Технологии сжатия и кодеки Windows Media Audio (WMA)

Audio data compression technology
audio encoding and decoding software
Since 2008 Microsoft has also been using WMA Professional in its Protected Interoperable File Format (PIFF) based on the ISO Base Media File Format and most commonly used for Smooth Streaming, a form of adaptive bitrate streaming over HTTP. Related industry standards such as DECE UltraViolet and MPEG DASH have not standardized WMA as a supported audio codec, deciding in favor of the more industry prevalent MPEG and Dolby audio codecs.

Each WMA file features a single audio track in one of the four sub formats: WMA, WMA Pro, WMA Lossless, or WMA Voice. These formats are implemented differently from one another, such that they are technically distinct and mutually incompatible; that is to say, a device or software compatible with one sub format does not therefore automatically support any of the other codecs. Each codec is further explained below.

Windows Media Audio (WMA) is the most common codec of the four WMA codecs. The colloquial usage of the term WMA, especially in marketing materials and device specifications, usually refers to this codec only. The first version of the codec released in 1999 is regarded as WMA 1. In the same year, the bit stream syntax, or compression algorithm, was altered in minor ways and became WMA 2. which reduces latency for encoding and decoding. Fundamentally, WMA is a transform coder based on modified discrete cosine transform (MDCT), somewhat similar to AAC, Cook and Vorbis. The bit stream of WMA is composed of superframes, each containing 1 or more frames of 2048 samples. If the bit reservoir is not used, a frame is equal to a superframe. Each frame contains several blocks, which are 128, 256, 512, 1024, or 2048 samples long after being transformed into the frequency domain via the MDCT. In the frequency domain, masking for the transformed samples is determined, and then used to requantize the samples. Finally, the floating point samples are decomposed into coefficient and exponent parts and independently huffman coded. Stereo information is typically mid/side coded. At low bit rates, line spectral pairs (typically less than 17 kbit/s) and a form of noise coding (typically less than 33 kbit/s) can also be used to improve quality. Like AAC and Ogg Vorbis, WMA was intended to address perceived deficiencies in the MP3 standard. Given their common design goals, the three formats ended up making similar design choices. All three are pure transform codecs. Furthermore, the MDCT implementation used in WMA is essentially a superset of those used in Ogg and AAC such that WMA iMDCT and windowing routines can be used to decode AAC and Ogg Vorbis almost unmodified. However, quantization and stereo coding is handled differently in each codec. The primary distinguishing trait of the WMA Standard format is its unique use of 5 different block sizes, compared to MP3, AAC, and Ogg Vorbis which each restrict files to just two sizes. WMA Pro extends this by adding a 6th block size used at 88.2/96 kHz sampling rate. Certified PlaysForSure devices, as well as a large number of uncertified devices, ranging from portable hand held music players to set top DVD players, support the playback of WMA files. Most PlaysForSure certified online stores distribute content using this codec only. In 2005, Nokia announced its plans to support WMA playback in future Nokia handsets. In the same year, an update was made available for the PlayStation Portable (version 2.60) which allowed WMA files to be played on the device for the first time.

Windows Media Audio Professional (WMA Pro) is an improved lossy codec closely related to WMA standards. It retains most of the same general coding features, but also features improved entropy coding and quantization strategies as well as more efficient stereo coding. Notably, many of the WMA standard's low bitrate features have been removed, as the core codec is designed for efficient coding at most bitrates. Its main competitors include AAC, HE AAC, Vorbis, Dolby Digital, and DTS. It supports 16 bit and 24 bit sample bit depth, sampling rates up to 96 kHz, and up to eight discrete channels (7.1 channel surround). WMA Pro also supports dynamic range compression, which reduces the volume difference between the loudest and quietest sounds in the audio track. According to Microsoft's Amir Majidimehr, WMA Pro could theoretically go beyond 7.1 surround sound and support "an unlimited number of channels"; however, Microsoft chose to limit its current capability to eight (7.1 discrete channels). The codec's bit stream syntax was frozen at the first version, WMA 9 Pro. Later versions of WMA Pro introduced low bit rate encoding, low delay audio, frequency interpolation mode, and an expanded range of sampling rate and bit depth encoding options. A WMA 10 Pro file compressed with frequency interpolation mode comprises a WMA 9 Pro track encoded at half the original sampling rate, which is then restored using a new compression algorithm. In this situation, WMA 9 Pro players which have not been updated to the WMA 10 Pro codec can only decode the lower quality WMA 9 Pro stream. Starting with WMA 10 Pro, eight channel encoding starts at 128 kbit/s, and tracks can be encoded at the native audio CD resolution (44.1 kHz, 16 bit), previously the domain of WMA Standard. Despite a growing number of supported devices and its superiority over WMA, WMA Pro still has little hardware and software support. Some notable exceptions to this are the Microsoft Zune (limited to stereo), Xbox 360, Windows Mobile powered devices with Windows Media Player 10 Mobile, newer Toshiba Gigabeat and Motorola devices, and devices running recent versions of the Rockbox alternative firmware. In addition, WMA Pro is a requirement for the WMV HD certification program. On the software side, Verizon utilizes WMA 10 Pro for its V CAST Music Service, and Windows Media Player 11 has promoted the codec as an alternative to WMA for copying audio CD tracks. WMA Pro is supported in Silverlight as of version 2 (though only in stereo mode). In the absence of the appropriate audio hardware, WMA Pro can automatically downmix multichannel audio to stereo or mono, and 24 bit resolution to 16 bit during playback. A notable example of WMA Pro being used instead of WMA Standard is the NBC Olympics website which uses WMA 10 Pro in its low bitrate mode at 48 kbit/s.

Windows Media Audio 9 Lossless is a lossless incarnation of Windows Media Audio, an audio codec by Microsoft, released in early 2003. It compresses an audio CD to a range of 206 to 411 MB, at bit rates of 470 to 940 kbit/s. The result is a bit for bit duplicate of the original audio file; in other words, the audio quality on the CD will be the same as the file when played back. WMA Lossless uses the same WMA file extension as other Windows Media Audio formats. It supports 6 discrete channels and up to 24 bit/96 kHz lossless audio. The format has never been publicly documented, although an open source decoder has been reverse engineered for non Microsoft platforms by the libav and ffmpeg projects. Windows Media Audio Lossless (WMA Lossless) is a lossless audio codec that competes with ATRAC Advanced Lossless, Dolby TrueHD, DTS HD Master Audio, Shorten, Monkey's Audio, FLAC, Apple Lossless, and WavPack (Since late 2011, the last three have the advantage of being open source software and available for nearly any operating system.) Designed for archival purposes, it compresses audio signals without loss of quality from the original using VBR. When decompressed, the audio signal is an exact replica of the original. The first version of the codec, WMA 9 Lossless, and its revisions support up to 96 kHz, 24 bit audio for up to 6 discrete channels (5.1 channel surround) with dynamic range compression control. The typical compression ratio for music varies between 1.7:1 and 3:1. Hardware support for the codec is available on the Cowon A3, Cowon S9, Bang & Olufsen Serenata Sony Walkman NWZ A and NWZ S series, Zune 4, 8, 80 30, Zune 120 (with firmware version 2.2 or later) and the Zune HD, Xbox 360, Similar to WMA Pro, the WMA Lossless decoder can perform downmixing when capable audio hardware is not present. As of 2012, the ffmpeg and libav projects have open source WMA Lossless decoders based on reverse engineering of the official decoder. Only 16 bit WMA files can be successfully decoded by ffmpeg as of June 20, 2012.

Windows Media Audio Voice (WMA Voice) is a lossy audio codec that competes with Speex (used in Microsoft's own Xbox Live online service), ACELP, and other codecs. Designed for low bandwidth, voice playback applications, it employs low pass and high pass filtering of sound outside the human speech frequency range to achieve higher compression efficiency than WMA. It can automatically detect sections of an audio track containing both voice and music and use the standard WMA compression algorithm instead.

Microsoft's claims of WMA sound quality have frequently drawn complaints. "Some audiophiles challenge Microsoft's claims regarding WMA's quality", according to a published article from EDN. At the early stages of WMA's development, a representative from RealNetworks claimed that WMA was a "clear and futile effort by Microsoft to catch up with RealAudio 8". Microsoft has sometimes claimed that the sound quality of WMA at 64 kbit/s equals or exceeds that of MP3 at 128 kbit/s (both WMA and MP3 are considered near transparent at 192 kbit/s by most listeners). In a 1999 study funded by Microsoft, National Software Testing Laboratories (NSTL) found that listeners preferred WMA at 64 kbit/s to MP3 at 128 kbit/s (as encoded by MusicMatch Jukebox). Both MP3 and WMA encoders have undergone active development and improvement for many years, so their relative quality may change over time.

Apart from Windows Media Player, most of the WMA compression formats can be played using ALLPlayer, VLC media player, Media Player Classic, MPlayer, RealPlayer, Winamp, Zune Software (with certain limitations—DSP plugin support and DirectSound output is disabled using the default WMA plugin), and many other software media players. The Microsoft Zune media management software supports most WMA codecs, but uses a variation of Windows Media DRM which is used by PlaysForSure. The FFmpeg project has reverse engineered and re implemented the WMA codecs to allow their use on POSIX compliant operating systems such as Linux. The Rockbox project further extended this codec to be suitable for embedded cores, allowing playback on portable MP3 players and cell phones running open source software. RealNetworks has announced plans to support playback of DRM free WMA files in RealPlayer for Linux. On the Macintosh platform, Microsoft released a PowerPC version of Windows Media Player for Mac OS X in 2003, but further development of the software has ceased. Microsoft currently endorses the third party Flip4Mac WMA, a QuickTime component that allows Macintosh users to play WMA files in any player that uses the QuickTime framework. Flip4Mac, however, does not currently support the Windows Media Audio Voice codec. The core Android platform does not itself support WMA, but there is third party WMA software for Android devices. WMA format can be played on almost all Windows Mobile and later Windows Phone devices.

There are many proprietary and open source software packages that can export audio in WMA format, including amongst many others Windows Media Player, Windows Movie Maker, Microsoft Expression Encoder, Sony Sound Forge, GOM Player, RealPlayer, Adobe Premiere Pro, Adobe Audition, Adobe Soundbooth, and VLC media player. Microsoft Office OneNote supports encoding in all WMA codecs, and Windows Media Encoder supports all available bit rate and resolution options as well.

The WMA codecs are most often used with the ASF container format, which has an optional DRM facility. Windows Media DRM, which can be used in conjunction with WMA, supports time limited music subscription services such as those offered by unlimited download services, including MTV's URGE, Napster, Rhapsody, Yahoo! Music Unlimited, and Virgin Digital. Windows Media DRM, a component of PlaysForSure and Windows Media Connect, is supported on many modern portable audio devices and streaming media clients such as Roku, SoundBridge, Xbox 360, and Wii. Players that support the WMA format but not Windows Media DRM cannot play DRM protected files.