it technically logs in and streams.. but thats it, complete rewrite almost everything broken

1 files changed, 0 insertions, 2375 deletions

diff --git a/modules/id3/changelog.txt b/modules/id3/changelog.txt
deleted file mode 100644
index 253f43b2..00000000
--- a/modules/id3/changelog.txt
+++ /dev/null
@@ -1,2375 +0,0 @@
-/////////////////////////////////////////////////////////////////
-/// getID3() by James Heinrich <info@getid3.org>               //
-//  available at http://getid3.sourceforge.net                 //
-//            or http://www.getid3.org                         //
-/////////////////////////////////////////////////////////////////
-//                                                             //
-// changelog.txt - part of getID3()                            //
-// See readme.txt for more details                             //
-//                                                            ///
-/////////////////////////////////////////////////////////////////
-
-        »  denotes a major feature addition/change
-        €  denotes a change in the returned structure
-* Bugfix:  denotes a fixed bug
-
-Version History
-===============
-
-1.7.2: [2004-10-18] Allan Hansen
-    * Bugfix: Large ID3v2 tags inside ASF not parsed properly under PHP5.
-    * Bugfix: Certain Wavpack3 files failed under PHP5 due to new 
-      undocumented tmpfile() limit (same problem as above).
-    * Bugfix: New iTunes crashes PHP - temp fix - no tags on those 
-      files.
-    € Trim/unset wavpack encoder_options to match 2.0.0b2 output.
-    » Added support for WavPack v4.0+ (thanks ahØartemis*dk)
-    » Removed code for parsing EXE files (thanks ahØartemis*dk)
-      Removed file: module.misc.exe.php
-    * Bugfix: ['nsv']['NSVs']['framerate_index'] might be wrong
-      (thanks ahØartemis*dk)
-    * Bugfix: transparent color was wrong from truecolor PNG
-      (thanks ahØartemis*dk)
-    * Bugfix: Changed MPC SV7 header size from 30 to 28, this will
-      change hash values for MPC files (thanks ahØartemis*dk)
-    * Bugfix: Changed MPC SV4-6 header size from 28 to 8, this will
-      change hash values for MPC files (thanks ahØartemis*dk)
-    € Commented-out unknown/unused values in NSV and ISO modules
-      (thanks ahØartemis*dk)
-    
-
-1.7.1b1: [July-26-2004] James Heinrich
-    » Added support for Apple Lossless Audio Codec
-    » Added support for RealAudio Lossless
-    » Added support for TTA v3
-    » Added support for TIFF
-      New file: /getid3/module.graphic.tiff.php
-    » Modified iconv_fallback to work with UTF-8, UTF-16, UTF-16LE,
-      UTF-16BE and ISO-8859-1 even if iconv() and/or XML support is
-      not available. This means that iconv() is no longer required
-      for most users of getID3()
-      (thanks Jeremia, khleeØbitpass*com)
-    » Added support for Monkey's Audio v3.98+ (thanks ahØartemis*dk)
-    » Included new demo showing most-basic getID3() usage
-      New file: /demos/demo.basic.php
-    * Bugfix: LAME3.94+ presets cached incorrectly if multiple files
-      are scanned in one batch and first file is LAME3.93 or earlier
-      (thanks enoyandØyahoo*com)
-    * Bugfix: Added warning if compressed ID3v2 frame decompression
-      fails. (thanks Mike Billings)
-    * Bugfix: Assorted small fixes to ensure compatability with PHP5
-    * Bugfix: ID3v1 genre "Blues" could not be written
-      (thanks Jeremia)
-    * Bugfix: ['bitrate_mode'] typo in module.audio-video.real.php
-      (thanks asukakenjiØusers*sourceforge*net)
-    * Bugfix: ['zip']['files'] is now populated with filenames even
-      if End Of Central Directory couldn't be parsed
-    * Bugfix: ['audio']['lossless'] was incorrect for FLAC
-      (thanks WaldoMonster)
-    * Bugfix: MD5 File was incorrect in directory browse mode for
-      /demo/getid3.browse.php
-    * Bugfix: PHP v5 compatability changes (float array keys, fread()
-      calls with zero data length)
-      (thanks getid3Øjsc*pp*ru)
-    * Bugfix: was dying if on compressed ID3v2 frames if
-      gzuncompress() function was unavailable
-    * Bugfix: ['vqf']['COMM'] was always empty
-    * Bugfix: MIDI playtime was missing for single-track MIDI files
-    * Bugfix: removed &#0; characters from ['comments_html']
-      (thanks p*quaedackersØplanet*nl)
-    * Bugfix: improved MIDI playtime accuracy
-      (thanks joelØoneporpoise*com)
-    * Bugfix: BMP subtypes 4 and 5 were not being identified
-    * Bugfix: frame_rate in AVI was incorrectly truncated to integer
-    * Bugfix: FLAC cuesheet track index was incorrect
-      (thanks tetsuo*yokozukaØoperamail*com)
-    € ['quicktime']['display_scale'] now contains the playback scale
-      multiplier for QuickTime movies - a movie set to playback at
-      double-size will have "2" here. Other values are "1" and "0.5"
-    € Added LAME preset guessing for --preset medium with v3.90.3
-      (thanks phwipØfish*co*uk)
-    € Added $encoding_id3v1 to allow for ID3v1 encodings other than
-      the standard ISO-8859-1
-    € Default AVI video bitrate_mode is now 'vbr'
-      (thanks eltoderØpisem*net)
-    Force getID3() to abort if Shorten files have ID3 or APE tags
-      (thanks ahØartemis*dk)
-    Editable textbox for parent directory in demo.browse.php
-      (thanks eltoderØpisem*net)
-
-
-1.7.0-hotfix [2004-03-17] Allan Hansen
-    (hotfix version released by Allan Hansen)
-    * Bugfix: PHP 4.1.x compatiblity - fgets($fp) => fgets($fp, 1024)
-    * Bugfix: Added default charset to TextEncodingNameLookup() in
-      module.tag.id3v2.php
-    Ø Removed option_no_iconv
-      iconv() support is only a requirement for WMA/WMW/ASF, and for
-      destination encodings other than ISO-8859-1 and UTF-8, iconv is
-      not needed otherwise. New 'iconv_req' in GetFileFormatArray()
-      only set for WMA/WMV/ASF. analyze() now refuses to analyse
-      WMA/ASF file if iconv is not present.
-    iconv_fallback() only dies on internal errors not missing iconv()
-
-
-1.7.0: [January-19-2004] James Heinrich
-    » Added support for RIFF/CDXA files (MPEG video in RIFF container
-      format (thanks chrisØdigitekdesign*com)
-    » Added support for TTA v2  (thanks ahØartemis*dk)
-    € ID3v2 unsynchronisation scheme disabled by default because most
-      tag-reading programs cannot read unsynchronised tags. Can be
-      overridden by setting id3v2_use_unsynchronisation to true.
-      (thanks mikeØdelusion*org)
-    € extention.*.php renamed to extension.*.php
-      (thanks tp62Øcornell*edu)
-    € /demo/demo.check.php renamed to /demo/demo.browse.php
-    € Added id3v2_paddedlength configuration parameter to WriteTags()
-      and renamed tag_language to id3v2_tag_language
-    € MPEG audio layers are now represented as 1, 2 or 3 instead of
-      'I', 'II', or 'III'
-    € Added [audio][wformattag] and [video][fourcc] for WAV and AVI
-    € Added [audio][streams] which contains one entry for each audio
-      stream present in the file (usually only one). The data is a
-      copy of what is usually found in [audio]. If there are multiple
-      audio streams then [audio] will contain a sum of the bitrates
-      of all audio streams, and the data format of the first stream
-      (if streams are of different data types)
-    € Added BruteForce mode to mp3 scanning. Disabled by default as
-      it is extremely slow and only files that are broken enough to
-      not really play will gain any benefit from this.
-    € Suppress '--resample xxxxx' appended to encoder options for mp3
-      with low-quality presets for default sampling frequencies
-    € Enhanced LAME preset guessing for pre-3.93 with a better lookup
-      table, --resample/--lowpass guessing (thanks phwipØfish*co*uk)
-    € RIFF files with non-MP3 contents no longer have
-      [audio][encoder_options] set
-    € Added [audio][encoder_options] to audio formats where possible
-      (including LiteWave, LPAC, OptimFROG, TTA)
-    € Moved [quantization] and [max_prediction_order] from
-      [lpac][flags] to just [lpac]
-    € WavPack flags are now parsed into [wavpack][flags]
-    * Bugfix: APEtags with ReplayGain information stored with comma-
-      seperated decimal values (ie "0,95" instead of "0.95") were
-      giving wrong peak and gain values
-    * Bugfix: Filesize > 2GB not always detected correctly
-    * Bugfix: Some ID3v2 frames had data key unset incorrectly
-      (thanks chrisØdigitekdesign*com)
-    * Bugfix: Warnings on empty-strings-only comments
-    * Bugfix: ID3v2 tag writing may have had incorrect padding length
-      if padded length less than current ID3v2 tag length and
-      merge_existing_data is false  (thanks mikeØdelusion*org)
-    * Bugfix: hash_data() for SHA1 was broken under Windows
-    * Bugfix: BigEndian2Float()/LittleEndian2Float() were broken
-    * Bugfix: LAME header calculated track peaks were incorrect for
-      LAME3.94a15 and earlier
-    * Bugfix: AVIs with VBR MP3 audio data reported incorrect bitrate
-      and bitrate_mode
-    * Bugfix: AVIs sometimes had incorrect or missing video and total
-      bitrates
-    * Bugifx: AVIs sometimes had incorrect ['avdataend'] and
-      therefore also incorrect data hashes (md5_data, sha1_data)
-    * Bugfix: ID3v1 genreid no longer returned for Unknown genre
-    * Bugfix: ID3v1 SCMPX genres were broken
-    Modified LAME header parsing to correctly process peak track
-      value for LAME3.94a16+ (thanks Gabriel)
-    md5_file() and sha1_file() now work under Windows in PHP < 4.2.0
-      and 4.3.0 respectively with helper apps
-    Default md5_data() tempfile location is now system temp directory
-      instead of same directory as file (thanks towbØtiscali*de)
-    Improved list of RIFF ['INFO'] comment key translations
-    More helpful error message when GETID3_HELPERAPPSDIR has spaces
-    /demo/demo.browse.php now autogets both MD5 and SHA1 hashes for
-      files < 50MB
-    Replaced PHP_OS comparisons with GETID3_OS_ISWINDOWS define
-      (thanks necroticØusers*sourceforge*net)
-
-
-1.7.0b5: [December-29-2003] James Heinrich
-    » Windows only: Various binary files are now required for some
-      file formats, especially for tag writing, as well as md5sum
-      (and other) calculations. These binaries are now stored in the
-      directory defined as GETID3_HELPERAPPSDIR in getid3.php
-      (default is /helperapps/ parallel to /getid3/).
-      Note: This directory must not have any spaces in the pathname.
-      All neccesary files are available as a seperate download.
-      See /helperapps/readme.txt for more information
-      New file: /helperapps/readme.txt
-    » Unified tag-writing interface for all tag formats
-      New file: /getid3/write.php
-                /getid3/write.apetag.php
-                /getid3/write.id3v1.php
-                /getid3/write.id3v2.php
-                /getid3/write.lyrics3.php
-                /getid3/write.metaflac.php
-                /getid3/write.vorbiscomment.php
-    » Added support for Shorten - requires shorten binary (head.exe
-      is also required under Windows).
-      New file: /getid3/module.audio.shorten.php
-    » Added support for RKAU
-      New file: /getid3/module.audio.rkau.php
-    » Added (minimal) support for SZIP
-      New file: /getid3/module.archive.szip.php
-    » Added MySQL caching extention  (thanks ahØartemis*dk)
-      New file: /getid3/extention.cache.mysql.php
-    » Added DBM caching extention  (thanks ahØartemis*dk)
-      New file: /getid3/extention.cache.dbm.php
-    » Added sha1_data hash option  (thanks ahØartemis*dk)
-    » Added option to allow getID3() to skip ID3v2 without parsing it
-      for faster scanning when ID3v2 data is not required. If you
-      want to enable this feature delete /getid3/module.tag.id3v2.php
-      (thanks ahØartemis*dk)
-    € 8-bit WAV data now calculates MD5 checksums as normal, not
-      converting to signed data as before, so stored md5_data_source
-      in FLAC files will no longer match md5_data for the equivalent
-      decoded 8-bit WAV. A warning will be generated for 8-bit FLAC
-      files
-    € Added option_no_iconv option to allow getID3() to work
-      partially without iconv() support enabled in PHP
-      (thanks ahØartemis*dk)
-    € All '*_ascii' keys removed for ASF/WMA/WMV files
-    € All 'ascii*' keys removed for ID3v2 tags
-    € Ogg filetypes now return MIME of "application/ogg" instead of
-      the previous "application/x-ogg"
-      (thanks blakewattersØusers*sourceforge*net)
-    € Force contents of ['id3v2']['comments'] to UTF-8 format from
-      whatever encoding each frame may have (text encoding can vary
-      from frame to frame in ID3v2)
-    € MP3Gain information from APE tags suppressed from ['tags'] and
-      parsed into ['replay_gain']
-    € ReplayGain information (all formats) changed from "Radio" and
-      "Audiophile" to "Track" and "Album" respectively
-    € ['volume'] and ['max_noclip_gain'] are now available in both
-      ['replay_gain']['track'] and ['replay_gain']['album'] for all
-      formats that calculate ReplayGain.
-    € ['video']['total_frames'] is available for AVIs
-    € All parsed ID3v2 frame data is now in ['id3v2'][XXXX][#]
-      (previously some frame types would have numeric array keys if
-      multiple instances of that frame type were allowed and other
-      frame types would not)
-    € ASF/WMA "WM/Picture" images are now parsed in the same manner
-      as ID3v2 with the image (ex JPEG) data returned in [data]
-      rather than [value]
-    * Bugfix: Optional tag processing options were being ignored (ie
-      ID3v1 still processed even if option_tag_id3v1 == false)
-      (thanks ahØartemis*dk)
-    * Bugfix: fixed MultiByteCharString2HTML() for UTF-8
-    * Bugfix: Quicktime files not always reporting video frame_rate
-    * Bugfix: False ID3v1 synch patterns in APE or Lyrics3 tags are
-      now detected and incorrect ID3v1 data not returned
-      (thanks sebastian_maresØusers*sourceforge*net for the idea)
-    * Bugfix: WMA9 Lossless now reported as lossless
-    * Bugfix: two typos in ID3v1 genre list
-    * Bugfix: MPEG-2/2.5 ABR/VBR MP3 files had doubled playtime
-    * Bugfix: MPEG-2/2.5 LayerII (ie MP2: 24/22.05/16kHz) files were
-      not detected due to incorrect frame length calculation
-    * Bugfix: MPEG LayerI files were not detected due to incorrect
-      frame length calculation (must be multiple of slot length)
-    Added alternative md5_data via system call - twice as fast. Needs
-      "getID3()-WindowsSupport" to work under Windows.
-      (thanks ahØartemis*dk)
-    ID3v2.4 compressed frames are now supported
-    php_uname() calls changed to use PHP_OS constant
-    Added SCMPX extensions to ID3v1 genres (0xF0-0xFE)
-    Obfuscated contributor email address in changelog and sourcecode
-    Added memory-saving EmbeddedLookup() function for lookup tables
-      in RIFF and ID3v2 modules (thanks ahØartemis*dk)
-    Major memory patches to many modules by using
-      $var = &$INFO_ARRAY_AT_SOME_INDEX
-      in place of large multi-dimensional array declarations.
-      Memory saved:  RIFF: ~200kB;  ID3v2: ~475kB;  ASF: ~50kB  etc.
-      (thanks ahØartemis*dk)
-
-
-1.7.0b4: [November-19-2003] James Heinrich
-   » Support added for MPC files with old SV4-SV6 structure
-   » RealVideo now properly supported with resolution, framerate, etc
-     (thanks jcsston)
-   » RealAudio files with old-style file format (v2-v4) are now
-     fully supported
-   » Support added for DolbyDigital WAV files (thanks ahØartemis*dk)
-   € ['RIFF'] is now ['riff'] to conform to make all root key names
-     lowercase
-   € ['OFR'] is now ['ofr'] to conform to make all root key names
-     lowercase
-   € ['tags_html'] is now available as a copy of ['tags'] but
-     with all text replaced with an HTML version of all characters
-     above chr(127), translated according to whatever the encoding
-     of the source tag is, in the HTML form &#1234;
-   € CopyTagsToComments() is now available in getid3_lib
-   € QuicktimeVR files now return a ['video']['dataformat'] of
-     'quicktimevr' instead of 'quicktime' (thanks gtsØtsu*biz)
-   € Quicktime video files with DivX, Xvid, 3ivx or MPEG4 video
-     streams now return those names as ['video']['dataformat']
-   € MPEG video files are now identified with ['video']['codec'] set
-     to either 'MPEG-1' or 'MPEG-2' (rather than just 'MPEG'). If you
-     see a file wrongly identified, please report it!
-     (thanks fccHandler)
-   € All bitrate values in ['mpeg']['audio'] is now reported in bps
-     rather than kbps (ie 128000 instead of 128) for consistancy
-   € AVIs with MP2 audio now report ['audio']['dataformat'] as 'mp2'
-     rather than 'wav'  (thanks metalbrainØnetian*com)
-   € Added ['md5_data_source'] for OptimFROG
-   € AC3 in RIFF-WAV now identified with ['audio']['dataformat']
-     returning 'ac3'
-   € WavPack ['extra_bc'] now returned as integer
-   € WavPack ['extras'] now returned as 3-element array of integers
-   € MP3 ['audio']['encoder options'] now returns 'VBR' or 'CBR' only
-     (no bitrate) if no LAME preset is used, or 'VBR q??' where ?? is
-     a number 0-100 for Fraunhofer-encoded VBR MP3s
-   * Bugfix: VBR MP3s could have incorrect bitrate reported
-   * Bugfix: Quicktime files with MP4 audio were not returning
-     ['video']['dataformat'] (thanks robØmassive-interactive*nl)
-   * Bugfix: strpad vs str_pad typo in module.riff.php
-     (thanks nicojunØusers*sourceforge*net)
-   * Bugfix: ReplayGain information was often wrong for MPC files
-   * Bugfix: MD5 and other post-TAIL chunks were not being processed
-     in module.audio.optimfrog.php
-   * Bugfix: Undefined variable in table_var_dump() in demo/check.php
-   * Bugfix: QuickTime files now only return information in [audio]
-     or [video] if those exist in the file
-   * Bugfix: WavPack no longer tries to read entire compressed data
-     chunk
-   * Bugfix: Properly handle VBR MP3s with "Info" (rather than
-     "Xing") header frame. foobar2000 adds this to MP3 files when
-     "Fix MP3 Header" function is used (thanks ahØartemis*dk)
-   * Bugfix: Fraunhofer VBRI headers for MP3s were assuming 2-byte
-     entries for TOC rather than using stride, and were ignoring the
-     scaling value. (thanks sebastianØmaresweb*net)
-   Several QuickTime atoms have been added to an exclusion list
-     because they have been observed, but I have no idea what they
-     are supposed to do so I can't add real support for them, but
-     they should not generate warnings (robØmassive-interactive*nl)
-   Old MPC encoder (before v1.06) was return as v0.00, now returned
-     as 'Buschmann v1.7.0-v1.7.9 or Klemm v0.90-v1.05'
-     (thanks ahØartemis*dk)
-   Added check for magic_quotes_runtime and code to disable it if
-     neccesary (thanks stefan*kischkelØt-online*de)
-   Added 3ivx fourCCs to module.audio-video.quicktime.php
-   MP3 and AC3 streams are now parsed when contained inside RIFF-WAV
-     or RIFF-AVI container formats
-   Better detection of named presets in LAME 3.93/3.94
-
-
-1.7.0b3: [October-17-2003] James Heinrich
-   » AC-3 (aka Dolby Digital) is now supported.
-     New file: /getid3/module.audio.ac3.php
-   * Bugfix: ID3v2-writing function Unsynchronise() was broken, which
-     made ID3v2 tag containing binary data (typically pictures) get
-     corrupted. (thanks t*coombesØbendigo*vic*gov*au,
-     i*kuehlbornØndh*net, mikeØdelusion*org, mikeØftl*com)
-   * Bugfix: Zip comments now returned as array instead of string,
-     as they're supposed to be.
-   * Bugfix: Quicktime/MP4 files may have reported extremely low
-     bitrates (thanks spunkØdasspunk*com)
-   Improved double-ID3v1 check to prevent false detection when string
-     "TAG" is present in APE or Lyrics3
-   Fixed /demo/simple.php
-   Fixed /demo/joinmp3.php
-   Fixed /demo/mimeonly.php
-   Fixed /demo/write.php
-
-
-1.7.0b2: [October-15-2003] James Heinrich
-   » TTA Lossless Audio Compressor format now supported.
-     (http://tta.iszf.irk.ru)
-     New file: /getid3/module.graphic.tta.php
-   » PhotoCD (PCD) format now supported. Image data for the three
-     lowest resolutions (192x128, 384x256, 768x512) can be optionally
-     extracted.
-     New file: /getid3/module.graphic.pcd.php
-   € RIFF-MP3 files now should return the same ['md5_data'] as the
-     identical MP3 file outside the RIFF container
-   € Name of LAME preset used (if available, needs LAME v3.90+)
-     returned in ['mpeg']['audio']['LAME']['preset_used'] and also as
-     part of ['audio']['encoder_options']
-   € VQF module now sets ['audio']['encoder_options'] to i.e. CBR96
-   € MP3 module now sets ['audio']['encoder_options'] on CBR files
-     and all LAME-encoded files
-   € MPC module now sets ['audio']['encoder_options']
-   € Monkey module now sets ['audio']['encoder_options']
-   € AAC module now sets ['audio']['encoder_options'] to profile name
-   € ASF module now sets ['audio']['encoder_options']
-   € Ogg module adds ['audio']['encoder_options'] -b 128 on
-     Ogg Vorbis 1.0+ ABR files
-   € Ogg module adds ['audio']['encoder_options'] -q N   on
-     Ogg Vorbis 1.0+ VBR files 44k/48k sample rate/stereo files only.
-   € Ogg module ['audio']['encoder_options'] "Nominal birate: 80k" to
-     other Ogg Vorbis files.
-   € ID3v2 track number now returned as string (with leading zeros,
-     if present in data) rather than integer (thanks Plamen)
-   € ASF module returns ['asf']['comments']['encoding_time_unix'] if
-     available (from WM/EncodingTime)
-   € Fixed /demo/mysql.php and added some new features:
-     - encoder options
-     - ID3v2 "Encoded By"
-     - non-empty comments
-     - total entries in database summary (totals & averages)
-     - database version update
-   * Bugfix: 'UNICODE' iconv() charset changed to 'UTF-16LE' or
-     'UTF-16BE' as appropriate
-   * Bugfix: iconv_fallback() function created in case iconv() fails
-   * Bugfix: fixed MD5 calls in demo/check.php
-   * Bugfix: reenabled detection of APE + Lyrics3 tags in same file
-   * Bugfix: ASF module now returns ID3v1 genre as string instead of
-     number - patch from Eugene Toder.
-   * Bugfix: ASF module now reads non-standard field names,
-     i.e. "date" as well as WM/Year - patch from Eugene Toder.
-   * Bugfix: ASF module now returns genre as-is if it is not a
-     standard ID3v1 genre (thanks wonderboy)
-   * Bugfix: Eliminated false-synch problem in MP3 module
-   * Bugfix: Fixed missing root ['bitrate'] for most formats
-   * Bugfix: ['audio']['compression_ration'] missing for MPC
-     (thanks WaldoMonster)
-   * Bugfix: NSV module died in 1.7.0b1
-   * Bugfix: ASF module died in 1.7.0b1 when WM/Picture preset
-   * Bugfix: ASF tracknumber incorrect when specified by WM/Track
-     rather than WM/TrackNumber (thanks jgriffiniiiØhotmail*com)
-   * Bugfix: MPEG audio+video playtime should now be pretty accurate
-     (ie within 0.1% variation at most)
-     (thanks mgrimmØhealthtvchannel*org)
-   * Bugfix: ID3v2 not being copied to ['tags'] in some cases
-   * Bugfix: LAME CBR files with Info tag were being incorrectly
-     flagged as VBR (thanks Jojo)
-   * Bugfix: LAME tag not being detected for LAME 3.90 (original)
-   Changed regex pattern match for MP3 to include 3rd byte for more
-     reliable/accurate pattern matching
-   Added duplicate-ID3v1 tag checking (two ID3v1 tags, one after the
-     other) that has been known to occur with iTunes
-     (thanks towbØtiscali*de)
-   Added instructions for enabling iconv() support under Windows
-   Removed some unneccesary debugging code
-   Suppressed duplicate PHP warnings for missing include files
-   Included some missing dependencies in various files
-   /demo/audioinfo.class.php now copies ['audio']['encoder_options']
-
-
-1.7.0b1: [2003-09-28] Allan Hansen 
-   This beta version was not made by James Heinrich. It was made by
-   Allan Hansen <ahØartemis*dk> - please send bug reports on this
-   beta directly to me.
-
-   James Heinrich will release 1.7.0 final, but it may take some time
-   to work out the bugs from the major rewrite.
-
-   This version could be called getID3lite. It makes a lot of checks
-   optional and makes it easy to remove support for undesired formats
-
-   It also is more library-like. Older versions of getID3() declared
-   an incredible amount of global scope functions and defined several
-   constants. 1.7.0beta1 still declares constants, but they are all
-   prepended by GETID3_. It declares no global scope functions - they
-   are all wrapped into classes.
-
-   » Made getID3() depend on iconv library: compile PHP --with-iconv
-   » Created new directory structure
-       Moved all demos to demos/
-       Moved all getID3() files to getid3/
-       Renamed most files to module.something
-       Changed header in all module.something to explain what they do
-       Simply remove all modules you don't need
-       Wrapped all modules into classes
-   * Bugfix: Implemented misc patches from Eugene Toder
-   * Bugfix: Implemented misc patches from "six"
-   € Added root key 'encoding'
-   € Added prefix GETID3_ to all defined constants.
-   € Wrapped getid3.php into getid3 class
-   € Wrapped getid3.functions.php into getid3_lib class
-       Removed unused functions
-       Moved several functions away from getid3.functions.php and
-         into the files where they are actually used.
-       Renamed getid3.functions.php to getid3.lib.php
-       Moved getid3.rgad.php functions into getid3_lib
-       Moved getid3.getimagesize.php funcitons ingo getid3_lib
-   € Moved getid3.ogginfo.php into ogg module
-   € Combined GetTagOnly() and GetAllFileInfo() in method analyze
-   € Removed redundant and unuseful root keys
-       'file_modified_time' == filemtime($filename)
-       'md5_file' == md5_file($filename)
-       'exist' == file_exists($filename)
-   € Changed root key ['tags'] from array of string to array of array
-     of comments.
-   Simplified code for detecting base path.
-   Removed ob_ from InitializeFilepointerArray(). That was really a
-     ugly HACK to get output from fopen. If user want the reason,
-     he should open the file himself!
-   Checking for APE tags before lyrics3 - makes Lyrics3 not depend
-     on APE tag. It seems to work on my test file.
-   Changed ['error'] and ['warning'] in multiple files to append to
-     array instead of appending to string. That simplified code in
-     getid3.php too.
-   Simplified clean-up procedure: simply remove all empty root keys
-   Setting tags in individual modules instead of main getid3.php
-   Made Bonk and ASF modules non-dependent on id3 modules - id3
-     optional.
-   Rewrote HandleAllTags() - simplified and convert comments to
-     desired encoding.
-   Replaced all calls to RoughTranslateUnicodeToASCII() in ASF module
-     with a TrimConvert() method. This uses iconv() for conversion.
-     It also converts from UNICODE instead of UTF-16BE, as the spec
-     says it should.
-   Replaced all calls to RoughTranslateUnicodeToASCII() in id3v2
-     module with iconv(). id3v2 module also reads
-     $ThisFileInfo['encoding'] and converts all comments to this
-     format. All other formats just add their comments in their
-     native charset, but every comment field in id3v2 can have a
-     different encoding, so this is needed.
-   Did same thing as above with ISO module. However - it does not
-     work. I need to find out how to specify big-endian unicode !=
-     UNICODING encoding name given to iconv().
-   Built-in assume mp3 format in getid3.php
-   Temporarily nuked root key ['comments'] and CopyCommentsToRoot()
-   Updated demo/audioinfo.class.php
-   Updated demo/check.php - some thing don't work!
-   Other demos are out of order!
-
-
-1.6.5: [October-06-2003] James Heinrich
-   » Added support for LiteWave (thanks supportØclearjump*com)
-   Ø Split out speedup info from ['OFR']['OFR']['compression'] into
-     ['OFR']['OFR']['speedup']
-   Ø If EXIF functions for JPEG not available, now warning not error
-   Ø ID3v2 track number now returned as string (with leading zeros,
-     if present in data) rather than integer (thanks Plamen)
-   * Bugfix: now correctly parses cbSize element of WAVEFORMATEX
-     structure (thanks supportØclearjump*com)
-   * Bugfix: ASF module now reads non-standard field names,
-     i.e. "date" as well as WM/Year - patch from Eugene Toder.
-   * Bugfix: ASF module now returns genre as-is if it is not a
-     standard ID3v1 genre (thanks wonderboy)
-   * Bugfix: ['audio']['compression_ration'] missing for MPC
-     (thanks WaldoMonster)
-   Prevent infinite loop in MP3 histogram if framelength == 0
-   Added wFormatTag values 0x00FF and 0x2001 - 0x2005
-     (thanks steveØheadbands*com)
-   Added "twos" and "sowt" FourCCs for Mac AIFC
-
-
-1.6.4: [June-30-2003] James Heinrich
-   » Added support for free-format MP3s
-     (thanks Sebastian Mares for the idea)
-   » Compressed (Flash 6+) SWF files are now handled properly
-     (thanks alan*cheungØalumni*ust*hk)
-   » Added DeleteLyrics3() to getid3.lyrics3.php
-   » Added FixID3v1Padding() to getid3.putid3.php
-   » Added new simple MP3-splicing sample file
-     (thanks tommybobØmailandnews*com for the idea)
-     New file: getid3.demo.joinmp3.php
-   » Moved all contents of getid3.putid3.php into either
-     getid3.id3v1.php or getid3.id3v2.php or getid3.functions.php as
-     appropriate
-     Removed file: getid3.putid3.php
-   € ['error'] and ['warning'] keys now return as arrays, not strings
-   € New root key for all files: ['file_modified_time'] (UNIX time)
-   € getid3.demo.scandir.php renamed to getid3.demo.mysql.php
-   € New demo file returns the MIME type only for a single file
-     (thanks adminØe-tones*co*uk for the idea)
-     New file: getid3.demo.mimeonly.php
-   € Added check for valid ID3v1 padding (strings should be padded
-     with null characters but some taggers incorrectly use spaces).
-     A warning will be generated if padding is invalid. New boolean
-     key ['id3v1']['padding_valid'] indicates padding validity.
-   € CleanUpGetAllMP3info() removes more useless root keys for
-     unknown-format files
-   € Extended LAME information in ['mpeg']['audio']['LAME'] is now
-     only returned for LAME v3.90+
-   € LAME-encoded MP3s now return
-     ['mpeg']['audio']['LAME']['long_version'] as well as
-     ['mpeg']['audio']['LAME']['short_version'] - these are identical
-     in LAME v3.90+ but older versions will report longer more
-     detailed version information if available
-   € New Lyrics3 values: ['lyrics3']['raw']['offset_start'] and
-     ['lyrics3']['raw']['offset_end']
-   € New optional parameter on getAPEtagFilepointer() to scan from a
-     defined offset rather than end-of-file to allow scanning of APE
-     tags before Lyrics3 tags
-   € ['tag_offset_start'] and ['tag_offset_end'] are now present in
-     ['ape'], ['lyrics3'], ['id3v1'] and ['id3v2']
-   € Numerous changes to the returned structure and content for La
-     files, including parsing the seektable (if applicable) and
-     parsing RIFF data occuring after the end of the compressed audio
-     data (notably RIFF comments)
-     (thanks mikeØbevin*de)
-   € getSWFHeaderFilepointer() now has optional 3rd parameter
-     $ReturnAllTagData (default == false) which if set to true will
-     return data on all tags in ['swf']['tags']
-   € ['swf']['bgcolor'] now returns the 6-character string
-     representing the background color in HTML hex color format
-     (thanks ubergeekØubergeek*tv)
-   € ['swf']['header']['frame_delay'] is no longer returned
-   € getQuicktimeHeaderFilepointer() now has two additional optional
-     parameters: $ReturnAtomData (default == true) and
-     $ParseAllPossibleAtoms (default == false). Setting
-     $ReturnAtomData to false will reduce the size of the returned
-     data array by unsetting ['quicktime']['moov'] before returning.
-     Leaving $ParseAllPossibleAtoms as false now suppresses parsing
-     of several atom types that contain very large tables of data
-     that are not typically useful. Atom type suppressed are:
-     stts, stss, stsc, stsz, and stco
-     (thanks ubergeekØubergeek*tv)
-   € ['fileformat'] no longer set to 'id3' if ID3v1 or ID3v2 tag
-     detected but no other data format recognized
-   * Bugfix: La files now return the correct values for
-     ['avdataoffset'] and ['avdataend'] and therefore the correct
-     values for ['md5_data'] - note that ['md5_data'] values will not
-     match values from previous versions of getID3() - the previous
-     versions were incorrect
-     (thanks mikeØbevin*de)
-   * Bugfix: A temporary file was being created in the web server's
-     root directory (not DocumentRoot) each time ['md5_data'] was
-     calculated, and not removed due to lack of permissions. Temp
-     file is now created (as it was supposed to be) in the directory
-     of the file being examined, or the system temp directory, and
-     properly removed when done.
-   * Bugfix: Several incorrect values were being returned inside
-     ['mpeg']['audio']['LAME'] (thanks bouvigneØmp3-tech*org)
-   * Bugfix: SWF frame rates values were usually incorrect.
-     (thanks alan.cheungØalumni*ust*hk and ubergeekØubergeek*tv)
-   * Bugfix: ID3v2.2 files always flagged 4 bytes of invalid padding
-     (thanks marcaØmac*com)
-   * Bugfix: Lyrics3 without ID3v1 was not working properly
-   * Bugfix: Lyrics3, APE & ID3v1 can all now exist in the same file.
-     A warning is issued if APE comes after Lyrics3 (because Lyrics3-
-     aware taggers probably are not APE-aware and therefore won't be
-     able to find the Lyrics3 tag)  (thanks mp3gainØhotmail*com)
-   * Bugfix: WriteAPEtag() now writes the APE tag before any Lyrics3
-     tags (if present) and removes any incorrect ones that are after
-     existing Lyrics3 tags  (thanks mp3gainØhotmail*com)
-   * Bugfix: RIFF-WAVE file with incorrect NumberOfSamples values in
-     the 'fact' chunk no longer cause incorrect playtime calculation
-     (thanks stprasadØindusnetworks*com)
-   * Bugfix: getid3.demo.simple.php had undefined variables if the
-     file needed to be deep-scanned with assumeFormat
-   * Bugfix: fixed previously-incorrect ['avdataend'] values for APE
-     and Lyrics3 tags in some cases, which in some cases means that
-     ['md5_data'] is different than previously (now correct)
-   Much-improved detection of AAC-ADTS, which also means MP3
-     format detection should now be nearly twice as fast
-   Truncated AVIs and WAVs are now reported
-   Number of new features and bugfixes in getid3.demo.mysql.php
-   Quicktime 'meta' atoms now parsed, so Quicktime MP4 files can now
-     return artist, title, album, etc  (thanks spunkØdasspunk*com)
-   Consolidated all comments processing functions (processing the
-     ['comments'] and ['tags'] keys) into HandleAllTags() which now
-     also checks to ensure that APE tags are really better than ID3v2
-     before using them in ['comments']
-   Known issue with Meracl ID3 Tag Writer v1.3.4 truncating last byte
-     of MP3 file when appending new ID3v1 tag now specifically noted
-     (rather than generic Probably Truncated File message)
-   getid3.demo.mysql.php now stores last-modified time for each file
-   getid3.demo.mysql.php is now case-sensitive for filenames
-   getid3.demo.mysql.php can generate M3U playlists of any of the
-     groups of files it can select (duplicate filenames, tag types,
-     etc.)
-   getid3.demo.mysql.php can now find mismatched tag contents and
-     filenames
-   getid3.demo.check.php now shows total number of errors & warnings
-   GetFileFormatArray() now matches actual patterns for MP3 files
-     based on the first two bytes of the file, rather than just the
-     first one
-   Simplified DeleteAPEtag() and made it work properly with Lyrics3
-
-
-1.6.3: [May-17-2003] James Heinrich
-   » Added support for Bonk  (thanks ahØartemis*dk)
-     New file: getid3.bonk.php
-   » Added support for AVR  (thanks ahØartemis*dk)
-     New file: getid3.avr.php
-   € Contents of getid3.id3.php moved to getid3.id3v1.php
-     Removed file: getid3.id3.php
-   € Contents of getid3.frames.php moved to getid3.id3v2.php
-     Removed file: getid3.frames.php
-   € Returned data structure documentation improved and updated and
-     now stored in getid3.structure.txt rather than getid3.readme.txt
-     New file: getid3.structure.txt
-   € Now including the GNU General Public License in the distribution
-     as getid3.license.txt
-     New file: getid3.license.txt
-   € Added new, optional, parameter to WriteAPEtag() (and also
-     GenerateAPEtag()) which must be set to TRUE if the values you
-     are passing are already UTF8-encoded, otherwise all data is
-     encoded to UTF8 by default. For all ASCII/ANSI data this value
-     should be left at the defaul value of FALSE.
-   € Added third, optional, parameter to getID3v2Filepointer() -
-     $StartingOffset (default == 0) which can parse an ID3v2 tag
-     in a file at a position other than the start-of-file.
-   € ['video']['pixel_aspect_ratio'] now returned when known
-   € AVI files with WMA audio now return ['audio']['dataformat']
-     of 'wma' rather than 'wav'
-   € ASF-WMA files now return the artist value from WM/AlbumArtist
-     in ['comments']['artist']  (thanks msibbaldØsaebauld*com)
-   € ASF-WMA files now return the 'author' value from
-     ['asf']['content_description'] in ['comments']['artist']
-     instead of ['comments']['author']
-   € ASF-WMA files now return the 'description' value from
-     ['asf']['content_description'] in ['comments']['comment']
-     instead of ['comments']['description']
-   * Bugfix: APE tag writing with multiple values for a tag (more
-     than one ARTIST for example) was not being correctly written
-     (thanks ahØartemis*dk)
-   * Bugfix: CreateDeepArray() was returning an empty-string key as
-     the top-level returned value - ['iso']['files'] now directly
-     contains the file listing without an empty array in between.
-   * Bugfix: ID3v2 genreid was not being returned in some cases.
-   * Bugfix: APEv1 tags would generate error messages
-   * Bugfix: APE tags would sometimes show phantom second entry for
-     each item (title, artist, etc) with no data
-   * Bugfix: APE tag writing was not UTF8-encoding the data -
-     non-ASCII characters (above chr(127)) were being incorrectly
-     stored  (thanks ahØartemis*dk)
-   * Bugfix: getid3.demo.scandir.php had undefined function error
-   * Bugfix: getid3.demo.scandir.php would not display list of files
-     with no tags
-   Added link to getid3.demo.check.php from list of specific-tags
-     files in getid3.demo.scandir.php
-
-
-1.6.2: [May-04-2003] James Heinrich
-   » New official mirror site for getID3() - http://www.getid3.org
-   » Added basic support for SWF (Flash)  (thanks n8n8Øyahoo*com)
-     New file: getid3.swf.php
-   » Added experimental support for parsing the audio portion of
-     MPEG-video files. I don't have any actual documentation for
-     this, so this part is experimental and not guaranteed accurate,
-     but it seems to be working OK as far as I have been able to test
-     it. Bug reports (or even better - documentation!) are welcome at
-     info@getid3.org
-   » Added new simple directory-scanning sample file
-     New file: getid3.demo.simple.php
-   » getid3.demo.write.php now writes APE tags as well.
-   € Renamed getid3.write.php to getid3.demo.write.php
-   € Renamed audioinfo.class.php to getid3.demo.audioinfo.class.php
-   € getid3.php now automatically includes the getid3.functions.php
-     function library file, no need to include it seperately.
-   € getLyrics3Filepointer() has been changed to be consistant with
-     all the other similar function structures - the parameters have
-     changed. The old function has been renamed to getLyrics3Data()
-   € Added DeleteAPEtag() function to getid3.ape.php
-   € HandleID3v1Tag() now only handles ID3v1. Lyrics3 processing is
-     now done by HandleLyrics3Tag()
-   € If BitrateHistogram is enabled in getOnlyMPEGaudioInfo() it now
-     also returns ['mpeg']['audio']['version_distribution'] showing
-     the number of frames of each MPEG version (1, 2 or 2.5) - all
-     frames *should* be of the same MPEG version
-   € getID3v1Filepointer() always returns TRUE now, even if it didn't
-     find a valid ID3v1 tag
-   € getOnlyMPEGaudioInfo() now looks for MPEG sync in the first 128k
-     bytes rather than the first 64k bytes
-   € Added dummy function GetAllMP3info() to generate warning not to
-     use that deprecated function.
-   € ['video']['codec'] is now 'MPEG' for all MPEG video files (this
-     will change to 'MPEG-1' or 'MPEG-2' as soon as I figure out how
-     to determine that)  (thanks jigalØspill*nl)
-   € ['mpeg']['audio']['LAME']['mp3_gain'] renamed to
-     ['mpeg']['audio']['LAME']['mp3_gain_db'] (gain in dB)
-   € Added ['mpeg']['audio']['LAME']['mp3_gain_factor'] (gain as a
-     multiplication factor)
-   € Added support for Preset and Surround Info bytes from LAME VBR
-     tag (http://gabriel.mp3-tech.org/mp3infotag.html)
-   * Bugfix: APE tag writing would put the string 'Array' for all
-     values rather than the actual data  (thanks ahØartemis*dk)
-   * Bugfix: Warning now generated for VBR MPEG-video files because
-     getID3() cannot determine average bitrate. If you know of
-     documentation that would tell me how to do this, please email
-     info@getid3.org
-   * Bugfix: Replay Gain values from Vorbis comments are now
-     returned in ['replay_gain'] (and not in ['comments'])
-     (thanks ahØartemis*dk)
-   * Bugfix: Replay Gain values from APE comments are now correctly
-     returned in ['replay_gain']  (thanks ahØartemis*dk)
-   * Bugfix: getid3.demo.check.php is now case-insensitive when
-     assuming a format for a corrupted file if standard detection
-     does not identify the file type.
-   * Bugfix: RIFF comments were overwriting/suppressing ID3 comments
-     for RIFF-MP3 files  (thanks wmØwofuer*com)
-   * Bugfix: RIFF-MP3 files with 'RMP3' chunks instead of 'WAVE' were
-     not being correctly identified.
-   * Bugfix: ID3v2 padding shorter than the length of an ID3v2 frame
-     header was not correctly detected
-   * Bugfix: getid3.demo.check.php now does in-depth scanning for MP2
-     and MP1 files the same as for MP3 files based on file extension
-     if a MPEG-audio structure isn't found immediately at the start
-     of the file
-   * Bugfix: removed condition where RIFF-WAV was being scanned for
-     MPEG-audio signature when it shouldn't be present (non-MP3 WAV)
-   * Bugfix: ASF files were not always showing correct audio datatype
-   * Bugfix: array_merge_clobber() and array_merge_noclobber() were
-     not being conditionally defined in getid3.functions.php
-     (thanks rich.martinØreden-anders*com)
-   * Bugfix: stream_numbers was not being correctly returned in
-     bitrate_mutual_exclusion_object chunks of ASF files
-   * Bugfix: Added support for 24kHz and 12kHz audio in ASF files
-   * Bugfix: Removed possible undefined offset error in MP3s where
-     cannot find synch before end of file
-   * Bugfix: Removed potential out-of-memory crash situation when
-     parsing Real files with chunks larger than the available memory
-     (thanks jigalØspill*nl)
-   * Bugfix: ID3v1 was incorrectly taking precedence over ID3v2 in
-     the ['comments'] array (thanks lionelflØwanadoo*fr)
-   * Bugfix: No longer calculates overall bitrate and playtime for
-     VBR MPEG video files based on the audio bitrate.
-   * Bugfix: AssumeFormat was not working properly
-   Added summary footer line to getid3.demo.check.php
-   Added '.mpeg' to the list of assume-format-from-filenames list in
-     getid3.demo.check.php
-   MPEG-video files now more reliably detected
-   A number of additional features have been added to
-     getid3.demo.scandir.php
-   Added many RIFF-AVI audio types and fourcc video types to the
-     lookup functions in getid3.riff.php
-   Now identifes files with Lyrics3 v1 tags that are of incorrect
-     length (v1 Lyrics3 is supposed to be 5100 bytes long, but
-     [unknown program] writes variable-length tags (which is illegal
-     for Lyrics3 v1)). getID3() now correctly parses these tags and
-     issues a warning.
-   Split GetFileFormat() to GetFileFormat() and GetFileFormatArray()
-   HTML colors in getid3.demo.check.php are now defined as constant
-     variables at the top of the file (if you want to change them)
-   Added support for OptimFROG v4.50x (non-alpha) (new header fields)
-     (thanks floringhidoØyahoo*com)
-   Added support for Lossless Audio v0.4 (thanks mikeØbevin*de)
-
-
-1.6.1: [March-03-2003] James Heinrich
-   » Added support for writing APE v2.
-     WriteAPEtag() in getid3.ape.php
-     NOTE: APE v1 writing support will *not* be added to future
-     versions of getID3()
-     (thanks ahØartemis*dk and adamØphysco*com for the idea)
-   » Added support for AIFF (Audio Interchange File Format) including
-     AIFF, AIFC and 8SVX  (thanks ahØartemis*dk for the idea)
-     Removed file: getid3.aiff.php
-   » Added support for OptimFROG (v4.50a and v4.2x)
-     (thanks ahØartemis*dk for the idea)
-     New file: getid3.optimfrog.php
-   » Added support for WavPack  (thanks ahØartemis*dk for the idea)
-   » Added support for LPAC  (thanks ahØartemis*dk for the idea)
-   » Added support for NeXT/Sun .au format
-     New file: getid3.au.php
-   » Added support for Creative SoundBlaster VOC format
-     New file: getid3.voc.php
-   » Added support for the BWF (Broadcast Wave File) RIFF chunks
-     "bext" and "MEXT"  (thanks Ryan and njhØsurgeradio*co*uk)
-   » Added support for the CART (Broadcast Wave File) RIFF chunks
-     (thanks Ryan)
-   » Added getid3.demo.scandir.php - a sample recursive scanning demo
-     that scans every file in a given directory, and all sub-
-     directories, and stores the resulting data in MySQL database,
-     and then displays a list of duplicate files based on md5_data
-   € ['md5_data_source'] now contains the MD5 value for the original
-     uncompressed data for formats that store that information
-     (currently only FLAC v0.5+). ['md5_data'] (if chosen to be
-     calculated) will contain the calculated MD5 value for the
-     compressed file. To check if 2 files are identical in every way,
-     including all comments: compare ['md5_file']. To check if two
-     files were compressed from the same source file: compare
-     ['md5_data_source']. To check if the compressed audio/video data
-     of two files is identical, even if comments or even the
-     container file format is different (MP3 in RIFF container,
-     FLAC in Ogg container, etc): compare ['md5_data'].
-   € ['md5_data'] for 8-bit WAV files is now calculated based on a
-     converted version of the data from unsigned to signed (MSB
-     inverted) to match the MD5 value calculated by FLAC
-   € New optional parameter added to GetAllFileInfo() -
-     $MD5dataIfMD5SourceKnown (default: false). If false the md5_data
-     value will NOT be calculated for files (such as FLAC) that have
-     ['md5_data_source'] set, even if $MD5data == true.
-     (thanks ahØartemis*dk)
-   € getid3.check.php renamed to getid3.demo.check.php
-   € Added GetTagOnly() function to getid3.php - similar to
-     GetAllFileInfo() except only takes a filename as a parameter and
-     only returns ID3v2, APE, Lyrics3 and ID3v1 tag information - no
-     attempt is made to parse the data contents of the file at all.
-     (thanks Phil for the idea)
-   € Added ['audio']['lossless'] and ['video']['lossless'] for all
-     formats (when known). Both are boolean values - true means the
-     data is lossless-compressed, false means the data is lossy-
-     compressed.
-   € Added ['audio']['compression_ratio'] and/or
-     ['video']['compression_ratio'] for all formats. Returns a number
-     (usually) less than 1, where 1 represents no compression and 0.5
-     represents a compressed file half the size of the original file
-   € Added ['video']['bits_per_sample'] to all video formats (when
-     known)
-   € Added ['video']['frame_rate'] to all video formats (when known)
-   € ['fileformat'] set to 'mp1' or 'mp2' instead of 'mp3' when
-     ['audio']['dataformat'] is one of those  (thanks ahØartemis*dk)
-   € Added 4th parameter to md5_data(), $invertsign, which will invert
-     the MSB of each byte before MD5'ing. This is needed for 8-bit
-     WAV files because FLAC calculates the stored MD5 value on
-     signed data rather than the original byte values. ['md5_data']
-     of an 8-bit WAV will now match the ['md5_data_source'] value
-     (thanks lichvarmØphoenix*inf*upol*cz)
-   € ['ape']['items']['data'] and ['ape']['items']['data_ascii'] now
-     contains an array of values, if the tag contains UTF-8 text (as
-     opposed to binary data)
-   € ['mpeg']['audio']['bitratemode'] renamed to
-     ['mpeg']['audio']['bitrate_mode']
-   * Bugfix: Removed potential bug that could replace all MP3 file
-     contents with only the new ID3v2 tag in getid3.putid3.php
-   * Bugfix: md5_data values calculated for RIFF (WAV, AVI) files
-     were incorrect  (thanks ahØartemis*dk)
-   * Bugfix: MP3 data in an MP4 wrapper fileformat could not identify
-     bitrate  (thanks ahØartemis*dk)
-   * Bugfix: ['audio'] and/or ['video'] keys would sometimes get
-     removed even if not empty
-   * Bugfix: Prevented creation of null entries in
-     ['RIFF']['WAVE']['INFO'] if a comment entry was not present
-   * Bugfix: Potential infinite-loop condition in getid3.ogg.php
-     (thanks afshin.behniaØsbcglobal*net)
-   * Bugfix: Ogg files with illegal ID3v1 (and/or APE or Lyrics3)
-     tags were not finding the last Ogg page
-     (thanks afshin.behniaØsbcglobal*net)
-   * Bugfix: replay-gain values not properly set from LAME tag
-   * Bugfix: RIFF-MP3 had incorrect md5_data
-   * Bugfix: the LAME DLL CBR problem of not re-writing the LAME
-     frame at the beginning of the data is now detected for MP3s
-     with ID3v2 tags as well
-   * Bugfix: APE tags with multiple values (ie multiple entries in
-     the "artist" tag) are now shown properly in ['ape']['items']
-   * Bugfix: fixed condition where APE tag with no ID3v1 tag could be
-     mistaken for APE tag with ID3v1 (and incorrectly parsed)
-   * Bugfix: added warning if ID3v2 frame has zero-length data
-     (thanks cmassetØclubinternet*fr)
-   * Bugfix: getid3.frames.php looking for non-existant key in USER
-     frames
-   Improved detection of RIFF-MP3 data. [unknown program] encodes
-     RIFF-WAV data with a chunk name of 'RMP3' instead of the
-     standard 'RIFF'
-   Encoder now returned in both ['comments'] and ['audio']['encoder']
-     for RIFF-WAV files with an INFO.ISFT chunk
-   Generate a warning for FLAC files encoded with v0.3 or v0.4
-     because audio_signature is not calculated during encoding
-     (thanks ahØartemis*dk)
-   Modified getid3.check.php to display md5_data_source as well as
-     md5_file and md5_data if display-MD5 mode is selected
-   Modified getid3.check.php to assume-format based on file extension
-     in browse mode if fileformat is found to be 'id3' (formerly only
-     if the fileformat was null)
-   Changed scaling of BitrateColor() from representing 1-256kbps to
-     representing 1-768kbps for better display of high-bitrate files,
-     specifically lossless-compressed CD-audio (FLAC, LA, etc)
-
-
-1.6.0: [January-30-2003] James Heinrich
-   » Added support for OggFLAC (FLAC data stored in an Ogg container)
-     (thanks ahØartemis*dk for the idea)
-   » Added support for Speex (the data stored in an Ogg container)
-   » Comments are now available in the root 2-dimensional array
-     ['comments'] - each entry in this array will contain one or more
-     strings. For example, if there are two artists then
-     ['comments']['artist'][0] will contain the first one and
-     ['comments']['artist'][1] the other. All keys are forced
-     lowercase. Comments will be stored in the ['comments'] array in
-     this order of precedence:
-     1) Native format tags (ASF, VQF, NSV, RIFF, Quicktime, Vorbis)
-     2) APE tags
-     3) ID3v2
-     4) Lyrics3
-     5) ID3v1
-     Lower-priority tags will not overwrite or append existing values
-     of higher-priority tags (for example, 'artist' in ID3v1 will be
-     ignored if already specified in APE), but missing values will be
-     filled in (for example, if 'album' is specified in ID3v2 but not
-     in APE, it will be included in the ['comments'] array).
-     Note: Root keys (['title'], ['artist'], etc) are NOT available
-     in this or future versions of getID3().
-     (thanks ahØartemis*dk)
-   » MD5 hashes are now available for all formats for both the entire
-     file (['md5_file']) and the portion of the file containing only
-     the audio/video data, stripped of all prepended/appended tags
-     like ID3v2, ID3v1, APE, etc - ['md5_data']
-     (thanks ahØartemis*dk for alternate md5_file() function that
-     runs on UNIX system running PHP < 4.2.0)
-     NOTE: Ogg files require the use of vorbiscomment to obtain the
-     md5_data value. vorbiscomment must be downloaded from
-     http://www.vorbis.com/download.psp and placed in the getID3()
-     directory. All Ogg formats (Vorbis, OggFLAC, Speex) are affected
-     by this problem, but only OggVorbis files can be processed with
-     vorbiscomment. OggFLAC and Speex files will be processed by
-     getID3(), but this may result in an incorrect value for md5_data
-     in the event that VorbisComments are larger than 1 page (4-8kB).
-     NOTE: md5_data for Ogg will not work if PHP is running in Safe
-     Mode
-   » There is now a wrapper class available, written by Allan Hansen,
-     which should simplify extracting most common basic information
-     (such as format, bitrate, comments).
-     New file: audioinfo.class.php
-   » OggWrite() in getid3.ogginfo.php has been replaced with a new
-     version that uses vorbiscomment to write the comments, because
-     of a reported bug that can corrupt OggVorbis files such they
-     cannot be played.
-     NOTE: Ogg comment writing now requires the use of vorbiscomment
-     which must be downloaded from http://www.vorbis.com/download.psp
-     and placed in the getID3() directory.
-     NOTE: Ogg comment writing will not work if PHP is running in
-     Safe Mode
-   € New root key ['tags'] is now always returned for all formats.
-     It is an array that may contain any of:
-     * Native format tags: 'vqf', 'riff', 'vorbiscomment', 'asf',
-       'nsv', 'real', 'midi', 'zip', 'quicktime'
-     * Appended data tags:  'ape', 'lyrics3', 'id3v2', 'id3v1'
-   € New root key ['audio'] is an array containing any or all of:
-       codec, channels, channelmode, bitrate, bits_per_sample,
-       dataformat, bitrate_mode, sample_rate, encoder
-       Note: This replaces several root keys, including:
-         bitrate_audio, bits_per_sample, frequency, channels
-   € New root key ['video'] is an array containing any or all of:
-       bitrate_mode, bitrate, codec, resolution_x,  resolution_y,
-       resolution_y, frame_rate, encoder
-       Note: This replaces several root keys, including:
-         bitrate_video, resolution_x, resolution_y, frame_rate
-   € ['id3']['id3v1'] has moved to ['id3v1']
-   € ['id3']['id3v2'] has moved to ['id3v2']
-   € ['audiodataoffset'] and ['audiodataend'] have been renamed to
-     ['avdataoffset'] and ['avdataend'] respectively
-   € GetAllMP3info() has been changed to GetAllFileInfo() with a
-     different parameter list ($allowedFormats is no longer a
-     parameter).  Check your code where you're calling
-     GetAllMP3Info() - you will need to change both the function
-     name and the parameter list if you pass more than 2 parameters
-   € All formats now return ['audio']['dataformat'] and/or
-     ['video']['dataformat'] where appropriate - this goes along with
-     ['fileformat'] - ['fileformat'] will return the actual structure
-     of the file, whereas ['dataformat'] will return the format of
-     the data inside that structure. For example, an Ogg file can
-     contain Vobis data (normal), or it can contain FLAC data in the
-     Ogg container format. In that case, ['fileformat'] would be
-     'ogg', but ['dataformat'] would be 'flac'.
-     Note: this means that WAV and AVI files now return a
-     ['fileformat'] of 'riff' rather than 'wav' or 'avi'.
-   € ['filesize'] is no longer returned for files larger than 2GB
-     because PHP does not support large file access. Attempting to
-     parse a file larger than 2GB will result in a message stored in
-     ['error'] and ['filesize'] not set.
-   € APEtag, ID3v1, and ID3v2 are now supported on ALL multimedia
-     files - even if illegal by format. Ogg will return warning if
-     ID3/APE tags are present.  (thanks ahØartemis*dk)
-   € All files: non-critical errors are now returned in the root key
-     ['warning'] rather than ['error'] (only critical errors that
-     prevent getID3() from correctly parsing the file are returned in
-     ['error']  (thanks ahØartemis*dk)
-   € Renamed all references to $MP3fileInfo to $ThisFileInfo
-   € Joliet now supported for ISO-9660.
-     ['iso']['supplementary_volume_descriptor'] is now returned, if
-     available, and ['iso']['files'] will contain ASCII equivalents
-     of the Unicode directory structure & filenames stored.
-   € Moved Monkey's Audio code from getid3.ape.php to seperate file.
-     New file: getid3.monkey.php
-   € Added new keys for ISO-9660: ['name_ascii'] for directories,
-     ['file_identifier_ascii'] for files
-   € Added root key ['track'] for CD-audio files
-   € Ogg/Vorbis-comment files now have comments returned inside
-     ['ogg']['comments_common'] as an array of strings, rather than
-     simple strings in ['ogg']
-   € Quicktime files now have comments returned inside
-     ['quicktime']['comments'] as an array of strings, rather than
-     simple strings in ['quicktime']
-   € ['mime_type'] is a new root key returned for all supported
-     formats (thanks ahØartemis*dk)
-   € ['fileformat'] now returns 'mp1' instead of 'mp3' for MPEG-1
-     layer-I audio files (thanks ahØartemis*dk)
-   € ['mpeg']['audio']['bitratemode'] now returns lowercase
-   € MPEG-4 audio files which consist of MP3 data wrapped in a
-     Quicktime fileformat will now return the usual data in
-     ['mpeg']['audio']
-   € Type-1 DV AVIs are now supported
-   € DV AVIs will return 1 or 2 in ['RIFF']['video'][x]['dv_type']
-   € Changed ['fileformat'] from 'mpg' to 'mpeg' for MPEG video files
-   € ASF comments are now stored in ['asf']['comments'] instead of
-     ['asf']
-   € RealMedia chunk data is now returned inside ['real']['chunks']
-     instead of ['real']
-   € ['replay_gain'] now properly populated from APE tags
-   € Added support for ASF_Old_ASF_Index_Object in ASF files
-     (thanks ahØartemis*dk)
-   € AAC-ADTS files now return ['aac']['bitrate_distribution']
-   € ParseVorbisComments() has been replaced with
-     ParseVorbisCommentsFilepointer() (with different parameters)
-   € All references to any key ['frequency'] are now ['sample_rate']
-   € Moved ID3v2 comments from ['id3v2'] into common root
-     ['comments'] structure, and now returns more values than before
-   * Bugfix: ['iso']['files'] and ['zip']['files'] could potentially
-     contain duplicate entries (in a numeric-indexed array) for files
-     if the directory structure specifies files multiple times.
-     Entries are now guaranteed unique, with the last entry for the
-     file overwriting any former ones.
-   * Bugfix: RIFF parsing had numerous issues, including:
-     - large AVIs would take a very very long time to parse
-     - chunks with odd (not even) sizes would cause the parser fail
-     - video and/or audio codecs not always identified
-     The ParseRIFF() function has been completely rewritten and fixes
-     all known issues with RIFF parsing. Users are, however,
-     encouraged to double-check output of any parsed (AVI/WAV/CDDA)
-     files.
-   * Bugfix: Modified getid3.riff.php to return correct total
-     bitrates for AVIs with multiple audio streams
-   * Bugfix: GetFileFormat() was not creating array structure
-     correctly  (thanks ahØartemis*dk)
-   * Bugfix: LAME tag for MP3s can only specify up to 255kbps, so any
-     files with actual CBR bitrate of >=256 were reported incorrectly
-   * Bugfix: Lyrics3 synched lyrics were not being correctly returned
-   * Bugfix: CreateDeepArray() was broken for non-nested cases, which
-     meant ZIP and ISO ['files'] structures were broken
-   * Bugfix: Incorrect pattern matching for ZIP files meant no zip
-     files were being detected as such
-   * Bugfix: AAC-ADIF was returning an incorrect number of channels
-     (too few) in some cases  (thanks ahØartemis*dk)
-   * Bugfix: Vorbis comments were returning an incorrect value for
-     ['dataoffset'] in some cases
-   * Bugfix: MPEG video ['marker_bit'] and ['vbv_buffer_size'] were
-     incorrect
-   * Bugfix: ['playtime_string'] could potentially have a value of
-     x minutes and 60 seconds (ie 3:60 instead of 4:00)
-   Added support for FLAC cuesheets (FLAC 1.1.0+)
-     (thanks ahØartemis*dk)
-   Improved parsing speed in MP3, MP2 and AAC  (thanks ahØartemis*dk)
-   Extra error-checking added to try and identify corrupt files for
-     most audio formats  (thanks ahØartemis*dk)
-   More accurate playtime calculation for RealMedia
-     (thanks ahØartemis*dk)
-   Changed all relevant files to use ['audiodataoffset'] and
-     ['audiodataend'] rather than ['filesize'] where appropriate
-     (thanks ahØartemis*dk)
-   Added text encoding type 255 as a duplicate of UTF-16BE but with
-     Big-Endian rather than Little-Endian byte order
-   Added many RIFF-AVI audio types and fourcc video types to the
-     lookup functions in getid3.riff.php
-   Added numerous new known GUIDs to getid3.asf.php
-   Added PoweredBygetID3() function to easily get a "powered by"
-     string with the current getID3() version.
-   Added "Morgan Multimedia Motion JPEG2000" (MJ2C), "DivX v5" (DX50)
-     and "XviD" (XVID) codecs to list of known codecs in
-     getid3.riff.php
-   Changed GETID3_INCLUDEPATH path seperators to forced /
-     (from \ for Windows)
-   Modified getid3.check.php to only change \ directory seperators to
-     / on Windows operating systems
-   Modified getid3.check.php to handle larger-than-2GB files (which
-     now do not return a filesize)
-   Modified getid3.check.php to handle ['dataformat_audio'] and
-     ['dataformat_video']
-   Modified getid3.check.php to show a list of present tags in one
-     column rather than one column for each of ID3v1, ID3v2, etc
-   Modified getid3.check.php to show MD5 values. Initially disabled
-     but can be enabled for a directory with a click. md5_file is
-     always calculated when displaying detailed info about a single
-     file; md5_data is calculated if the file is < 50MB
-   Modified getid3.check.php to show errors and warnings. Details are
-     visible with a mouseover or a click.
-   Changed getid3.check.php to use SafeStripSlashes instead of a
-     manual conditional directory name replacement for special
-     characters
-   Added sample recursive scanning sample code to getid3.readme.txt
-     (thanks lipisinØmail*ru for the idea)
-
-
-1.5.7: [January-10-2003] James Heinrich
-   » Added support for ISO 9660 (CD-ROM image) format. Most-useful
-     data is directory structure returned under ['iso']['files']
-     Note: Only ISO-9660 supported, not (yet) Joliet extension
-     (thanks nebula_djØsofthome*net for the idea)
-     New file: getid3.iso.php
-   € ZIP files are now parsed by getID3() itself without relying on
-     built-in PHP functions and/or ZZipLib support.
-     (thanks Vince for the idea)
-   € ZIP files now return a simple directory listing with filename
-     and filesize info only under ['zip']['files'].
-     Note: empty subdirectories will note appear in here, only files
-     and non-empty subdirectories. Information for all entries,
-     including empty subdirectories, is available under
-     ['zip']['central_directory'] (or under ['zip']['entries'] if the
-     Central Directory cannot be located (usually due to a trucated
-     file).
-   € RIFF-WAV files with MP3 data (or MP3s with RIFF headers, if you
-     want to think of it that way) now have the MPEG audio portion
-     scanned and the usual data returned in ['mpeg']['audio'] if the
-     RIFF audio codec has wFormatTag of "85" (identified by getID3()
-     as "MPEG Layer 3")
-     (thanks ahØartemis*dk for the idea)
-   € EXIF data (if present) is returned for JPEG files under
-     ['jpg']['exif']  (thanks nebula_djØsofthome*net)
-   € ['filepath'] now returned for all files with the directory part
-     of the full filename.
-   € ['filenamepath'] is now returned for all files (equivalent to
-     ['filepath'].'/'.['filename'])
-   * Bugfix: ['id3']['id3v2'][<framename>]['dataoffset'] was wrong
-   * Bugfix: MP3s tagged with iTunes have an invalid comment field
-     frame name ('COM ' - should be 'COMM') but the data is valid
-     otherwise; the frame is now renamed to 'COMM' and parsed
-     normally (with the error noted in ['error'])
-     (thanks kheller2Ømac*com for the sample file)
-   * Bugfix: Some ASF/WMA audio files were not being identified as
-     any format  (thanks ahØartemis*dk)
-   * Bugfix: Warning now generated and ASCII format assumed for
-     invalid text encoding values in ID3v2
-   * Bugfix: Changed ZIP detection pattern from 'PK' to 'PK\x04\x03'
-   * Bugfix: Ogg/FLAC files with large Vorbis comments were dying in
-     an infinite loop with lots of error messages due to missing $fd
-     parameter on ParseVorbisComments()  (thanks ahØartemis*dk)
-   * Bugfix: ['data'] and ['image_mime'] were being returned for all
-     Ogg comments even if they were not images for versions of PHP
-     that have image_type_to_mime_type() built in (ie PHP 4.3.0+)
-
-
-1.5.6: [December-31-2002] James Heinrich
-   » Added support for NSV (Nullsoft Streaming Video)
-     (www.nullsoft.com/nsv/)
-     (thanks demonØsoundplanet*com for the idea)
-     New file: getid3.nsv.php
-   » Added support for CD-audio track files (track01.cda etc)
-   € Added standard ['frame_rate'] root value when known (AVI, NSV,
-     MPEG-video)
-   € ASF files now report ['fileformat'] of:
-     'wmv' when Windows Media Video codec v7/v8/v9 is used;
-     'wma' when any 'Windows Media Audio' named audio codec is used
-           and no video stream is present;
-     'asf' in all other cases (audio-only, video-only, or both)
-   € Removed support for ZIP functions (will be rewritten to not
-     require ZZIPlib support in future versions)
-   € Added function SafeStripSlashes() as a drop-in replacement for
-     stripslashes(), but that only strips slashes if magic_quotes_gpc
-     is set
-   € Removed support for remote file scanning (HTTP / FTP)
-   € Added ['aac']['frames'] (number of AAC frames in file)
-   € Added ['mpeg']['audio']['frame_count'] when a bitrate histogram
-     is created
-   € Average bitrate for VBR MP3/MP2 is calculated from actual counts
-     of frames of various bitrates (rather than relying on the header
-     values or filesize) when a bitrate histogram is created
-   € RecursiveFrameScanning() split out into seperate function
-     (getid3.mp3.php)
-   € Removed old function getMP3header() from getid3.mp3.php
-   € Changed default MPEG_VALID_CHECK_FRAMES (number of mp3 frames
-     scanned to ensure a valid audio sequence has been located) from
-     10 to 25. This means scanning will be slightly slower, but more
-     reliable/accurate
-   * Bugfix: ID3v2.2 - valid frame names not correctly detected
-     (thanks maeckerØweb*de for the sample file)
-   * Bugfix: ID3v2.2 - valid padding not correctly detected
-     (thanks maeckerØweb*de for the sample file)
-   * Bugfix: MIDI files with flat key signatures were not being
-     correctly reported (thanks alexleeisØshaw*ca for sample file)
-   * Bugfix: now returns message in ['error'] if file does not exist
-   * Bugfix: ['RIFF']['video'][x]['codec'] wasn't always being
-     correctly populated
-   * Bugfix: ['bitrate'] was incorrect for multi-stream RealMedia
-   * Bugfix: ['playtime_seconds'] was sometimes null or incorrect
-     for multi-stream RealMedia
-   * Bugfix: ChannelTypeID was incorrect in RVA2 ID3v2.4 frames
-   * Bugfix: Fixed potential divide-by-zero error for corrupt FLAC
-     files  (thanks ahØartemis*dk)
-   * Bugfix: AAC-ADTS was not returning ['bitrate_mode'] unless
-     $ReturnExtendedInfo was TRUE  (thanks ahØartemis*dk)
-   * Bugfix: LAME-encoded CBR MP3s now properly identified as CBR
-     with correct bitrate  (thanks ahØartemis*dk)
-   * Bugfix: VBR MP2 (or headerless MP3) is now identified as VBR
-     rather than CBR. Note: to obtain VBR bitrate for headerless
-     files, the entire file is scanned and a histogram distribution
-     of bitrates is created, and the average bitrate calculated from
-     that.  (thanks ahØartemis*dk for sample file)
-   Added support for DSIZ chunks in VQF, and checks to make sure size
-     of audio data matches DSIZ value, if present
-     (thanks ahØartemis*dk for sample file)
-   Rewrote GetAllMP3info() - removed some unneccesary code, changed
-     format-detection routine from ParseAsThisFormat() to
-     GetFileFormat() to allow for more flexible format parsing
-     (needed for ISO CD-ROM images, helpful for Quicktime and others)
-   Changed references in all files from string-cast indexes: ["$i"]
-     to non-cast indexes: [$i] where appropriate
-   Put a sans-serif 9pt style on all text in getid3.check.php
-   getAACADTSheaderFilepointer() now return TRUE if synch is lost
-     after the first frame has been successfully parsed (previously
-     it would return FALSE if synch was lost at any time, meaning the
-     file is most likely MP3, which was incorrect)
-     (thanks ahØartemis*dk for sample file)
-   Speed improvement code changes to getid3.mp3.php (up to 24% faster
-     in some cases)  (thanks ahØartemis*dk for the code)
-   Changed all include_once() to require_once()
-
-
-1.5.5: [November-25-2002] James Heinrich
-   » Added support for La (Lossless Audio - www.lossless-audio.com)
-     (thanks ahØartemis*dk for the idea)
-     New file: getid3.la.php
-   € Moved lookup functions from getid3.lookup.php to the files where
-     they are used.
-     New file: getid3.id3.php
-     New file: getid3.rgad.php
-     Removed file: getid3.lookup.php
-   € getID3v1Filepointer() returns FALSE if ID3v1 tag not found
-   € Added new paramter "ReturnExtendedInfo" to the function
-     getAACADTSheaderFilepointer() in getid3.aac.php which now
-     defaults to FALSE - if TRUE then the data for every frame is
-     returned (containing aac_frame_length, adts_buffer_fullness and
-     num_raw_data_blocks, which aren't usually very useful). Speed
-     improvement with FALSE is about 35%.
-   € Now returns fopen() errors in ['error'], for example if a remote
-     file is not accessible.
-   € Changed default number of MP3 audio frames to scan to determine
-     if a valid stream has been found from 5 to 10, now also defined
-     as a constant at the top of getid3.mp3.php  This will result in
-     slightly slower MP3 parsing, but greater reliability in
-     detecting false/invalid/corrupted VBR headers.
-   € fopen() errors now displayed in getid3.putid3.php
-     (thanks miguel.dieckmannØhamburg*de)
-   € Added 4th parameter to decodeMPEGaudioHeader() $ScanAsCBR which
-     will force an MP3 audio frame sequence to be force-scanned in
-     CBR mode. You should never need to call this directly, it's only
-     used internally to scan for MP3 files that have an illegal VBR
-     header with CBR data. (thanks fletchØpobox*com)
-   * Bugfix: ASF_Marker_Object in getid3.asf.php was always returning
-     an error in non-existant "reserved_1" and failing
-   * Bugfix: VBR bitrate calculations in getid3.mp3.php only occur if
-     ['mpeg']['audio']['VBR_frames'] is defined.
-     (thanks fletchØpobox*com)
-   * Bugfix: getid3.putid3.php no longer deletes original MP3 if
-     ID3v2 tag writing fails (thanks miguel*dieckmannØhamburg*de)
-   * Bugfix: incorrect order of if-statement error messages in
-     getid3.putid3.php (thanks miguel*dieckmannØhamburg*de)
-   getid3.asf.php now notes the error and continues parsing rather
-     than failing when it encounters an error parsing a chunk
-   Now actually scan 1000 frames for AAC ADTS as reported in the
-     v1.5.4 changelog, rather than 100. (thanks ahØartemis*dk)
-   Improved scanning speed in getAACADTSheaderFilepointer() by ~30%
-     (thanks ahØartemis*dk for the fix)
-   Added FileSizeNiceDisplay() function to getid3.functions.php for
-     formatting filesize output in kB, MB, GB, etc.
-
-
-1.5.4: [October-07-2002] James Heinrich
-   » Added support for Quicktime.
-     New file: getid3.quicktime.php
-   » Added support for AAC files, both ADTS and ADIF header formats.
-     ADIF format is a pain because it's very similar to standard MP3
-     header format, and it's hard to distinguish between the two. I
-     have tried to make the detection accurate, but I have a limited
-     number of AAC test files to play with so if you have an AAC file
-     that gets detected as MP3/MP2 (or vice-versa), please send me
-     the details via email at getid3Øsilisoftware*com
-     ADTS format is very slow to parse because to get the bitrate of
-     VBR files the whole file has to be stepped through frame by
-     frame (getID3() scans up to the first 1000 frames and assumes
-     that to be close enough).
-     Note: I would suggest commenting out support for AAC (see top of
-     GetAllMP3info() function in getid3.php) unless you need it.
-     (thanks jfaulØgmx*de for the idea and sample Delphi source code)
-     New file: getid3.aac.php
-   » Added bitrate distribution analysis option for MP3 VBR files. A
-     new boolean parameter for getOnlyMPEGaudioInfo() enabled this
-     feature which steps through the MP3 file frame by frame and
-     counts how many frames of each bitrate exist. This information
-     is returned in ['mpeg']['audio']['bitrate_distribution']
-     Caution: this feature is very inefficient for large files and
-     takes a very long time and does lots of disk I/O. Use with care.
-   € Changed layout of allowedFormats in GetAllMP3info() function in
-     getid3.php to allow easy removal of support for any of the
-     supported format. As stated above, I recommend commenting out
-     AAC unless needed.
-   € Added ['flac']['compressed_audio_bytes'],
-     ['flac']['uncompressed_audio_bytes'], and
-     ['flac']['compression_ratio']
-   € Replaced FXPT2DOT30toFloat() function with FixedPoint2_30()
-   * Bugfix: getid3.mpc.php was slightly miscalculating the number of
-     samples, therefore also bitrate and playtime
-     (thanks ahØartemis*dk for the fix)
-   * Bugfix: MonkeyCompressionLevelNameLookup() didn't know about
-     'insane' compression (thanks ahØartemis*dk for the fix)
-   * Bugfix: MonkeySamplesPerFrame() was incorrect for MAC v3.95+
-     (thanks ahØartemis*dk for the fix)
-   * Bugfix: getid3.check.php wasn't processing the assumeFormat
-     directive when (register_globals == off)
-   * Bugfix: detecting of synch pattern for MP3 files with invalid
-     data at the beginning wasn't always correct, also meant possible
-     incorrect bitrate/duration/etc info for such corrupt files.
-   getid3.functions.php now includes a replacement utf8_decode()
-     function for those PHP installations that are not configured
-     with the --with-xml option. (thanks stephaneØtekartists*com)
-
-
-1.5.3: [September-30-2002] James Heinrich
-   » Added support for VQF. (thanks mtØmansonthomas*com for the idea)
-     New file: getid3.vqf.php
-   » Added support for FLAC. Comments, if present, are returned under
-     ['ogg'] because they follow the Ogg Vorbis structure standard.
-     New file: getid3.flac.php
-   € OS/2-format bitmaps are now correctly interpreted. The format of
-     the bitmap is now returned in ['bmp']['type_os'] and
-     ['bmp']['type_version']. OS/2 bitmaps can be v1 or v2, Windows
-     can be v1, v4 or v5
-
-
-1.5.2: [September-25-2002] James Heinrich
-   » Support for RealMedia (audio & video) added
-     Note: only tested on G2 and v5 audio and video files - if anyone
-     has older and/or newer sample files, please test it and/or send
-     me the sample files.
-     (thanks stephaneØtekartists*com for idea)
-     New file: getid3.real.php
-   » Support for BMP added. Palette and pixel data can optionally be
-     extracted as well - this is slow and generally unneccesary, but
-     the option is there if you need it. Also includes PlotBMP()
-     which will take the extracted pixel data and output it as a true
-     color PNG. This function requires GD v2.0+
-     Note: Untested on 16-bit and 32-bit BMPs because I couldn't find
-     any sample files - if you know of a program that can create such
-     files, please email getid3Øsilisoftware*com
-     Note: Support for RGB (uncompressed), RLE8 and RLE4 is included
-     and tested. BITFIELDS support is also included for 16- & 32-bit
-     formats, but it's untested, so if anybody has any test files
-     please send them to getid3Øsilisoftware*com
-     Note: Support currently only for Windows-format BMPs, and trying
-     to parse an OS/2-format bitmap leads to unpredictable/invalid
-     results.
-     New file: getid3.bmp.php
-   » PNG now fully parsed, including all information chunks
-     New file: getid3.png.php
-   € Support for GIF/JPG/PNG moved to seperate files and expanded,
-     including standard ['resolution_x'] and ['resolution_y'] as well
-     as more thorough parsing of header information
-     New file: getid3.gif.php
-     New file: getid3.jpg.php
-   table_var_dump() simplified and now outputs &#123;-style character
-     entities for characters outside the normal alphanumeric range
-   CleanOggCommentName() changed to a regular expression
-     (thanks chris-getid3Øbolt*cx for rewriting the function)
-
-
-1.5.1: [September-20-2002] James Heinrich
-   » Added support for MPEGplus/Musepack SV7. ['fileformat'] is 'SV7'
-     for version 7 files (versions 4, 5 ,6 and 8 are not supported
-     yet, but will be of ['fileformat'] SV4, SV5, SV6 and SV8) when
-     they are supported (thanks Christian Fritz for the idea)
-     New file: getid3.mpc.php
-   € ['bitrate_audio'], ['bitrate_video'], ['bitrate_mode'],
-     ['channels'], ['resolution_x'], and ['resolution_y'] keys added
-     for all appropriate formats
-   € Ogg files with a COVERART comment now save and display the
-     attached image the same way as is done with ID3v2 APICs
-   € ['ogg']['comments'][n]['data'] and
-     ['ogg']['comments'][n]['dataoffset'] is now returned for all
-     comments. ['ogg']['comments'][n]['data'] is only useful if
-     the field is supposed to contain binary data. It is a
-     base64_decode()'d version of ['value'].
-     ['ogg']['comments'][n]['dataoffset'] is the byte offset in the
-     file at which the 'COMMENTNAME=value string' starts, not the
-     start of just 'value'
-   € ['ogg']['comments'][n]['image_mime'] is now returned if
-     ['ogg']['comments'][n]['data'] contains valid image data.
-   € More than 3 Ogg pages may now be read in, if the comment data
-     is longer than 1 page (usually about 4kB)
-   € ['fileformat'] is now 'mp2' rather than 'mp3' if it's MPEG-1,
-     Layer-II audio
-   € ASF bitrates now calculated even if stream_bitrate_properties
-     object not present
-   € ['asf']['stream_properties_object'] is now a numeric-key array
-     with one entry for each stream - the key being the stream number
-   € ['replay_gain'] is returned for all audio formats that support
-     it (MP3-LAME, ID3v2, Ogg) (thanks Christian Fritz for the idea)
-   € ['mpeg']['audio']['LAME']['RGAD']['radio_replay_gain'] is now
-     ['mpeg']['audio']['LAME']['RGAD']['radio'] (same for audiophile)
-   € ASF/WMA files now use WM/Track to get track number from if
-     WM/TrackNumber is not available (thanks stephaneØtekartists*com)
-   € ASF/WMV files now returns ['year'] and ['asf']['year']
-   € ASV/WMV files now use ['content_description']['description'] for
-     the ['comment'] field (thanks stephaneØtekartists*com)
-   € ['track'] is now always returned as an integer
-   * Bugfix: Ogg comments that are larger than one data page (usually
-     about 4kB) are now correctly parsed (thanks Christian Fritz)
-   * Bugfix: Ogg comment data is now UTF8-decoded
-   * Bugfix: Ogg comment writing now UTF8-encodes the data
-   * Bugfix: playtime for ASF files was off by <preroll> (usually
-     between 3 and 12 seconds)
-   * Bugfix: ['asf']['stream_properties_objects']['flags'] data was
-     possibly incorrect
-   * Bugfix: ASF Padding Object was overwriting
-     Stream Bitrate Properties Object data (now returned correctly in
-     ['asf']['padding_object']
-   * Bugfix: ASF Marker Object Reserved_2 field was incorrect
-   * Bugfix: ASF Bitrate Mutual Exclusion Object had incorrect stream
-     numbers
-   Warning displayed if incorrectly-formatted Ogg comment is present
-     (known to be an issue with CDex v1.40, but fixed by v1.50b7)
-     (thanks Christian Fritz)
-   Ogg comment writing now checks for valid comment names
-   Added bitrate column in getid3.check.php, and added some formatting
-     (font, colour)
-   Performance tweaks using bitwise math instead of binary string
-     operations
-
-
-1.5.0: [September-18-2002] James Heinrich
-   » Ogg comment writing support added. getid3.write.php has been
-     updated to allow for writing comment tags to both MP3 and Ogg.
-     Big thanks to Chris Bolt <chris-getid3Øbolt*cx> for writing the
-     OggWrite() function and offering it for inclusion in getID3()
-     New file: getid3.ogginfo.php
-   » Support for Monkey's Audio and APE tag added.
-     (thanks Christian Fritz for the idea)
-     New file: getid3.ape.php
-     ['fileformat'] now returns 'mac' for Monkey's Audio files, or
-     'ape' for files with an APE tag (Monkey's Audio or other format)
-   » getid3.thumbnail.php has been removed from the distribution and
-     the table_var_dump() function now outputs APICs as seperate
-     files in the same directory as the analyzed file. This should
-     make the image-displaying more reliable as well as reduce
-     complexity. The naming convention for the images is
-     filename.ext.[byte offset of APIC data].[jpg|gif|png]
-     If anybody still has any problems with corrupted images please
-     let me know at getid3Øsilisoftware*com
-   » Support for extended Xing/LAME tag
-     (see http://users.belgacom.net/gc247244/extra/tag.html)
-     Data is returned in ['mpeg']['audio']['LAME']
-   € ['ogg']['tracknumber'] has been renamed to ['ogg']['track'] and
-     ['track'] is now returned in the root of the array
-   € ['ogg']['pageheader'][n]['flag'] has been renamed to
-     ['ogg']['pageheader'][n]['flags'] and the unprocessed flag byte
-     is available in ['ogg']['pageheader'][n]['flags_raw']
-   € ['frequency'] is now returned for WAVE files in the root of the
-     array (thanks danielØelectroteque*org)
-   € ASF files now return codec, bitrate, resolution, etc information
-     under ['asf']['video_media'] or ['asf']['audio_media']
-   * Bugfix: RVA2 and EQU2 writing in getid3.putid3.php were
-     incorrectly writing Volume Adjustment field
-   * Bugfix: EQU2 in getid3.frames.php was reading Volume Adjustment
-     as unsigned integer instead of signed integer
-   * Bugfix: handling of remote files over HTTP & FTP was broken
-     (thanks Vince)
-   * Bugfix: incorrect handling of some ASF packets
-   ASF/Windows Media format now more fully parsed, including Index
-     Objects
-   Added several new fourCC video codecs
-
-
-1.4.3: [September-15-2002] James Heinrich
-   » Now parses ASF / WMV / WMA files
-   € New file: getid3.asf.php
-   * Bugfix: RoughTranslateUnicodeToASCII() would return nothing
-     if didn't find a terminator it was expecting
-   Added FILETIMEtoUNIXtime() function (for converting 64-bit
-     Microsoft FILETIME timestamps, used in ASF files and elsewhere,
-     to UNIX Epoch timestamps)
-   Added GUIDtoBytestring() and BytestringToGUID() functions
-
-
-1.4.2: [September-12-2002] James Heinrich
-   » getID3() now requires PHP v4.1.0 or higher because it now is
-     designed to work with register_globals = off and the new auto-
-     globals ($_GET, $_SERVER, etc).
-   * Bugfix: VBR MP3 files with Fraunhofer-style VBR header were not
-     being correctly detected in most cases
-     (thanks dkushnerØoddcast*com and mikeØftl*com for sample files)
-   * Bugfix: IsValidTextEncoding() was broken
-   * Bugfix: Add stripslashes($EditorFilename) to getid3.write.php
-     (writing was broken for files with ' or " in the filename)
-     (thanks mikeØftl*com and kthejoker)
-   * Bugfix: If there is garbage data between a valid VBR header
-     frame and a sequence of valid MPEG-audio frames the VBR data is
-     no longer discarded. (thanks to mikeØftl*com for sample
-     Fraunhofer-style VBR file produced with MusicMatch v7.2)
-   € Changed variable system to work with (register_globals = off)
-   € Moved relevant code into seperate PlaytimeString() function
-   € Added nl2br() to table_var_dump() for cleaner output
-   € Now returns the following keys from Fraunhofer-VBR files:
-     ['VBR_seek_offsets'], ['VBR_seek_offsets_stride'],
-     ['VBR_offsets_relative'] and ['VBR_offsets_absolute']
-   € Added ID3v1matchesID3v2() function and implemented in
-     getid3.check.php (thanks to "Guest" in the forums for the idea)
-   Changed amount of data read in getid3.getimagesize.php from 10kB
-     to entire file. (thanks mikeØftl*com)
-   Wrapped function_exists() checks around function definitions in
-     getid3.functions.php
-   Fixed a lot of E_WARNING and E_NOTICE situations, especially in
-     ID3-writing code (getid3.putid3.php, etc)
-   Added checks to make sure all needed data is available for writing
-     ID3v2 tags
-
-
-1.4.1b5: [May-30-2002] James Heinrich
-   * Bugfix: Unsynchronise() was broken, now fixed
-     (thanks mikeØftl*com)
-   * Bugfix: GenerateID3v2Tag() now correctly uses non-synchsafe
-     integers for frame size descriptors in ID3v2.3 and ID3v2.2
-     (thanks mikeØftl*com)
-   € Added ['artist'], ['title'], etc keys to root of returned
-     array to provide a common place to access any returned info
-     from any file type. Currently gets info from ID3v1, ID3v2,
-     Ogg, and RIFF/WAVE. Possible returned keys are:
-     title, artist, album, year, genre, comment, track
-   € Modified LookupGenre() function to search for either genre based
-     on numeric ID, or now reverse lookup as well
-   € Added ['artist'], ['title'], etc keys to ['RIFF'] information
-     if info tags are present
-   Added functionality to attach a picture to the ID3v2 tag in
-     getid3.write.php
-   Sorted genres into alphabetical order (special 3 at end of list)
-     in getid3.write.php
-   Changed the comment-edit field in getid3.write.php to a multi-line
-     <textarea> from a single-line <input>
-   getid3.write.php now only writes ID3v2 frames that have data
-   Added default TXXX field to getid3.write.php to put a tagger info
-     field when writing ID3v2 tags. Description is "ID3v2-tagged by"
-     and data is "getID3() v[version] (www.silisoftware.com)"
-   Changed getid3.check.php to use the new common info keys
-   Improved file-format detection in getid3.check.php - if the auto-
-     detect based on the first few bytes of the file doesn't find a
-     known format (for example if the header is corrupt), a more
-     thorough scan is done based on the file extension
-   Added 'Edit ID3' link from getid3.check.php to getid3.write.php for
-     MP3 files  (thanks maxØgutalin*com for the idea)
-   Added 'Delete file' link from getid3.check.php to getid3.write.php
-     allowing you to permanently delete a file (be careful with this!!)
-     (thanks maxØgutalin*com for the idea)
-   Added some mouse-over titles for links in getid3.check.php
-
-
-1.4.1b4: [May-15-2002] James Heinrich
-   * Bugfix: getid3.check.php wasn't parsing MP3s with invalid headers
-     or padding at the beginning of the file - added 'assumeFormat'
-     parameter and 'Parse this file as:' options to force parsing in a
-     particular format  (thanks Alcohol for the sample file)
-   * Bugfix: unset(['fileformat']) and ['error'] added in cases where
-     file cannot be parsed in the assumed or forced format
-
-
-1.4.1b3: [May-01-2002] James Heinrich
-   € For Ogg files, now calculates the real average bitrate (returned
-     in ['ogg']['bitrate_average']) and so the playtime of the file is
-     calculated on actual average bitrate, not nominal bitrate, so it
-     should be accurate now  (thanks to stephaneØtekartists*com for
-     telling me it was wrong)
-   * Bugfix: ID3v2FrameIsAllowed() wasn't behaving properly if the
-     writing functions were called for more than one file, because of
-     the static array not being cleared between uses. This is an
-     updated fix because the one in 1.4.1b2 didn't work :o)
-     (thanks soulcatcherØevilsoft*org and yoyo)
-   Added rawurlencode() to the filename parameter in table_var_dump()
-     for images (wouldn't work with path/file names containing special
-     characters (#, &, ", +)  (thanks Christian Fritz)
-   getid3.check.php no longer attempts to scan all MIDI tracks in
-     directory-browse mode, since this can take a long time. Detailed
-     single-file view is still fully scanned (new third parameter for
-     getMIDIHeaderFilepointer() controls this)
-   Small improvements to MoreNaturalSort()
-
-
-1.4.1b2: [April-18-2002] James Heinrich
-   € GetAllMP3Info()'s 2nd parameter has changed from boolean to string
-     (now specifying the parse-this-file-as-this format, like 'mp3',
-     but also can be FALSE to mean don't assume any format, auto-detect
-     only), and a third parameter (array containing allowed formats)
-     has been added. The assumedFormat parameter allows a file to be
-     forced to be parsed as a certain format rather than relying on the
-     auto-detection of getID3() (ex: an MP3 wrapped in a RIFF/WAV
-     header will be auto-detected as RIFF/WAV, but if forced to parse
-     as MP3 will extract the original MP3 information)
-     (thanks reel_tazØusers*sourceforge*net)
-   * Bugfix: ID3v2FrameIsAllowed() wasn't behaving properly if the
-     writing functions were called for more than one file, because of
-     the static array not being cleared between uses (thanks yoyo)
-   * Bugfix: Lyrics3 data wasn't being properly copied from the ['raw']
-     keys to the easy keys (['title'], etc.)  (thanks Christian Fritz)
-   * Bugfix: some testing code was accidentally left in
-     getid3.thumbnail.php  (thanks Christian Fritz)
-   * Bugfix: RIFF/WAVE files are now more likely to have all their
-     chunks parsed.
-   * Bugfix: RIFF/WAVE bitrate & playtime now better calculated
-   * Bugfix: MP3 scanning for synch doesn't go beyond 64k now, to stop
-     intensive scanning through large file that don't have a synch
-     (thanks soulcatcherØevilsoft*org for a weird sample file)
-   Improved performance when scanning for MP3 synch (about 600% faster
-     if the synch is never found)
-   ZIP files no longer return the contents of each compressed file, as
-     that would very easily be more data than PHP could handle.
-     (thanks davidbullockØtech-center*com)
-   getid3.check.php now displays entries in a more natural sort order:
-     case insensitive, ignores most punctuation, treats accented chars
-     the same as their unaccent equivalent  (thanks mikeØftl*com)
-   Added support for SmartSound-format RIFF files (which are regular
-     RIFF/WAVE files with the first 4 chars changed from RIFF to SDSS)
-   All instances of while(list() = each()) replaced with foreach()
-
-
-1.4.1b1: [April-11-2002] James Heinrich
-   » Parses MIDI files.
-     NOTE: very slow at parsing, much slower than any other file type
-     NOTE: playtime is generally mostly accurate, but not always 100%
-   » Parses ZIP files (if ZZIPlib available, and only in PHP 4.0.7RC1
-     and later (see http://www.php.net/manual/en/ref.zip.php)
-     NOTE: currently untested as I'm unable to find php_zip.dll for
-     PHP/Win32 - if someone has a copy of this file, please email me:
-     getid3Øsilisoftware*com
-   » Parses JPEG files (requires GD installed)
-   » Parses PNG files  (requires GD v1.6+ installed)
-   » Parses GIF files  (requires GD < v1.6 installed)
-   » For MP3s, once a valid synch is detected, the next 5 frames are
-     also scanned for valid synch signatures, to prevent false
-     identification of synch. For corrupt MP3 files this will be a bit
-     slower, but hopefully produce more reliable results.
-     (Thanks to mpdjØbtinternet*com for bringing this to my attention,
-     and xbhoffØpacbell*net for explaining what was happening)
-     (Thanks also to macik for helping me with MP3 frame lengths:
-     http://66.96.216.160/cgi-bin/YaBB.pl
-     ?board=c&action=display&num=1018474068)
-   » The actual image data is now displayed (for JPEG, PNG and GIF
-     images only) rather than a binary text dump in getid3.check.php
-     (specifically table_var_dump()) for APIC frames. Made possible
-     by the inclusion of (a modified version of) GetURLImageSize() by
-     Filipe Laborde-Basto (www.rezox.com). You can right-click, save-as
-     to extract the image to a file.
-     NOTE: The actual image data is still returned in ['data']
-   € ['image_mime'], ['image_width'], ['image_height'], ['image_bytes']
-     are now returned for APICs
-   € split parsing functions out into seperate files: lyrics3, id3v1,
-     id3v2, mp3, ogg, riff, mpeg, midi, zip
-   € ['ogg']['bitrate_ave'] -> ['ogg']['bitrate_nominal'] (thanks to
-     stephaneØtekartists*com for pointing out that "nominal" bitrate
-     may actually differ significantly from the "average" bitrate)
-     The real average bitrate seems to be only extractable by parsing
-     the entire file and calculating the average bitrate. This is not
-     yet an option, but hopefully in a future version of getID3()
-   € ['filename'] now returned for all files
-   € ['ogg']['date'] and ['ogg']['description'] now returned when
-     available  (thanks stephaneØtekartists*com)
-   € ['mpeg']['audio']['crc'] now contains the CRC (if present)
-   € ['bitrate'] is now returned as a double instead of an int
-   € ['dataoffset'] is now returned for all ID3v2 frames
-   * Bugfix: MP3 CRC presence ['mpeg']['audio']['protection'] was being
-     reported as opposite of what it actually should be
-   * Bugfix: MPEG videos weren't being detected (they were being
-     parsed as MP3), and even if they were, there was a typo in
-     getMPEGHeaderFilepointer()  (thanks Christian Fritz)
-   * Bugfix: getid3.functions.php wasn't being included in
-     getid3.write.php  (thanks mikeØftl*com)
-   * Bugfix: Browse:___ directory name in getid3.check.php wasn't
-     correct with directory names with ' and other strange characters
-     (thanks Christian Fritz)
-   ID3v2FrameProcessing() now checks to see if the next frame is valid
-     after it encounters an invalid FrameID, and if the next frameID
-     appears valid, it will just skip the current (invalid) frame and
-     continue processing (it would previously abort at the first sign
-     of incorrect structure)   (thanks stephaneØtekartists*com)
-   getid3.check.php now scans filetypes based on content, not filename
-     extension, and shows the filetype in the displayed output. Files
-     are only scanned as MP3 if ID3v2 or MPEG-audio signatures are at
-     the immediate beginning of the file (MP3 used to be the default
-     format), so a corrupt file may not show up as MP3 format in the
-     browse screen, but in detail it will scan in-depth
-   getid3.check.php now has columns to show the presence of ID3v1,
-     ID3v2 and Lyrics3 content
-   Helium2 (www.helium2.com) has been known to write ID3v2.4 tags with
-     non-synchsafe-integer framesizes, getID3() now checks for this and
-     will override and parse the tag as ID3v2.3 if the tag would parse
-     fine as ID3v2.3 when it's really specified as ID3v2.4  (thanks
-     Christian Fritz for the test files)
-
-
-1.4.0b9: [April-05-2002] James Heinrich
-   » Ogg files now return bitrate and playtime (playtime calculated
-     from nominal bitrate and filesize, so it's only approximately
-     accurate).  (thanks stephaneØtekartists*com for the idea)
-   * Bugfix: ID3v1 tags were not properly being parsed - track, genre
-     and comment fields were incorrect.  (thanks Christian Fritz)
-   * Bugfix: getid3.check.php would not browse directories with single
-     quotes (') or double quotes (") in the directory name.
-     (thanks Christian Fritz)
-   * Bugfix: Improved detection of MPEG-video files (a sample MP3 file
-     had a false MPEG video signature at the beginning), and the MPEG-
-     video parsing function now only looks for the MPEG-video header
-     in the first 100k bytes of the file, to prevent needlessly
-     scanning very large files. Also will not infinitely loop if it
-     does not find what it's looking for.  (thanks Christian Fritz)
-   ['error'] now returned if MP3 synch doesn't occur at beginning of
-     file if ID3v2 not used (ie there's some kind of padding there that
-     should not be)
-   Reduced use of fread() in getMPEGHeaderFilepointer() (now faster)
-   Added "file parsed in x.xxx seconds" to getid3.check.php
-   Added "browse: <directory>" link to getid3.check.php
-   Changed default ID3v2 majorversion from 2.4 to 2.3 in
-     getid3.write.php because Winamp (and probably many other
-     ID3v2-aware tools) can only read up to ID3v2.3
-     (thanks mikeØftl*com)
-
-
-1.4.0b8: [April-04-2002] James Heinrich
-   » Lyrics3 support added  (thanks Christian Fritz for the idea)
-   € check.php renamed to getid3.check.php
-   € write.php renamed to getid3.write.php
-   € ['id3']['id3v2']['error'] (if present) now reported in ['error']
-   € ['mpeg']['audio']['error'] (if present) now reported in ['error']
-   * Bugfix: RoughTranslateUnicodeToASCII() was completely mangling
-     UTF-16/UTF-16BE encoded text
-   * Bugfix: The warning about MP3ext wasn't always showing up
-     (thanks davidbullockØtech-center*com)
-   getID3v1Filepointer() cleaned up & shortened
-   Moved the include_once() statements around so that a minimum of code
-     is included
-
-
-1.4.0b7: [April-03-2002] James Heinrich
-   » RIFFs (specifically AVIs) are now more completely parsed,
-     almost everything in the returned ['RIFF'] array has been moved
-     around and/or restructured. A lot of new data is in there too -
-     codecs, frame size, etc.
-   € Better recursive parsing of RIFFs (sub-arrays are now in the right
-     place)
-   * Bugfix: the isset() idea introduced in beta 5 was incorrectly
-     implemented, such that ['asciidata'] and ['asciidescription'] were
-     never returned - this had the side effect that ID3v2 comments were
-     not copied to ['id3']['id3v2']['comment']  (thanks mikeØftl*com)
-   * Bugfix: MPEG audio synch wasn't being detected, and therefore MPEG
-     audio data not parsed, if no ID3v2 header present in an MP3
-   ID3v1 track number only returned if greater than zero
-   Removed !== FALSE (introduced in 1.4.0b6) from while(fread()) loops,
-     some users were reporting problems with that syntax.
-   Changed substr($string, 0, 1) to $string{0} syntax in most files
-   Reformatted changelog.txt to 72-column width
-
-
-1.4.0b6: [April-01-2002] James Heinrich
-   * Bugfix: 1.4.0b5 introduced a bug where any RIFF file other than
-     PCM WAVE (this includes any compressed WAV, as well as all AVIs)
-     would crash getID3()
-   Reduced use of fread() in getOggHeaderFilepointer() for increased
-     speed
-   Added constant FREAD_BUFFER_SIZE for many fread() operations
-   Added !== FALSE check to while(fread()) loops
-     (thanks davidbullockØtech-center*com)
-   Added more entries to RIFFwFormatTagLookup()
-     (still looking for a good complete list)
-   Converted use of hexdec() in getid3.lookup.php to 0x1234 notation
-
-
-1.4.0b5: [March-28-2002] James Heinrich
-   € Renamed decodeheader() to decodeMPEGaudioHeader()
-   * Bugfix: Fixed infinite loop problem for RIFF/WAV files with
-     unknown chunks
-   * Bugfix: WXXX frames were incorrectly writing from ['URL'] instead
-     of ['url']
-   * Bugfix: RoughTranslateUnicodeToASCII() wasn't properly decoding
-     UTF-16/UTF-16BE
-   Changed all quoted strings from " to ' to hopefully improve speed
-     (although benchmarks have not yet shown any significant
-     improvement in speed)  (thanks davidbullockØtech-center*com)
-   Improved code in check.php for dealing with symbolic links
-     (thanks davidbullockØtech-center*com)
-   Changed '<?' tags to '<?php'  (thanks davidbullockØtech-center*com)
-   Added processing time indicator in check.php
-     (ie 'directory scanned in 2.45 seconds')
-   Replaced all instances of feof() to prevent infinite loop conditions
-   Moved lookup portions of decodeMPEGaudioHeader() to
-     getid3.lookup.php
-   Replaced $arrayname[$index] with $arrayname["$index"] to avoid PHP
-     E_NOTICEs  (thanks davidbullockØtech-center*com)
-   Wrapped isset() around many if statements, to avoid PHP E_NOTICEs,
-     hence improve speed (up to 30x speed improvement reported in some
-     cases :)
-
-
-1.4.0b4: [March-26-2002] James Heinrich
-   € RIFF/WAV file format now parsed, returned under ['riff']
-   € Support for Relative Gain Adjustment in RIFF/WAV files
-   € ['channels'] (1 or 2) now returned for MP3 and WAV files
-   € ['bitrate'] now returned (in bits-per-second) at root level for
-     MP3 and WAV files
-   Added support for RGAD (Relative Gain ADjustment) ID3v2 frames, both
-     reading & writing
-     (see http://privatewww.essex.ac.uk/~djmrob/replaygain/ for details
-     on RGAD)  (thanks Christian Fritz for the idea)
-   Removed some test data-dumping from the ID3v2 writing functions
-   Language code 'XXX' now returns descriptive string 'unknown' instead
-     of NULL
-   Seperated out comments from top of getid3.php into getid3.readme.txt
-     and changelog.txt
-   Split out non-lookup functions from getid3.lookup.php to
-     getid3.functions.php
-
-
-1.4.0b3: [March-25-2002] James Heinrich
-   € ['asciidata'] for WXXX frames now returns correct information, but
-     under ['asciidescription']  (thanks Christian Fritz)
-   € Added ['framenamelong'] to all returned frame data arrays with
-     text description of that frame (ie 'RVA2' would return 'Relative
-     volume adjustment (2)')  (thanks Christian Fritz)
-   € ['datalength'] is now ['indexeddatalength'] in ASPI frames (was
-     confliciting with the all-frames ['datalength'] as introduced in
-     v1.4.0b1
-   € ['datalength'] now returned as integer (rather than double) where
-     possible
-
-
-1.4.0b2: [March-21-2002] James Heinrich
-   € ['mpeg']['audio']['bitrate'] now returned as int rather than
-     double for VBR files
-   * Bugfix: MPEG audio information wasn't being parsed on files that
-     had neither ID3v1 or ID3v2
-   * Bugfix: COMM/WXXX frames weren't returning 'asciidata' in
-     ID3v2.2, which also meant the ['id3']['id3v2']['comment'] field
-     wasn't being returned  (thanks stephaneØtekartists*com)
-   * Bugfix: file might not be found if filename actually contains
-     escaped chars or %xx-formatted characters
-     (thanks reel_tazØusers*sourceforge*net)
-   Added support for running with Register Globals turned off
-     (thanks reel_tazØusers*sourceforge*net)
-   Added urlencode() where needed in check.php
-     (thanks reel_tazØusers*sourceforge*net)
-   Fixed IE buffering/display problem in progress counter in check.php
-
-
-1.4.0b1: [March-11-2002] James Heinrich
-   » ID3v2 writing support via WriteID3v2() in putid3.php
-     RemoveID3v2() and RemoveID3v1() functions now available in
-     putid3.php  All ID3v1 and ID3v2 writing functions have been moved
-     to putid3.php and example file write.php has been added to the
-     distribution
-   € MPEG audio frame information (bitrate, frequency, etc) now
-     returned inside ['mpeg']['audio'] instead of just ['mpeg']
-   € MPEG video information now parsed, returned in ['mpeg']['video']
-     Note: audio portion of video system files is not yet being parsed
-   € All flag bits are now returned as boolean rather than int or
-     string
-   € RVA2 data now returned as an array (multiple RVA2 tags are
-     allowed)
-   € RVA2/EQU2 description returned under ['description'] rather than
-     ['identification']
-   € RVAD/EQUA adjustments now returned as signed integers, rather than
-     absolute values which required you to check flag bytes
-   € RVRB/REV data no longer returns under ['reverb'] array
-   € WXXX/W???/LINK frames now return ['url'] instead of ['URL']
-   € USER now properly returns both ['language'] and ['languagename']
-   € OWNE now returns ['purchasedateunix'] as a UNIX timestamp
-     (only if ['purchasedate'] is a valid date)
-   € ['id3']['id3v2']['padding'] now returned with information on padding
-   € ['headerlength'] now includes the initial 6 or 10 bytes of the
-     ID3v2 header
-   € ['artist'], ['title'], ['album'], ['tracknumber'], ['genre'] now
-     returned for easier access for Ogg files
-   € added ['datalength'] to all ID3v2 frames: length of frame data,
-     not including frame header
-   € ['fileformat'] now returns 'id3' if there are ID3v1 or ID3v2 tags
-     but no audio data
-   € ['fileformat'] now returns 'mpg' if it's an MPEG system (video +
-     audio) file
-   * Bugfix: RVAD was being parsed incorrectly
-   * Bugfix: ['currency'] and ['purchasedate'] now correctly returned
-     in OWNE
-   * Bugfix: Frequncies in 'EQU2' frames were incorrectly double
-   * Bugfix: ['bytedeviation'] and ['msdeviation'] now properly
-     returned as integer rather than binary string for 'MLLT' frames
-   * Bugfix: ['filename'] now properly returned for 'GEOB' frames
-   * Bugfix: ['imagetype'] now properly returned for 'PIC' frames in
-     ID3v2.2
-   * Bugfix: Genre not being written if not set in WriteID3v1()
-     (thanks reel_tazØusers*sourceforge*net)
-   * Bugfix: Changed write mode to 'r+b' from 'a+' because ID3v1 tags
-     were being appended instead of overwritten if they already existed
-     (thanks reel_tazØusers*sourceforge*net)
-   * Bugfix: open would fail on filenames containing quotes
-     (thanks javierØcrackdealer*com)
-   * Bugfix: various values were incorrectly returned (unneeded ord())
-     in these frames: COMR, USER, ENCR, GRID, PRIV, SIGN
-   * Bugfix: ASPI ['bitsperpoint'] was not correctly returned
-   * Bugfix: RoughTranslateUnicodeToASCII() was not returning the last
-     char for UTF-16
-   * Bugfix: ['audiobytes'] now correctly 0 if no synch found
-   * Bugfix: GenreLookup was incorrectly returning 'Remix' instead of
-     'Blues' for GenreID 0
-   Added sample directory browser to check.php
-   Seperated out MPEGaudio-parsing functionality into
-     getOnlyMPEGaudioInfo() which may be called directly if you don't
-     need any ID3 parsing  (thanks djpretzelØcox*rr*com for idea)
-   Reduced use of fread() for increased performance in
-     getID3v1Filepointer()
-   Added clearstatcache() before checking filesize - size after writing
-     tag now correct
-   Added hack for mp3Rage (www.chaoticsoftware.com) that puts
-     ID3v2.3-formatted MIME type instead of 3-char ID3v2.2-format image
-     type  (thanks xbhoffØpacbell*net for test file)
-
-
-1.3.2: [February-15-2002] James Heinrich
-   € UFID/UFI, USLT/ULT, COMM/COM, APIC/PIC, GEOB/GEO, CRM, RVA2, EQU2,
-     POPM/POP, AENC/CRA, ENCR and GRID frame data now returned under
-     numeric array index rather than by ownerID
-   € RVA2 frame data is now returned keyed by $channeltypeid instead of
-     $frame_idstring
-   € WXXX/WXX frame description now returned under ['description']
-     instead of ['data']
-   Trailing null bytes now trimmed from frame (W??? & T???) text data
-     (it shouldn't be there to begin with, but a sample file encoded by
-     [unknown program] had data padded to 50 chars with null bytes,
-     which caused ParseID3v2GenreString() to freeze).
-
-
-1.3.1: [February-13-2002] James Heinrich
-   * Bugfix: ['playtime_seconds'] and ['playtime_string'] were not
-     being returned
-   * Bugfix: ['fileformat'] was incorrectly being returned as a
-     2-element array
-   * Bugfix: USLT wasn't being correctly parsed
-   Improved RoughTranslateUnicodeToASCII()
-     (thanks reel_tazØusers*sourceforge*net for Unicode test file)
-
-
-1.3.0: [February-13-2002] James Heinrich
-   » ID3v1 writing support via WriteID3v1()
-   € MPEG audio frame information (bitrate, frequency, etc) now
-     returned inside ['mpeg']
-   € ['mpeg']['raw'] returns the integer values of the bits for MPEG
-     audio information as returned in ['mpeg'] by decodeheader()
-     (thanks reel_tazØusers*sourceforge*net)
-   € 'protection', 'padding', 'private', 'copyright' and 'original' now
-     return as boolean
-   € 'bitrate' and 'frequency' now return as int (except in special
-     case of 'free')
-   Language name as well as code retured where appropriate
-     (ie 'English' and 'eng')
-   Text frames with invalid TextEncoding value are now passed through
-     anyway
-   ID3v1 data (title, artist, album, year, comment) is now trimmed
-     (no more nulls)
-   RoughTranslateUnicodeToASCII() now uses utf8_decode() for UTF-8
-
-
-1.2.5: [January-30-2002] James Heinrich
-   * Bugfix: Playtime calculations for VBR files were off slightly
-     (rounding error)
-   * Bugfix: Extended header length was incorrectly calculated
-   * Bugfix: Genre strings such as '03' weren't being handled correctly
-   More complete support for ID3v2.3 FrameIDs
-   Split out getid3.frames.php (FrameID-specific parsing function)
-   Split out getid3.lookup.php (assorted lookup-table functions)
-   Searches for what directory getid3.*.php support files are in (must
-     be same as getid3.php, but doesn't have to be same as main file -
-     for example your main file could be /index.php, but including
-     /lib/getid3/getid3.php)
-   Simplified, tweaked, changed and/or eliminated several functions.
-
-
-1.2.4: [January-26-2002] James Heinrich
-   » Basic support for reading Ogg-Vorbis comment tags
-
-
-1.2.3: [January-24-2002] James Heinrich
-   » ID3v2.2.x 3-char FrameIDs are now fully parsed
-     Note: While I've included support for 22 FrameIDs as defined in
-     the specs, I don't have test files for all of them. If anyone
-     knows of programs that generate any of the untested tags, please
-     email getid3Øsilisoftware*com ! Here's what's tested and not:
-       Tested: T??, COM
-     Untested: UFI, TXX, W??, WXX, IPL, MCI, ETC, MLL, STC, ULT, SLT,
-               RVA, EQU, REV, PIC, GEO, CNT, POP, BUF, CRM, CRA, LNK
-   table_var_dump() now displays boolean variables as TRUE or FALSE
-   table_var_dump() now uses htmlspecialchars() to avoid broken-table
-     problems
-
-
-1.2.2: [January-18-2002] James Heinrich
-   € Parses ID3v2 genres into ['id3']['id3v2']['genreid'] and
-     ['id3']['id3v2']['genrelist'] where appropriate
-     (thanks stephaneØtekartists*com for the idea)
-   Added ID3v2 genre abbreviations 'RX' (remix) and 'CR' (cover)
-
-
-1.2.1: [January-17-2002] James Heinrich
-   * Bugfix: 'mp3' was being returned in ['format'], but 'zip' was
-     being returned in ['fileformat'], both are now returned in
-     ['fileformat']
-   € Splits ['id3']['id3v2']['track'] in the format '5/12' into
-     ['track'] = '5' and ['totaltracks'] = '12'
-   € Enabled ['id3']['id3v2']['title'] etc for ID3v2.2.x
-     (3-char frame names)  (thanks stephaneØtekartists*com)
-   € Changed v1.?? version number format to v1.?.?
-   Scans through the file until it finds the MPEG synch (start of audio
-     frame) - some files encoded by LAME 3.91 had undocumented padding
-     after the ID3v2 header; getMP3headerFilepointer() now scans until
-     it finds synch (or EOF)  (thanks adamØtrekjapan*com)
-   Improved Unicode conversion in RoughTranslateUnicodeToASCII()
-
-
-1.20:  [January-15-2002] James Heinrich
-   » Support for variable-bitrate (VBR) files, both Xing and Fraunhofer
-     headers
-   » All 4-character FrameIDs are now fully parsed according to the
-     specs at http://www.id3.org/id3v2.4.0-frames.txt
-     € This means that most no longer return ['flags'] and ['data']
-     Note: While I've included support for 30 FrameIDs as defined in
-     the specs, I don't have test files for all of them. If anyone
-     knows of programs that generate any of the untested tags, please
-     email getid3Øsilisoftware*com ! Here's what's tested and not:
-       Tested: UFID, T???, WXXX, USLT, SYLT, COMM, APIC, GEOB
-     Untested: TXXX, W???, MCDI, ETCO, MLLT, SYTC, RVA2, EQU2, RVRB,
-               PCNT, POPM, RBUF, AENC, USER, OWNE, COMR, ENCR, GRID,
-               PRIV, SIGN, SEEK, ASPI
-   € Added 'title', 'artist', etc names to ID3v2 data (easier to access
-     than the 4-character FrameIDs of the ID3v2 standard)
-     (thanks jaksonØgmx.net)
-   * Bugfix: added fclose() at end of GetAllMP3Info()
-     (thanks stephaneØtekartists*com)
-   * Bugfix: ID3v1 wasn't being parsed if ID3v2 wasn't present
-     (thanks jaksonØgmx.net)
-   * Bugfix: several flags were being parsed incorrectly (the structure
-     had changed from ID3v2.3 to ID3v2.4) - v2.3 flags were being
-     incorrectly parsed
-   Much more compact implementation of decodeheader()
-     (thanks jaksonØgmx.net for the idea)
-   ID3v1 genres 126 through 147  (thanks jaksonØgmx.net)
-   New table_var_dump() function in check.php
-     (based partially on idea by jaksonØgmx.net)
-   Seperated ID3v1 retrieval into seperate function
-
-
-1.11:  [December-23-2001] James Heinrich
-   All functions merged into file getid3.php
-   Updated documentation to reflect new returned information
-
-
-1.10:  [December-20-2001] James Heinrich
-   * Bugfix: ID3v1 Track# was incorrectly being parsed whether it
-     existed or not
-   Changed calling procedure to recommend using
-     GetAllMP3info($filename) from getmp3header.php
-   Now includes check.php - example file
-   € Checks to see if file is in ZIP or MP3 format
-     (returned in ['format'])
-     [Ed. Note: ['fileformat'] as of v1.2.1]
-
-
-1.06:  [November-05-2001] James Heinrich
-   * Bugfix: ID3v2.2.x frames weren't being parsed since they use
-     6-byte rather than 10-byte headers as v2.3+ does
-     (thanks spunkØmac*com for pointing that out)
-
-
-1.05:  [September-06-2001] James Heinrich
-   * Bugfix: ID3v2 was being parsed even if it didn't exist
-
-
-1.04:  [July-16-2001] James Heinrich
-   * Bugfix: typo in Extended Header section (strpad() should be
-     str_pad()) (thanks jurroonØyahoo*com)
-
-
-1.03:  [May-07-2001] James Heinrich
-   * Bugfix: Added missing ['id3']['id3v1']['genreid'] and
-     ['id3']['id3v1']['genre']
-
-
-1.02:  [May-05-2001] James Heinrich
-   € Added ['getID3version']
-
-
-1.01:  [May-04-2001] James Heinrich
-   » Added support for frame-level de-unsynchronisation (as per
-     ID3v2.4.0 specs) in addition to ID3v2.3.x tag-level
-     de-unsynchronisation
-
-
-1.00:  [May-04-2001] James Heinrich
-   » Initial public release
-
-
-///////////////////////////////////////////////////////////////////////
-
-Future Plans
-============
-
- Features
- --------
-
-  * Writing support for Real
-  * Better support for MP4 container format
-  * Support for Matroska (www.matroska.org)  (thanks ahØartemis*dk)
-    http://corecodec.com/modules.php?op=modload&name=PNphpBB2&file=viewtopic&t=227
-  * scan for appended ID3v2 tag at end of file per ID3v2.4 specs (Section 5.0)
-  * Support for JPEG-2000 (http://www.morgan-multimedia.com/jpeg2000_overview.htm)
-  * Support for MOD (mod/stm/s3m/it/xm/mtm/ult/669) (thanks ahØartemis*dk)
-  * Support for FROG (http://ghido.shelter.ro/FROG.php)
-  * Lyrics3 v1 & v2 writing support
-  * Support for gzip
-  * Support for ACE (thanks Vince)
-  * Support for Ogg other than Vorbis, Speex and OggFlac (ie. Ogg+Xvid)
-    (thanks ahØartemis*dk)
-  * Ability to create Xing/LAME VBR header for VBR MP3s that are missing VBR header
-  * Ability to "clean" ID3v2 padding (replace invalid padding with valid padding)
-  * Ability to convert RIFF-MP3 to regular MP3 (strip RIFF headers)
-  * Warn if MP3s change version mid-stream (in full-scan mode)
-  * check for corrupt/broken mid-file MP3 streams in histogram scan
-  * Support for lossless-compression formats
-    (http://www.firstpr.com.au/audiocomp/lossless/#Links)
-    (http://compression.ca/act-sound.html)
-    (http://web.inter.nl.net/users/hvdh/lossless/lossless.htm)
-    * Support for Sonarc (http://www.firstpr.com.au/audiocomp/lossless/sonarc/)
-    * Support for WavArc (http://www.firstpr.com.au/audiocomp/lossless/wavarc/)
-    * Support for WaveZip/MUSICompress (http://hometown.aol.com/sndspace)
-    * Support for LTAC (http://www.nue.tu-berlin.de/wer/liebchen/ltac.html)
-  * Support for RIFF-INFO chunks
-    * http://lotto.st-andrews.ac.uk/~njh/tag_interchange.html
-      (thanks Nick Humfrey <njhØsurgeradio*co*uk>)
-    * http://abcavi.narod.ru/sof/abcavi/infotags.htm
-      (thanks Kibi)
-  * Better support for Bink video
-  * http://www.hr/josip/DSP/AudioFile2.html
-  * http://www.pcisys.net/~melanson/codecs/
-  * http://sox.sourceforge.net/AudioFormats-11.html
-    * DiamondWare Digitized .dwd
-    * Tandy Deskmate .snd
-    * Sample Vision
-  * Detect mp3PRO
-  * Support for PSD
-  * Support for JPC
-  * Support for JP2
-  * Support for JPX
-  * Support for JB2
-  * Support for IFF
-  * Support for ICO
-  * Support for ANI
-  * Support for EXE (comments, author, etc) (thanks p*quaedackersØplanet*nl)
-  * Support for DVD-IFO (region, subtitles, aspect ratio, etc)
-    (thanks p*quaedackersØplanet*nl)
-  * More complete support for SWF - parsing encapsulated MP3 and/or JPEG content
-      (thanks n8n8Øyahoo*com)
-  * Support for a2b
-  * MPC-SV8 (http://www.uni-jena.de/~pfk/mpp/sv8/components.html)
-    http://www.personal.uni-jena.de/~pfk/mpp/
-  * Optional scan-through-frames for AVI verification
-    (thanks rockcohenØmassive-interactive*nl)
-  * Support for TTF (thanks infoØbutterflyx*com)
-  * Support for DSS (http://www.getid3.org/phpBB2/viewtopic.php?t=171)
-  * Support for SMAF (http://smaf-yamaha.com/what/demo.html)
-    http://www.getid3.org/phpBB2/viewtopic.php?t=182
-  * Support for AMR (http://www.getid3.org/phpBB2/viewtopic.php?t=195)
-  * Support for 3gpp (http://www.getid3.org/phpBB2/viewtopic.php?t=195)
-  * Support for ID4 (http://www.wackysoft.cjb.net grizlyY2KØhotmail*com)
-  * Parse XML data returned in Ogg comments
-  * Parse XML data from Quicktime SMIL metafiles (klausrathØmac*com)
-  * ID3v2.2 tag writing support
-  * ID3v2 genre string creator function
-  * Support for optional XML-format output (thanks moisei for the idea)
-  * More complete parsing of JPG
-  * Support for all old-style ASF packets
-  * ASF framerate guess
-  * ASF/WMA/WMV tag writing
-  * Parse declared T??? ID3v2 text information frames, where appropriate
-      (thanks Christian Fritz for the idea)
-  * Recognize encoder:
-    http://www.guerillasoft.com/EncSpot2/index.html
-    http://ff123.net/identify.html
-    http://www.hydrogenaudio.org/?act=ST&f=16&t=9414
-    http://www.hydrogenaudio.org/?showtopic=11785
-  * Support for other OS/2 bitmap structures: Bitmap Array('BA'),
-    Color Icon('CI'), Color Pointer('CP'), Icon('IC'), Pointer ('PT')
-    http://netghost.narod.ru/gff/graphics/summary/os2bmp.htm
-  * Support for WavPack RAW mode
-  * ASF/WMA/WMV data packet parsing
-  * ID3v2FrameFlagsLookupTagAlter()
-  * ID3v2FrameFlagsLookupFileAlter()
-  * obey ID3v2 tag alter/preserve/discard rules
-  * http://www.geocities.com/SiliconValley/Sector/9654/Softdoc/Illyrium/Aolyr.htm
-  * proper checking for LINK/LNK frame validity in ID3v2 writing
-  * proper checking for ASPI-TLEN frame validity in ID3v2 writing
-  * proper checking for COMR frame validity in ID3v2 writing
-  * http://www.geocities.co.jp/SiliconValley-Oakland/3664/index.html
-  * decode GEOB ID3v2 structure as encoded by RealJukebox,
-    decode NCON ID3v2 structure as encoded by MusicMatch
-    (probably won't happen - the formats are proprietary)
-
-
- Known Bugs/Issues in getID3() that may be fixed eventually
- -----------------------------------------------------------
-
-  * Cannot determine bitrate for MPEG video with VBR video data
-    (need documentation)
-  * Interlace/progressive cannot be determined for MPEG video
-    (need documentation)
-  * MIDI playtime is sometimes inaccurate
-  * AAC-RAW mode files cannot be identified
-  * WavPack-RAW mode files cannot be identified
-  * mp4 files report lots of "Unknown QuickTime atom type"
-     (need documentation)
-  * Encrypted ASF/WMA/WMV files warn about "unhandled GUID
-    ASF_Content_Encryption_Object"
-  * Bitrate split between audio and video cannot be calculated for
-    NSV, only the total bitrate. (need documentation)
-  * All Ogg formats (Vorbis, OggFLAC, Speex) are affected by the
-    problem of large VorbisComments spanning multiple Ogg pages, but
-    but only OggVorbis files can be processed with vorbiscomment.
-  * The version of "head" supplied with Mac OS 10.2.8 (maybe other
-    versions too) does only understands a single option (-n) and
-    therefore fails. getID3 ignores this and returns wrong md5_data.
-
-
-
- Known Bugs/Issues in getID3() that cannot be fixed
- --------------------------------------------------
-
-  * Files larger than 2GB (of any format) cannot be parsed by
-    getID3() due to limitations in the PHP filesystem functions
-
-
- Known Bugs/Issues in other programs
- -----------------------------------
-
-  * Winamp (up to v2.80 at least) does not support ID3v2.4 tags,
-      only ID3v2.3
-      see: http://forums.winamp.com/showthread.php?postid=387524
-  * Some versions of Helium2 (www.helium2.com) do not write
-      ID3v2.4-compliant Frame Sizes, even though the tag is marked
-      as ID3v2.4)  (detected by getID3())
-  * MP3ext V3.3.17 places a non-compliant padding string at the end
-      of the ID3v2 header. This is supposedly fixed in v3.4b21 but
-      only if you manually add a registry key. This fix is not yet
-      confirmed.  (detected by getID3())
-  * CDex v1.40 (fixed by v1.50b7) writes non-compliant Ogg comment
-      strings, supposed to be in the format "NAME=value" but actually
-      written just "value"  (detected by getID3())
-  * Oggenc 0.9-rc3 flags the encoded file as ABR whether it's
-      actually ABR or VBR.
-  * iTunes (versions "X v2.0.3", "v3.0.1" are known-guilty, probably
-      other versions are too) writes ID3v2.3 comment tags using a
-      frame name 'COM ' which is not valid for ID3v2.3+ (it's an
-      ID3v2.2-style frame name)  (detected by getID3())
-  * MP2enc does not encode mono CBR MP2 files properly (half speed
-      sound and double playtime)
-  * MP2enc does not encode mono VBR MP2 files properly (actually
-      encoded as stereo)
-  * tooLAME does not encode mono VBR MP2 files properly (actually
-      encoded as stereo)
-  * AACenc encodes files in VBR mode (actually ABR) even if CBR is
-     specified
-  * AAC/ADIF - bitrate_mode = cbr for vbr files
-  * LAME 3.90-3.92 prepends one frame of null data (space for the
-    LAME/VBR header, but it never gets written) when encoding in CBR
-    mode with the DLL
-  * Ahead Nero encodes TwinVQF with a DSIZ value (which is supposed
-    to be the filesize in bytes) of "0" for TwinVQF v1.0 and "1" for
-    TwinVQF v2.0  (detected by getID3())
-  * Ahead Nero encodes TwinVQF files 1 second shorter than they
-    should be
-  * AAC-ADTS files are always actually encoded VBR, even if CBR mode
-    is specified (the CBR-mode switches on the encoder enable ABR
-    mode, not CBR as such, but it's not possible to tell the
-    difference between such ABR files and true VBR)
-  * STREAMINFO.audio_signature in OggFLAC is always null. "The reason
-    it's like that is because there is no seeking support in
-    libOggFLAC yet, so it has no way to go back and write the
-    computed sum after encoding. Seeking support in Ogg FLAC is the
-    #1 item for the next release." - Josh Coalson (FLAC developer)
-    NOTE: getID3() will calculate md5_data in a method similar to
-    other file formats, but that value cannot be compared to the
-    md5_data value from FLAC data in a FLAC file format.
-  * STREAMINFO.audio_signature is not calculated in FLAC v0.3.0 &
-    v0.4.0 - getID3() will calculate md5_data in a method similar to
-    other file formats, but that value cannot be compared to the
-    md5_data value from FLAC v0.5.0+
-  * RioPort (various versions including 2.0 and 3.11) tags ID3v2 with
-    a WCOM frame that has no data portion
-  * Earlier versions of Coolplayer adds illegal ID3 tags to Ogg Vorbis
-    files, thus making them corrupt.
-  * Meracl ID3 Tag Writer v1.3.4 (and older) incorrectly truncates the
-    last byte of data from an MP3 file when appending a new ID3v1 tag.
-    (detected by getID3())
-  * Lossless-Audio files encoded with and without the -noseek switch
-    do actually differ internally and therefore cannot match md5_data
-  * iTunes has been known to append a new ID3v1 tag on the end of an
-    existing ID3v1 tag when ID3v2 tag is also present
-    (detected by getID3())