable&section=main"> 1675 FFmpeg packages in Debian unstable</a> again. <strong>A big thank-you 1676 to Andreas Cadhalpun and all the people that made it possible.</strong> It has been anything but simple. 1677 </p> 1678 <p> 1679 Unfortunately that was already the easy part of this news. The bad news is the packages probably won't 1680 migrate to Debian testing to be in the upcoming release codenamed jessie. 1681 <a href="https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=763148">Read the argumentation over at Debian.</a> 1682 </p> 1683 <p> 1684 <strong>However things will come out in the end, we hope for your continued remarkable support!</strong> 1685 </p> 1686 1687 <h3 id="opw03">October 8, 2014, FFmpeg secured a place in OPW!</h3> 1688 <p> 1689 Thanks to a generous 6K USD donation by Samsung (Open Source Group), 1690 FFmpeg will be welcoming at least 1 "Outreach Program for Women" intern 1691 to work with our community for an initial period starting December 2014 1692 (through March 2015). 1693 </p> 1694 1695 <p> 1696 We all know FFmpeg is used by the industry, but even while there are 1697 countless products building on our code, it is not at all common for 1698 companies to step up and help us out when needed. So a big thank-you 1699 to Samsung and the OPW program committee! 1700 </p> 1701 1702 <p> 1703 If you are thinking on participating in OPW as an intern, please take 1704 a look at our <a href="https://trac.ffmpeg.org/wiki/SponsoringPrograms/OPW/2014-12">OPW wiki page</a> 1705 for some initial guidelines. The page is still a work in progress, but 1706 there should be enough information there to get you started. If you, on 1707 the other hand, are thinking on sponsoring work on FFmpeg through the 1708 OPW program, please get in touch with us at [email protected]. With your 1709 help, we might be able to secure some extra intern spots for this round! 1710 </p> 1711 1712 <h3 id="pr2.4">September 15, 2014, FFmpeg 2.4</h3> 1713 <p> 1714 We have made a new major release (<b><a href="download.html#release_2.4">2.4</a></b>) 1715 It contains all features and bugfixes of the git master branch from the 14th September. 1716 Please see the <b><a href="https://git.ffmpeg.org/gitweb?p=ffmpeg.git;a=blob;f=RELEASE_NOTES;hb=release/2.4">Release Notes</a></b> for a 1717 list of note-worthy changes. 1718 </p> 1719 <p> 1720 We recommend users, distributors and system integrators to upgrade unless they use 1721 current git master. 1722 </p> 1723 1724 <h3 id="pr2.3.3">August 20, 2014, FFmpeg 2.3.3, 2.2.7, 1.2.8</h3> 1725 <p> 1726 We have made several new point releases (<b><a href="download.html#release_2.3">2.3.3</a>, 1727 <a href="download.html#release_2.2">2.2.7</a>, 1728 <a href="download.html#release_1.2">1.2.8</a></b>). 1729 They fix various bugs, as well as CVE-2014-5271 and CVE-2014-5272. 1730 Please see the changelog for more details. 1731 </p> 1732 <p> 1733 We recommend users, distributors and system integrators to upgrade unless they use 1734 current git master. 1735 </p> 1736 1737 <h3 id="opw02">July 29, 2014, Help us out securing our spot in OPW</h3> 1738 <p> 1739 Following our previous post regarding our participation on this year's 1740 OPW (Outreach Program for Women), we are now reaching out to our users 1741 (both individuals and companies) to help us gather the needed money to 1742 secure our spot in the program.<br> 1743 We need to put together 6K USD as a minimum but securing more funds would 1744 help us towards getting more than one intern.<br> 1745 You can donate by credit card using 1746 <a href="https://co.clickandpledge.com/advanced/default.aspx?wid=56226"> 1747 Click&Pledge</a> and selecting the "OPW" option. If you would like to 1748 donate by money transfer or by check, please get in touch by 1749 <a href="mailto:[email protected]">e-mail</a> and we will get back to you 1750 with instructions.<br>Thanks! 1751 </p> 1752 1753 <h3 id="newweb">July 20, 2014, New website</h3> 1754 <p> 1755 The FFmpeg project is proud to announce a brand new version of the website 1756 made by <a href="http://db0.fr">db0</a>. While this was initially motivated 1757 by the need for a larger menu, the whole website ended up being redesigned, 1758 and most pages got reworked to ease navigation. We hope you'll enjoy 1759 browsing it. 1760 </p> 1761 1762 <h3 id="pr2.3">July 17, 2014, FFmpeg 2.3</h3> 1763 <p> 1764 We have made a new major release (<b><a href="download.html#release_2.3">2.3</a></b>) 1765 It contains all features and bugfixes of the git master branch from the 16th July. 1766 Please see the <b><a href="https://git.ffmpeg.org/gitweb?p=ffmpeg.git;a=blob;f=RELEASE_NOTES;hb=489d066">Release Notes</a></b> for a 1767 list of note-worthy changes. 1768 </p> 1769 <p> 1770 We recommend users, distributors and system integrators to upgrade unless they use 1771 current git master. 1772 </p> 1773 1774 <h3 id="opw01">July 3, 2014, FFmpeg and the Outreach Program For Women</h3> 1775 <p> 1776 FFmpeg has started the process to become an OPW includer organization for the 1777 next round of the program, with internships starting December 9. The 1778 <a href="https://gnome.org/opw/">OPW</a> aims to "Help women (cis and trans) 1779 and genderqueer to get involved in free and open source software". Part of the 1780 process requires securing funds to support at least one internship (6K USD), so 1781 if you were holding on your donation to FFmpeg, this is a great chance for you 1782 to come forward, get in touch and help both the project and a great initiative! 1783 </p> 1784 <p> 1785 We have set up an <a href="mailto:[email protected]">email address</a> you can use 1786 to contact us about donations and general inquires regarding our participation 1787 in the program. Hope to hear from you soon! 1788 </p> 1789 1790 <h3 id="pr2.2.4">June 29, 2014, FFmpeg 2.2.4, 2.1.5, 2.0.5, 1.2.7, 1.1.12, 0.10.14</h3> 1791 <p> 1792 We have made several new point releases (<b><a href="download.html#release_2.2">2.2.4</a>, 1793 <a href="download.html#release_2.1">2.1.5</a>, 1794 <a href="download.html#release_2.0">2.0.5</a>, 1795 <a href="download.html#release_1.2">1.2.7</a>, 1796 <a href="download.html#release_1.1">1.1.12</a>, 1797 <a href="download.html#release_0.10">0.10.14</a></b>). 1798 They fix a 1799 <a href="http://blog.securitymouse.com/2014/06/raising-lazarus-20-year-old-bug-that.html">security issue in the LZO implementation</a>, 1800 as well as several other bugs. See the git log for details. 1801 </p> 1802 <p> 1803 We recommend users, distributors and system integrators to upgrade unless they use 1804 current git master. 1805 </p> 1806 1807 1808 <h3 id="lt2014">May 1, 2014, LinuxTag</h3> 1809 <p> 1810 Once again FFmpeg will be represented at LinuxTag in Berlin, Germany. The event will 1811 take place from 8th to 10th of May. Please note that this year's LinuxTag is at a 1812 different location closer to the city center. 1813 </p> 1814 1815 <p> 1816 We will have a shared booth with XBMC and VideoLAN. 1817 <b> 1818 If you have media files that cannot be processed correctly with 1819 FFmpeg, be sure to have a sample with you so we can have a look! 1820 </b> 1821 </p> 1822 1823 <p> 1824 More information about LinuxTag can be found <a href="http://www.linuxtag.org/2014/">here</a> 1825 </p> 1826 1827 <p> 1828 We are looking forward to see you in Berlin! 1829 </p> 1830 1831 <h3 id="heartbleed">April 18, 2014, OpenSSL Heartbeat bug</h3> 1832 <p> 1833 Our server hosting the Trac issue tracker was vulnerable to the attack 1834 against OpenSSL known as "heartbleed". The OpenSSL software library 1835 was updated on 7th of April, shortly after the vulnerability was publicly 1836 disclosed. We have changed the private keys (and certificates) for all 1837 FFmpeg servers. The details were sent to the mailing lists by 1838 Alexander Strasser, who is part of the project server team. Here is a 1839 link to the user mailing list 1840 <a href="https://lists.ffmpeg.org/pipermail/ffmpeg-user/2014-April/020968.html">archive</a> 1841 . 1842 </p><p> 1843 We encourage you to read up on 1844 <a href="https://www.schneier.com/blog/archives/2014/04/heartbleed.html">"OpenSSL heartbleed"</a>. 1845 <b>It is possible that login data for the issue tracker was exposed to 1846 people exploiting this security hole. You might want to change your password 1847 in the tracker and everywhere else you used that same password.</b> 1848 </p> 1849 1850 <h3 id="pr2.2.1">April 11, 2014, FFmpeg 2.2.1</h3> 1851 <p> 1852 We have made a new point releases (<b><a href="download.html#release_2.2">2.2.1</a></b>). 1853 It contains bug fixes for Tickets #2893, #3432, #3469, #3486, #3495 and #3540 as well as 1854 several other fixes. 1855 See the git log for details. 1856 </p> 1857 1858 <h3 id="pr2.2">March 24, 2014, FFmpeg 2.2</h3> 1859 <p> 1860 We have made a new major release (<b><a href="download.html#release_2.2">2.2</a></b>) 1861 It contains all features and bugfixes of the git master branch from 1st March. 1862 A partial list of new stuff is below: 1863 </p> 1864 <pre> 1865 - HNM version 4 demuxer and video decoder 1866 - Live HDS muxer 1867 - setsar/setdar filters now support variables in ratio expressions 1868 - elbg filter 1869 - string validation in ffprobe 1870 - support for decoding through VDPAU in ffmpeg (the -hwaccel option) 1871 - complete Voxware MetaSound decoder 1872 - remove mp3_header_compress bitstream filter
1873 - Windows resource files for shared libraries 1874 - aeval filter 1875 - stereoscopic 3d metadata handling 1876 - WebP encoding via libwebp 1877 - ATRAC3+ decoder 1878 - VP8 in Ogg demuxing 1879 - side & metadata support in NUT 1880 - framepack filter 1881 - XYZ12 rawvideo support in NUT 1882 - Exif metadata support in WebP decoder 1883 - OpenGL device 1884 - Use metadata_header_padding to control padding in ID3 tags (currently used in 1885 MP3, AIFF, and OMA files), FLAC header, and the AVI "junk" block. 1886 - Mirillis FIC video decoder 1887 - Support DNx444 1888 - libx265 encoder 1889 - dejudder filter 1890 - Autodetect VDA like all other hardware accelerations 1891 </pre> 1892 <p> 1893 We recommend users, distributors and system integrators to upgrade unless they use 1894 current git master. 1895 </p> 1896 1897 <h3 id="clt2014">February 3, 2014, Chemnitzer Linux-Tage</h3> 1898 <p> 1899 We happily announce that FFmpeg will be represented at `Chemnitzer Linux-Tage' 1900 in Chemnitz, Germany. The event will take place on 15th and 16th of March. 1901 </p> 1902 1903 <p> 1904 More information can be found <a href="http://chemnitzer.linux-tage.de/2014/en/info/">here</a> 1905 </p> 1906 1907 <p> 1908 We invite you to visit us at our booth located in the Linux-Live area! 1909 There we will demonstrate usage of FFmpeg, answer your questions and listen to 1910 your problems and wishes. 1911 </p> 1912 <p> 1913 <b> 1914 If you have media files that cannot be processed correctly with 1915 FFmpeg, be sure to have a sample with you so we can have a look! 1916 </b> 1917 </p> 1918 <p> 1919 We are looking forward to meet you (again)! 1920 </p> 1921 1922 1923 <h3 id="trac_sec">February 9, 2014, trac.ffmpeg.org / trac.mplayerhq.hu Security Breach</h3> 1924 <p> 1925 The server on which FFmpeg and MPlayer Trac issue trackers were 1926 installed was compromised. The affected server was taken offline 1927 and has been replaced and all software reinstalled. 1928 FFmpeg Git, releases, FATE, web and mailinglists are on other servers 1929 and were not affected. We believe that the original compromise happened 1930 to a server, unrelated to FFmpeg and MPlayer, several months ago. 1931 That server was used as a source to clone the VM that we recently moved 1932 Trac to. It is not known if anyone used the backdoor that was found. 1933 </p> 1934 <p> 1935 We recommend all users to change their passwords. 1936 <b>Especially users who use a password on Trac that they also use 1937 elsewhere, should change that password at least elsewhere.</b> 1938 </p> 1939 1940 1941 <h3 id="ffmpeg_rfp">November 12, 2013, FFmpeg RFP in Debian</h3> 1942 <p> 1943 Since the splitting of Libav the Debian/Ubuntu maintainers have followed 1944 the Libav fork. Many people have requested the packaging of ffmpeg in 1945 Debian, as it is more feature-complete and in many cases less buggy. 1946 </p> 1947 <p> 1948 <a href="http://cynic.cc/blog/">Rogério Brito</a>, a Debian developer, 1949 has proposed a Request For Package (RFP) in the Debian bug tracking 1950 system. 1951 </p> 1952 <p> 1953 Please let the Debian and Ubuntu developers know that you support packaging 1954 of the real FFmpeg! See Debian <a href="http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=729203">ticket #729203</a> 1955 for more details. 1956 </p> 1957 1958 <h3 id="pr2.1">October 28, 2013, FFmpeg 2.1</h3> 1959 <p> 1960 We have made a new major release (<b><a href="download.html#release_2.1">2.1</a></b>) 1961 It contains all features and bugfixes of the git master branch from 28th October. 1962 A partial list of new stuff is below: 1963 </p> 1964 <pre> 1965 - aecho filter 1966 - perspective filter ported from libmpcodecs 1967 - ffprobe -show_programs option 1968 - compand filter 1969 - RTMP seek support 1970 - when transcoding with ffmpeg (i.e. not streamcopying), -ss is now accurate 1971 even when used as an input option. Previous behavior can be restored with 1972 the -noaccurate_seek option. 1973 - ffmpeg -t option can now be used for inputs, to limit the duration of 1974 data read from an input file 1975 - incomplete Voxware MetaSound decoder 1976 - read EXIF metadata from JPEG 1977 - DVB teletext decoder 1978 - phase filter ported from libmpcodecs 1979 - w3fdif filter 1980 - Opus support in Matroska 1981 - FFV1 version 1.3 is stable and no longer experimental 1982 - FFV1: YUVA(444,422,420) 9, 10 and 16 bit support 1983 - changed DTS stream id in lavf mpeg ps muxer from 0x8a to 0x88, to be 1984 more consistent with other muxers. 1985 - adelay filter 1986 - pullup filter ported from libmpcodecs 1987 - ffprobe -read_intervals option 1988 - Lossless and alpha support for WebP decoder 1989 - Error Resilient AAC syntax (ER AAC LC) decoding 1990 - Low Delay AAC (ER AAC LD) decoding 1991 - mux chapters in ASF files 1992 - SFTP protocol (via libssh) 1993 - libx264: add ability to encode in YUVJ422P and YUVJ444P 1994 - Fraps: use BT.709 colorspace by default for yuv, as reference fraps decoder does 1995 - make decoding alpha optional for prores, ffv1 and vp6 by setting 1996 the skip_alpha flag. 1997 - ladspa wrapper filter 1998 - native VP9 decoder 1999 - dpx parser 2000 - max_error_rate parameter in ffmpeg 2001 - PulseAudio output device 2002 - ReplayGain scanner 2003 - Enhanced Low Delay AAC (ER AAC ELD) decoding (no LD SBR support) 2004 - Linux framebuffer output device 2005 - HEVC decoder, raw HEVC demuxer, HEVC demuxing in TS, Matroska and MP4 2006 - mergeplanes filter 2007 </pre> 2008 <p> 2009 We recommend users, distributors and system integrators to upgrade unless they use 2010 current git master. 2011 </p> 2012
2013 <div class="text-right"> 2014 <a href="archive.html" class="btn btn-lg btn-success"><i class="fa fa-chevron-right"></i> Past news</a> 2015 </div> 2016 2017</div> <!-- index --> 2018 2019 <p class="text-right"><small>Hosting provided by <a href="https://telepoint.bg">telepoint.bg</a></small></p> 2020 </div> <!-- page-content-inset --> 2021 </div> <!-- page-content-wrapper --> 2022 </div> <!-- wrapper --> 2023 2024
2024<script src="js/jquery.min.js"></script>
2024 2025
2025<script src="js/bootstrap.min.js"></script>
2025 2026 2027 <!-- Custom JavaScript for the Menu Toggle --> 2028
2028<script> 2029 $("#menu-toggle").click(function(e) { 2030 e.preventDefault(); 2031 $("#wrapper").toggleClass("active"); 2032 }); 2033 </script>
2033 2034 2035 </body> 2036</html>
Line numbers count LF bytes from the start of the resource, as the search results do. Vendor segments are library code the classifier recognised; they are stored but not indexed. Bytes are shown as Latin1 characters, one per byte.