File: wavpack_doc.html

package info (click to toggle)
wavpack 5.9.0-1
links: PTS, VCS
area: main
in suites: forky, sid
size: 4,340 kB
sloc: ansic: 27,542; asm: 9,843; sh: 5,640; makefile: 172
file content (1832 lines) | stat: -rw-r--r-- 88,479 bytes
<!DOCTYPE html
	PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
	"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en">
	<head>
		<meta http-equiv="content-type" content="text/html; charset=iso-8859-1"/>
  <meta http-equiv="Expires" content="Sat, 13 Mar 1982 09:10:27 GMT"/>
      	<style type="text/css">
		body {
	background: #ffffff;
	color: #000000;
	font-family: "Verdana", sans-serif;
	font-size: 10pt;
	text-align: center;
}

.container {
	top: 0;
	bottom: 0;
	left: 0;
	right: 0;
	width: 960px;
	margin-left: auto; 
	margin-right: auto;
}

a {
	text-decoration: underline;
	color: #000000;
}

a:hover {
	color: #8394b2;
}

.pageTitle, .pageMenu, .boxHeader, .boxContent {
	padding: 2px;
}

.pageTitle, .pageMenu, .box, .boxHeader, .filesHeader {
	border: 1px solid;
}

.pageTitle, .pageMenu, .box, .boxContent {
	margin: 2px;
}

.pageTitle, .pageMenu, .box {
	border-color: #072a66;
}

.pageHeader, .subHeader, .boxHeader, .filesHeader {
	font-weight: bold;
}

.pageMenu, .boxHeader {
	background: #bcd0ed;
}

.boxHeader, .filesHeader {
	color: #5176b5;
	border-color: #ffffff;
}

.pageTitle {
	background: #618ece;
	color: #ffffff;
}

.pageHeader {
	font-size: 14pt;
}

.pageSubHeader {
	font-size: 12pt;
}

.box {
	background: #eef2f7;
	border-color: #072a66;
}

.boxHeader {
	letter-spacing: 2px;
	word-spacing: 2px;
}

.boxContent {
	text-align: justify;
}

.filesBox {
	width: 100%;
}

.filesHeader {
	border-color: #bcd0ed;
	text-align: center;
}

.hidden, .hidden a {
	color: #c0c0c0;
}
		</style>
   		<title>WavPack User Documentation</title>
	</head>
	<body>
		<div class="container">
			<div class="pageTitle">
				<div class="pageHeader">Using WavPack Version 5.9.0</div>
			</div>
			<div class="pageMenu">
				<a href="index.html">Home</a> - 
				<a href="#linux">Linux</a> -
				<a href="#wavpack">WavPack options</a> - 
				<a href="#wvunpack">WvUnpack options</a> - 
				<a href="#wvgain">WvGain options</a> - 
				<a href="#wvtag">WvTag options</a> -
				<a href="#usage">Usage Guide</a> - 
				<a href="#noncd">Non-CD</a> - 
				<a href="#plugins">Plugins</a>
			</div>
			<br/>
			<div class="box" id="Intro">
				<div class="boxContent">
					WavPack consists of two complementary command-line programs, WavPack and WvUnpack. These
					programs allow you to compress (and restore) audio files in several formats including the
     Microsoft .wav format. The audio files must be uncompressed PCM (not ADPCM, for example),
     but other than that there are few restrictions. The files may be any resolution from 8 to
     32 bits; they may be mono, stereo, or multichannel; they may even be IEEE floating point
     data or 1-bit DSD audio (in those formats that support DSD).
    <p>
     Two additional utilities are provided to manipulate metadata stored in APEv2 format
     tags located at the end of WavPack files. The first, WvGain, is used to apply (after
     loudness analysis) ReplayGain information to WavPack files and the other, WvTag, is a
     more general utility for appending or removing all kinds of other metadata items from
     WavPack files (including cover art and cuesheets).
    </p>
				<p>
     The programs are initiated from the command line with the following syntax:
    </p>
				<p>
					<b>Usage:</b>
					<br/>
					<b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;WAVPACK [-options] infile[.wav]|infile.ext|- [outfile[.wv]|outpath|-]</tt></b>
					<br/>
					<b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;WAVPACK --drop [-options] infile[.wav]|infile.ext [...]</tt></b>
					<br/>
					<b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;WVUNPACK [-options] infile[.wv]|- [outfile[.ext]|outpath|-]</tt></b>
					<br/>
					<b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;WVUNPACK --drop [-options] infile[.wv] [...]</tt></b>
					<br/>
     <b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;WVGAIN [-options] infile[.wv] [...]</tt></b>
					<br/>
     <b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;WVTAG [-options] infile[.wv] [...]</tt></b>
				</p>
				<p>
					The filename extensions will be defaulted if not provided and the input filename may 
					contain wildcards for doing batch operations. Specifying the output filename is 
					optional, and if more than one input file is specified (with wildcards) then the
					only acceptable output specification is an output path. If the input filename
					begins with a '@' then the file is assumed to contain a list of newline separated 
					filenames to be processed in batch mode (the output file may also be done this way 
					although it may contain only a single line). The Windows frontend uses this 
					mechanism for passing filenames, although it could be used for other purposes. Either 
					filename can be replaced with "-" to allow use of stdin or stdout, although because 
					the hybrid lossless mode uses two files it is not compatible with pipes.
				</p>
				<p>
     The input files for WavPack may be either Microsoft .wav files (as mentioned above), or
     any of the other formats listed below, including existing WavPack files.
     They may also be raw PCM files (specified with the --raw-pcm option).
     In the case of re&#235;ncoding from existing WavPack files, all tags are
     copied (and may be modified with the tag specification options) and unless an alternate output
     name or directory is specified, the source files will be overwritten (safely). Used with the
     filename wildcards, this can be used to easily re&#235;ncode an entire directory of WavPack
     files with a single command. Re&#235;ncoding from lossy to lossless is not allowed, but other
     combinations like pure lossless to hybrid lossless (or lossy) are fine. Note that when
     re&#235;ncoding from a configuration with a correction file to a configuration without a
     correction file, the source correction file will <i>not</i> be automatically deleted even if
     the source WavPack file is being overwritten. Adding the -d option will accomplish this,
     however.
				</p>
				<p>
					<b>File Formats:</b>
					<br/>
					<b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;Microsoft Waveform Audio, extension ".wav", PCM audio</tt></b>
					<br/>
					<b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;WavPack Compressed Audio, extension ".wv", PCM audio</tt></b>
					<br/>
     <b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;Apple AIFF Audio Format, "extension ".aif", PCM audio</tt></b>
					<br/>
     <b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;Apple Core Audio Format, "extension ".caf", PCM audio</tt></b>
					<br/>
     <b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;Sony Wave64 Audio Format, extension ".w64", PCM audio</tt></b>
					<br/>
     <b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;Philips DSDIFF Format, "extension ".dff", 1-bit DSD audio</tt></b>
					<br/>
     <b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;Sony DSD Stream Format, "extension ".dsf", 1-bit DSD audio</tt></b>
				</p>
				<p>
					Both programs will warn before overwriting any file (unless the -y or --no-overwrite switch is 
					specified) and file overwriting is implemented with temp files so that the overwritten
     file is not deleted until the operation is successful (including the verification pass
     if that has been requested with the -v switch). The programs will also (unless the -q
     switch is specified) display 
					progress in percentage complete and when finished with each file will give the 
					compression ratio (or bitrate for lossy files), whether the compression was lossy 
					or lossless, and the processing time. They will also (unless the -z option is used)
     display the progress percentage for the whole batch of files on the title bar which
     is useful when they are being run minimized. If a batch of files is run then both
     programs will indicate when complete whether any errors occurred and how many files
     were processed.
				</p>
				<p>
					The compressed file format (.wv) contains the filename extension, all the original header information, and the 
					compressed audio data for one audio file. The file can be used by WvUnpack to restore 
					the original file, or the .wv file can be played directly by a player that supports 
					WavPack files natively like Foobar2000 or JRiver Media Center, or a player that can play WavPack
     files using a plugin like Winamp, dBpowerAMP, or Apollo. Also, there are plugins
     available to read and write WavPack files with Adobe Audition (and CoolEdit) and Nero
     Burning ROM. Many DAW programs also support WavPack natively like Reaper, Steinberg WaveLab, and
     Audacity.
				</p>
    <p>
     Starting with version 5.5.0, the WavPack and WvUnpack programs allow options to be added to the
     executable (exe) filename and these options are parsed first every time the command is executed.
     This is useful when the programs are run by dropping files into their respective icons in the
     Windows File Explorer rather than using the Command Prompt to run them. For example, the WavPack
     executables might be renamed (see below for the exact meaning of the options listed):
    </p>
    <b><tt>&nbsp;&nbsp;&nbsp;wavpack{--pause}{--drop}{-hxvm}{--import-id3}.exe</tt></b>
					<br/>
    <b><tt>&nbsp;&nbsp;&nbsp;wvunpack{--pause}{--drop}.exe</tt></b>
    <p>
     When multiple input files are specified for the WavPack and WvUnpack programs (using wildcards)
     and the output is specified as stdout ("-") then the resulting output from all the files
     processed is sent to stdout in a continuous concatenated stream. Normally this is not very
     useful because data is interleaved with file headers, however one case can be. If the --raw or
     --raw-pcm option is used in WvUnpack then just the audio from all the files will be send to
     stdout, and this can be captured in a file or piped directly back to WavPack for re&#235;ncoding:
    </p>
    <b><tt>&nbsp;&nbsp;&nbsp;wvunpack --raw *.wv - > merged.raw</tt></b>
					<br/>
    <b><tt>&nbsp;&nbsp;&nbsp;wvunpack --raw *.wv - | wavpack --raw-pcm - merged.wv</tt></b>
				<p>
     Starting with version 5.8.0, WavPack, WvUnpack, and WvGain will all automatically use multiple
     threads to speed up their operations if they determine that they are running on a multicore
     machine. If a different behavior is desired, use --no-threads to prevent multithreading or
     --threads=&lt;n&gt; to fine-tune the number of threads utilized (from 1 to 12). Note that
     multithreaded <i>encoding</i> is not supported for the hybrid mode (except for multichannel
     files), but all modes of multithreaded <i>decoding</i> work fine.
				</p>
				<p>
				 Finally, a debugging mode is provided for the programs which is enabled by simply appending
				 "_debug" to the the name of the executable (for example, wavpack.exe becomes
     wavpack_debug.exe). The debugging versions work exactly like the normal versions except
				 that they report more detailed information about their operation. For instance, they
     will list their command-line arguments and wavpack.exe will display detailed
				 information about the files it compresses (including how the audio data is interpreted).
				 Also, all information displayed on the console (including any errors encountered) is
				 dumped to a text file called "wavpack.log" (under Application Data). This is very handy
				 for debugging in situations where the command-line window terminates before an error
				 message can be read (or doesn't get displayed at all). It is not recommended that these
				 debug version be used all the time because the log file will grow indefinitely.
				</p>
				</div>
			</div>

			<br/>
			<div class="box" id="linux">
				<div class="boxHeader">Linux</div>
				<div class="boxContent">
					<p>This documentation was written specifically for the Windows command-line version of
      WavPack, however the version for Linux (and MacOS) is essentially the same. There are
      some Windows-specific commands missing from the Linux version like the -l switch to run
      at low priority or the --pause and --drop options helpful with drag and drop. The "debug"
      mode described above does not create a log file, but it <i>does</i> provide the extra information.
     </p>
     <p>
      The most significant change is in the specification of the output filename or directory,
      which in the Linux case must be preceded with the -o switch. This was done to allow
      multiple input files to be specified, which is handy by itself but is especially useful
      in the default case where the shell is performing wildcard expansion.
					</p>
				</div>
			</div>
			
			<br/>
			<div class="box" id="wavpack">
				<div class="boxHeader">WavPack Options</div>
				<div class="boxContent">

				<p>
					<b><tt>-a = Adobe Audition (CoolEdit) mode for 32-bit floats</tt></b>
				</p>
				<p>
					The WAVEFORMATEXTENSIBLE structure is used (if present) to determine the format 
					details. However, there are some programs that use their own non-standard format 
					extensions. The most popular of these is Adobe's Audition (previously Syntrillium's 
					CoolEdit) which created two new 32-bit floating point formats. An option has been 
					added to WavPack (-a) to force the "adobe" interpretation of these floating point 
					formats. If you are compressing integer files do NOT use this option.
				</p>

				<p>
					<b><tt>--allow-huge-tags = allow tag data up to 16 MB (otherwise it's 1 MB)</tt></b>
				</p>
				<p>
     Normally WavPack allows the APEv2 tags to contain up to 1 MB of data. This limit was
     implemented to allow for their use on portable devices which may have limited memory or
     processing resources. However, in some situations it may be desirable to place more
     data in the tags (for high resolution cover art scans, for example) and this option
     permits that. Note that these files are not fully WavPack compliant and may not work
     in all situations or with older versions of WavPack programs and plugins.
				</p>

				<p>
					<b><tt>-b&lt;n&gt; = enable hybrid compression, n = 2.0 to 23.9 bits/sample, or</tt></b>
					<br/>
					<b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
					&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
					&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;n = 24-9600 kbits/second (kbps)</tt></b>
				</p>
				<p>
					The default operation of WavPack is pure lossless, which means that the .wv file 
					contains all the information that was in the original .wav file. The hybrid mode 
					allows the user to specify a target bitrate for the output file, either in kilobits 
					per second (kbps) or bits per sample. If the track can be losslessly compressed without 
					exceeding the specified bitrate, then it will be and WavPack will report the compression
					as lossless. If lossless compression would exceed the specified bitrate, then WavPack 
					will begin carefully discarding the least significant portion of the audio information 
					to stay within the limit. Every effort is made to keep this inaudible, including the 
					use of joint stereo, dynamic bit allocation and noise shaping. WavPack will report this 
					as "lossy" compression. Although the option accepts bitrates as low as 24 kbps, the 
					actual value that WavPack can achieve is usually much higher than that. For example,
					with CD-audio sampled at 44.1k the lower limit is about 196 kbps.
				</p>
				<p>
     When this option is used in combination with the <b>-c</b> option, then a "correction"
     file is also created and the result is hybrid lossless compression. Without the <b>-c</b>
     option the operation is lossy.
				</p>
				<p>
					The hybrid mode can be used quite successfully with floating-point audio, however
					it should <b>not</b> be used for scientific type floating-point
					data because the hybrid algorithm might not be application appropriate (and
					floating-point "exception" values like infinities or NaNs will not be properly encoded).
					Use only the pure lossless mode with non-audio floating-point data.
				</p>
				<p>
					The hybrid mode is <b>not</b> usable with DSD audio files; those are always lossless
     and attempting hybrid compression with them will generate an error.
				</p>

				<p>
     <b><tt>--blocksize=&lt;n&gt; = specify block size in samples (n = 128 - 131072)</tt></b>
				</p>
				<p>
     WavPack normally determines the optimum number of samples to place into each
     WavPack block, however this option allows it to be directly specified. The most likely
     use for this option would be to improve the handling of audio files that have a
     variable number of redundant MSBs. For example, the output of the lossyWAV program or
     the output of a software HDCD decoder. This option is also often combined with the
     <b>--merge-blocks</b> option (which also reduces the minimum allowed block size to 16 samples).
				</p>

				<p>
					<b><tt>-c = create correction file (.wvc) for hybrid mode (=lossless)</tt></b>
				</p>
				<p>
					If the -c option is specified (in addition to the -b option), then WavPack will 
					generate an additional file that will contain the information that was discarded 
					to generate the lossy .wv file. This file will have the same name as the .wv file 
					but will have the extension .wvc (the 'c' is for "correction"). When WvUnpack is 
					used to restore the .wav file, it will attempt to find the .wvc file in the same 
					directory as the .wv file. If it is found then the decompression will be lossless
					and will be reported as such, otherwise lossy decompression will be reported 
					(assuming that any information was actually discarded during the original compression).
					If -c is specified but no actual information is discarded, the correction file will 
					be deleted. The extra overhead involved with having these two files instead of a 
					single pure lossless file is usually less than 1% of the original .WAV file and can 
					be as low as 0.25% at high bitrates. Note that CRCs are stored for both the lossy 
					and lossless versions, so error detection works correctly whether the .wvc file is 
					used or not.
				</p>

				<p>
					<b><tt>-c&lt;n&gt; = enable hybrid lossless compression, n = 2.0 to 23.9 bits/sample, or</tt></b>
					<br/>
					<b><tt>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
					&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
					&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
     &nbsp;&nbsp;&nbsp;n = 24-9600 kbits/second (kbps)</tt></b>
				</p>
    <p>
     This is a new shortcut combining the <b>-b&lt;n&gt;</b> option to set hybrid bitrate and the
     <b>-c</b> option to create a "correction" file for hybrid lossless operation. The
     advantage of this new syntax is that it makes it harder to accidentally forget the
     <b>-c</b> option and get lossy operation unintentionally.
    </p>

				<p>
				<p>
					<b><tt>-cc = maximum hybrid compression (hurts lossy quality &amp; decode speed)</tt></b>
				</p>
				<p>
					Normally, when the -c option is used to create a correction file in the hybrid mode, 
					WavPack attempts to optimize for the quality of the lossy file and lets the combined 
					lossless compression of the two files fall where it may. This option tells WavPack 
					to optimize for the overall compression ratio instead, even if this means some 
					possible degradation of lossy quality (for example, dynamic noise shaping is not
     used in this mode). This can also have an effect on lossless decompression speed
     (however it does <b>not</b> affect lossy decoding). Keep in mind the effect of
     this option is not too significant either way.
				</p>

				<p>
     <b><tt>--channel-order=&lt;list&gt; = specify non-standard channel ordering</tt></b>
				</p>
				<p>
     For multichannel audio WAV files there is a Microsoft required WAVEFORMATEXTENSIBLE header to
     indicate which speakers are represented and those speakers must be in the standard Microsoft
     order (which is also specified by USB). However, some programs skip generating the WAVEFORMATEXTENSIBLE
     header and even write the channels in the wrong order. For files without the WAVEFORMATEXTENSIBLE
     header (or those with the header but with a zeroed channel mask), WavPack assumes Microsoft channel
     order and further assumes that all speakers (up to the
     number of channels) are present. If that is not the case, then this option allows the user to
     specify the exact channels present (and their order) from this list (which is in Microsoft order):
     FL,FR,FC,LFE,BL,BR,FLC,FRC,BC,SL,SR,TC,TFL,TFC,TFR,TBL,TBC,TBR.
    </p>
				<p>
     If not all of the channels in the file have speaker definitions, then it is possible to simply
     terminate the speaker list with "..." (e.g. FL,FR,...) to indicate that all following channels are
     unassigned. Specifying "..." alone indicates that all channels are unassigned.
    </p>
    <p>
     This option may just be used to specify <i>which</i> channels are present, even if they
     are in standard Microsoft order (e.g. standard quad is FL,FR,BL,BR). In this case the
     option just controls the channel "mask" and performs no reordering. But if the option
     does result in reordering, then this is done <i>before</i> encoding, so if the
     resulting WavPack file is subsequently unpacked it will <b>not</b> recreate the
     identical WAV file (because the channels will now be in the correct order). If you want
     the unpacked file to be identical to the original simply do not use this option (although
     this means that the resulting WavPack file will not have the correct channel information
     and so won't play correctly in multichannel software, and also means that the compression
     might not be as good because the "wrong" channels will be paired in stereo).
    </p>
    <p>
     If this option is used then it might be a good idea to add the <b>-r</b> option also to generate a
     valid WAVEFORMATEXTENSIBLE header with the correct channel information. In fact, one could even
     pipe the output of WavPack directly to WvUnpack for the sole purpose of reordering WAV files
     and adding the correct WAV headers!
    </p>

				<p>
					<b><tt>--cross-decorr = use cross-channel correlation in hybrid mode</tt></b>
				</p>
				<p>
					Cross-channel correlation is exploited in lossless mode, but it not used by default in hybrid mode
     because it can increase noise slightly and increases CPU requirements during hybrid lossless
     decoding. This switch is provided to force this mode without affecting noise shaping
     (which <b>-cc</b> does to maximize hybrid lossless compression).
				</p>

				<p>
					<b><tt>-d = delete source file if successful (use with caution!)</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

				<p>
					<b><tt>--drop = accept multiple files, no output specification allowed</tt></b>
				</p>
				<p>
					This Windows-only option should be specified before any source files and alters the syntax to allow multiple
     input files instead of just one. This is useful for using the WavPack executable icon as the target
     of a "drag and drop" operation in which case the option is added to the executable filename using
     braces (e.g., <b>wavpack{--drop}.exe</b>). Because this is mostly intended for drap and drop support,
     the specification of output names or folders is not supported.
				</p>

				<p>
					<b><tt>-f = fast mode (fast, but some compromise in compression ratio)</tt></b>
				</p>
				<p>
					The "fast" mode should be used when compression (or decompression) speed is more 
					important than compression ratio (or, in lossy mode, audio quality). This option
     has no effect on DSD audio.
				</p>

				<p>
					<b><tt>--force-even-byte-depth = round the bit-depth of source PCM files up to the next whole byte</tt></b>
				</p>
				<p>
     File formats that store uncompressed PCM allow the specification of bit depths
     that are not multiples of 8 (e.g., 12-bit or 20-bit). WavPack accepts these files,
     however it also checks to make sure that the appropriate LSB "padding" bits are zero
     in the source file (as they should be), and if not reports this as an error. One
     method of getting around this rare situation is to specify this option which will
     round up the bit depth to the next full byte and thereby save the non-zero padding
     bits. These files will restore losslessly (assuming a lossless mode is used) even
     though they are technically invalid. Note that the WavPack files will report as the
     rounded bit depth and be otherwise fine, and this does not result in worse
     compression. The other method is to use the <b><tt>--pre-quantize</tt></b> option
     to zero the padding bits before encoding, but that's only recommended if it's
     known that the padding bits are not valid audio.
				</p>

				<p>
					<b><tt>-g = general/normal mode (cancels any previously specified -f or -h options)</tt></b>
				</p>
				<p>
     This option is provided to revert to the normal default speed mode after having
     specified one of the other modes (i.e., "fast" or "high"). It really only makes sense
     if one of the other modes has been encoded directly into the executable filename using
     braces, but it could be specified to be unambiguous.
				</p>

				<p>
					<b><tt>-h = high quality (better compression in all modes, but slower)</tt></b>
				</p>
				<p>
					The "high" mode should be used when compression ratio (or, in lossy mode, audio 
					quality) is more important than compression (or decompression) speed. This option 
					slows both by about a factor of about 1.5.
				</p>

				<p>
					<b><tt>-hh = very high quality (best compression in all modes, but slowest)</tt></b>
				</p>
				<p>
					The "very high" mode should be used when compression ratio (or, in lossy mode, audio 
					quality) is much more important than compression (or decompression) speed. This option 
					slows both by about a factor of about 2, and is not recommended for use on
     vintage portable devices because of the high CPU load required for decoding, but should
     be fine for anything modern. This option simply activates the <b>high</b> mode for DSD
     audio files.
				</p>

				<p>
					<b><tt>--help = extended help display</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

				<p>
					<b><tt>-i = ignore length in wav header (no pipe output allowed)</tt></b>
				</p>
				<p>
					Some programs that pipe data to encoders do not always give the correct length in
					the wav headers that they provide (foobar's client and CDex are examples). In these
					cases use this option to force WavPack to ignore the header and accept the actual
					length. Because WavPack must seek to the beginning of the output file to write the correct
					length, this option cannot be used with piped output (override with <b>-y</b>). Note that
     WavPack will attempt to fix the supplied header with the correct length if it can do
     so safely (and of course in this case it's not strictly lossless, but we're just fixing
     an invalid header). <b>Only use this option when it's really needed; using with a valid
     header will cause any trailing metadata to be interpreted as audio.</b>
				</p>
    <p>
     When using <b>-i</b> it is not allowed to use pipes for both the input and output because the actual
     audio length will not be known in the beginning and there's no way to rewind the output file when
     we're done to update the length stored there. You can use the <b>-y</b> to override this but you will
     end up with sub-standard files that must be seeked to the end on decode to determine their length.
     A warning to this effect will be displayed when the operation completes.
    <p>

				<p>
					<b><tt>--import-id3 = import applicable tag items from ID3v2.3 or ID3v2.4 tag on DSF and other files</tt></b>
				</p>
				<p>
					Sony's DSF file format specifies that these files may contain an ID3v2 tag at the end, and it
     is not uncommon for other files to contain these tags even though it's not part of their official
     specs (e.g., WAV and DSDIFF files). WavPack considers this a trailing "wrapper" and stores it in the
     WavPack file as such so that the original file can be restored verbatim. However, stored this way
     it is not easily accessible for reading (and it is certainly not writable) because WavPack uses
     APEv2 tags for metadata. This option causes any trailing ID3v2.3 or ID3v2.4 tag in the DSF file (or
     any other file type) to be scanned and all applicable items imported into the APEv2 tag,
     including cover art.
    </p>
				<p>
     Note that if over 1 MB of image data is present, then the <b>--allow-huge-tags</b> option must
     be specified. Also, keep in mind that the image data will be duplicated in the APEv2 tag
     and will therefore be consuming twice as much space as required, so it might make sense to also
     specify the <b>-r</b> option if very large images are present, although this will obviously make
     it impossible to restore the original DSF file because the ID3v2 tag will be gone, and of course
     there might be fields in the original tag that are not copied to the APEv2 tag.
				</p>

				<p>
					<b><tt>-j&lt;n&gt; = joint-stereo override (0 = left/right, 1 = mid/side)</tt></b>
				</p>
				<p>
					WavPack normally defaults to joint stereo (sometimes called mid/side) in which 
					the left and right channels are combined to form an alternate representation 
					(essentially L+R and L-R) that compresses better in lossless mode and improves 
					quality in lossy mode. In the "extra" modes WavPack will choose whether or not to
     use joint-stereo on a block by block basis. This option allows this feature to be
     forced either always on or always off.
				</p>

				<p>
					<b><tt>-l = run at low priority (for smoother multitasking)</tt></b>
				</p>
				<p>
					This option can be used (in Windows only) to force WavPack to run at a low priority 
     and is handy for doing large WavPack batch conversions in the background.
				</p>

				<p>
					<b><tt>-m = compute &amp; store MD5 signature of raw audio data</tt></b>
				</p>
				<p>
					Calculate and display the MD5 checksum of the uncompressed audio data and store 
					it in the compressed file. These sums are commonly used in file trading communities 
					to compare versions of tracks, and as such the sums generated by WavPack match those 
					of FLAC, OptimFROG, Shntool, and get_id3(). They can also be used by WvUnpack during
					decompression to verify the data integrity of lossless files, but this is a secondary
     function because WavPack already verifies each block as they are decoded.
				</p>

				<p>
					<b><tt>--merge-blocks = merge consecutive blocks of equal redundancy</tt></b>
				</p>
				<p>
     This option is only valid when generating lossless files and is only used with the
     <b>--blocksize</b> option. WavPack will always scan each block of audio data before
     compressing to determine the actual number of valid bits that need to be stored because,
     for various reasons, this is not always the same as the number of bits present. In some
     situations there can be many redundant bits, as for example in the output of the
     lossyWAV program or the output of software HDCD decoders. In these cases it may be
     advantageous to use a specific block size (or simply a small block size), but WavPack
     does not work very well with very small block sizes because of its relatively large
     block header.
    </p>
    <p>
     The <b>--merge-blocks</b> option takes care of this situation by allowing WavPack to
     merge consecutive blocks with the same number of bits to remove. For example, if a
     blocksize of 512 is specified, and 10 512-sample blocks in a row all have exactly
     the same number of redundant bits, then these 10 blocks will be combined in to a
     single block of 5120 samples (which is much more efficient for WavPack). When the
     <b>--merge-blocks</b> option is specified the minimum block size that <b>--blocksize</b>
     will accept is reduced from 128 to 16 samples.
    </p>

				<p>
					<b><tt>-n = calculate average and peak quantization noise (hybrid only)</tt></b>
				</p>
				<p>
					This causes WavPack to calculate the average and peak quantization noise generated 
					in the lossy version of the hybrid mode, both referenced in decibels below full 
					scale. While it is impossible to use this as a guide to determine the audibility 
					of the noise, it is useful for comparing the various compression options and for 
					comparing WavPack's lossy performance to other programs. Note that this option does
					not currently produce meaningful results for floating point or multichannel files.
				</p>

    <p>
     <b><tt>--no-overwrite = don't overwrite existing files, and don't ask to</tt></b>
    </p>
    <p>
     Normally WavPack checks that a file it is about to create already exists, and if it
     does will prompt the user before overwriting it. When this option is specified
     WavPack will skip this prompt and simply go to the next file (if there are any).
     This might be useful to resume a long, multi-file WavPack operation that was cancelled
     because in that case all the previously encoded files are quickly skipped.
    </p>

    <p>
     <b><tt>--no-threads = do not use multiple threads for faster operation</tt></b>
    </p>
    <p>
     Normally, if WavPack can determine that it is running on a multicore system, it will
     use multiple threads to speed up the operation, sometimes significantly. Use this
     option to force single-threaded operation, for example in cases where multithreading
     is being done elsewhere. This is equivalent to specifing <b>--threads=1</b>.
    </p>

    <p>
     <b><tt>--no-utf8-convert = don't recode passed tags to UTF-8, assume they are UTF-8 already</tt></b>
    </p>
    <p>
     The text fields of APEv2 tags are encoded in the UTF-8 variant of Unicode, so when
     tag information is passed in on the command-line they are converted to UTF-8 before
     being stored. If your system is already passing the strings in UTF-8, use this option
     to prevent double conversion.
    </p>

    <p>
     <b><tt>--optimize-int32 = enable new optimizations for certain 32-bit integer files</tt></b>
    </p>
    <p>
     32-bit integer files are often originally sourced from floating-point audio, and
     previously these were not optimally compressed by WavPack. This option enables a
     new check for these (and similar) files and compresses them up to 10% more.
     Unfortunately this new mode is not 100% backward compatible, so files created
     with this option will decode with only 24 bits of resolution with older WavPack
     decoders or 3rd party decoders like FFmpeg. Of course, this difference is very
     likely inaudible, and in fact will be non-existent with common 24-bit DACs.
    </p>

				<p>
					<b><tt>--pair-unassigned-chans = encode unassigned channels into stereo pairs</tt></b>
				</p>
				<p>
     WavPack encodes all channels of multichannel files into stereo or mono streams. So, if
     two channels are related (like FL and FR) then these are encoded into one stream
     (which improves compression performance) whereas unrelated channels (like LFE) are
     encoded by themselves into mono streams. Unassigned channels (those not associated
     with any defined speaker) are normally encoded in mono streams for safety, however
     this option can be used to specify that they should be paired into stereo streams.
     This might improve compression in cases where all of the channels were very similar,
     or when it is known that the channels are, in fact, stereo pairs.
				</p>

				<p>
					<b><tt>--pause</tt></b>
				</p>
				<p>
     Pause before exiting the console program (Windows only), allowing the user to press
     any key to continue. This might be useful to include in situations where the console
     window disappears before the completion status can be seen (like with EAC or the
     WavPack FrontEnd) or when using the program icon as "drag and drop".
				</p>

				<p>
					<b><tt>--pre-quantize=bits = apply pre-quantization to specified bitdepth</tt></b>
				</p>
				<p>
     Even the finest 24-bits ADCs have a true resolution of only around 20 or 21 bits, and
     those lower "noise" bits are completely uncompressible. This option enables truncation
     of the source audio samples to a specified bit depth <b>before</b> encoding and MD5
     calculation, and can often greatly improve the lossless compression ratio. Keep in mind
     that although this is conceptually a lossy operation, it will be reported as lossless
     because this is performed by the command-line program itself just as the audio is read
     from the source file (but of course the source file is not actually modified).
    </p>
    <p>
     This option can be used with any source format (float data is rounded instead of
     truncated, and <i>not</i> clipped) and any WavPack mode, although it really does not
     make sense to use this in combination with WavPack's lossy mode because it wouldn't
     do much (at least not with reasonable parameters).
     In the hybrid lossless mode, this pre-quantizing <i>would</i> reduce the size of the
     correction file, but wouldn't have much effect on the lossy portion. It also works
     well as a re&#235;ncoding operation (with a WavPack source file), although it <i>will</i>
     modify the MD5 checksum stored in the file (which WavPack's lossy mode does not do).
				</p>
				<p>
					The pre-quantize option is <b>not</b> usable with DSD audio files because they are
     already only 1 bit deep, but an error will not be generated (the option is ignored).
				</p>

				<p>
					<b><tt>-q = quiet (keep console output to a minimum)</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>
				<p>
					<b><tt>-r = remove any extra chunk info from audio file</tt></b>
				</p>
				<p>
				 WavPack normally saves all the header information contained in the audio file
				 (including any chunks after the audio data). This is done so that WvUnpack.exe can
				 restore the original audio file <i>exactly</i>. The -r option causes WavPack to discard the
				 header contained in the source file (and any extra chunks). Obviously the source
     header is still used to determine the format and size of the file (i.e. this is not
     a "raw" mode). When the file is restored with WvUnpack, a generic header will be
     generated appropriate for the format. Note that if this option is used with
     certain floating-point WAV files generated by CoolEdit or Audition that are not
					normalized, the <b>--normalize-floats</b> option of WvUnpack will have to be used when
					converting them back to standard PCM formats.
				</p>

				<p>
					<b><tt>--raw-pcm = input data is raw pcm (44100 Hz, 16-bit signed, 2-ch, little-endian)</tt></b>
				</p>
				<p>
     Specifies that the source file (or stdin) contains raw PCM data with no header.
     and conforms to standard CD quality. Note that the data must match the format of WAV file
     audio (i.e., little endian and signed). 
				</p>
    <p>
     When encoding raw PCM, it is not allowed to use pipes for both the input and output because the actual
     audio length will not be known in the beginning and there's no way to rewind the output file when we're
     done to update the length stored there. You can use the <b>-y</b> to override this but you will end
     up with sub-standard files that must be seeked to the end on decode to determine their length.
     A warning to this effect will be displayed when the operation completes.
    <p>
     Note that WavPack files created in this way will still normally decode to WAV files (with headers)
     when unpacked. To avoid this, use the <b>-r</b>, <b>--raw</b>, or <b>--raw-pcm</b> option with the
     WvUnpack program to force raw unpacking.
				</p>

				<p>
					<b><tt>--raw-pcm=sr,bps[f|s|u],ch,[le|be] = input data is raw pcm with specified format</tt></b>
				</p>
				<p>
     Specifies that the source file (or stdin) contains raw PCM data with no header
     and the specified sampling rate, bit-depth, and number of channels. Parameters
     matching the default (44100,16,2,le) can be omitted. The valid range of bit-depths
     is from 1 to 32. The default data format can be overridden with "f" for floating-point
     audio (32-bit only) or "s" or "u" for signed/unsigned integers. The number of channels
     may be from 1 to 4096.
    </p>
    <p>
     To compress DSD audio, specify bps=1. Note that DSD audio must be in Philips DSDIFF
     format (i.e., channels interleaved by byte and MSB first temporally). Therefore, standard
     SACD stereo audio would be <b>--raw-pcm=2822400,1,2</b>.
    </p>
				<p>
     Note that WavPack files created in this way will still normally decode to WAV files (with
     headers) when unpacked (or DFF files for DSD audio). To avoid this, use the <b>-r</b> option
     with the WvUnpack program to force raw unpacking.
				</p>

				<p>
					<b><tt>-s&lt;n&gt; = noise shaping override (hybrid only, n = -1.0 to 1.0, 0 = off)</tt></b>
				</p>
				<p>
					WavPack uses first-order noise shaping to improve the perceived quality of lossy 
					files and to improve the hybrid lossless compression ratio. Normally WavPack will 
					choose the noise shaping most appropriate for the application (based on the source
     sampling rate and whether the -cc option is specified), but this option allows the
     user to override the default with a fixed value. The parameter range is +/- 1.0, where 
					positive values shift the noise higher in frequency and negative values shift the 
					noise lower in frequency. Values very close to -1.0 are clipped to prevent problems
					associated with very high gain near DC and the value zero results in no noise shaping
     at all (i.e., white noise). This option should not be used with the <b>--use-dns</b>
     option which enables dynamic noise shaping.
				</p>

				<p>
					<b><tt>-t = copy input file's time stamp to output file(s)</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

				<p>
     <b><tt>--threads[=n] = enable multithreaded operation (optional n = 1 for no threading to 12 threads maximum)</tt></b>
				</p>
				<p>
     On modern multicore CPUs WavPack can get a significant performance boost by utilizing
     multiple threads. This is fully applicable to the pure lossless mode, but in the hybrid
     modes it is only applicable to multichannel files. If the parameter <b>n</b> is omitted
     then the number of threads is determined automatically as a balance between operational
     speed and energy consumption. Using multiple threads is now the default operation on
     multicore CPUs, so this option is really only needed to fine-tune the number of threads.
				</p>

				<p>
					<b><tt>--use-dns = force use of dynamic noise shaping (hybrid only)</tt></b>
				</p>
				<p>
     For version 4.50, dynamic noise shaping was added to the WavPack hybrid mode. This feature
     causes the noise shaping to continuously adjust to the spectral characteristics of the source
     signal and results in significantly improved subjective quality on samples that previously caused
     difficulty with WavPack's lossy mode (particularly material with loud high-frequency transients).
     This feature is used by default except for high sampling rates (>= 64 kHz) or when the -cc option
     is specified (because it can adversely effect compression ratio), however the <b>--use-dns</b> option
     allows the user to force the use of dynamic noise shaping even when it would not normally be used.
     This option should not be specified with the <b>-s</b> option which overrides the default with fixed
     noise shaping.
				</p>

				<p>
					<b><tt>-v = verify output file integrity after write</tt></b>
				</p>
				<p>
					Causes WavPack to perform a separate verification pass over the output file to guarantee that the audio data
     is correctly encoded and stored to disk. For lossless compression this is performed with an MD5 hash even
     if the <b>-m</b> option is not specified. If an error occurs (indicating either a bug in the program or
     faulty hardware) then a message is displayed and the output file is deleted (and if an existing file is
     being overwritten, then it will be untouched). Because the output file is actually rewound and reread, this
     option cannot be used when writing to pipes.
				</p>

				<p>
					<b><tt>--version = display program version to stdout</tt></b>
				</p>
				<p>
					Both the version of the command-line program and the WavPack library are displayed.
				</p>

				<p>
					<b><tt>-w Encoder = write actual encoder metadata to APEv2 tag</tt></b>
				</p>
				<p>
     Write a metadata item to the APEv2 tag for the actual encoder being used
     (e.g., "Encoder=WavPack 5.9.0"). It's also possible to specify a value to override
     the default. Note that when re&#235;ncoding WavPack files this metadata item is
     <b>not</b> simply copied from source to destination, but is recreated with the
     appropriate value.
				</p>

				<p>
					<b><tt>-w Settings = write actual encoder settings metadata to APEv2 tag</tt></b>
				</p>
				<p>
     Write a metadata item to the APEv2 tag for the actual encoder settings being used
     (e.g., "Settings=-hb384cx3"). It's also possible to specify a value to override
     the default. Note that when re&#235;ncoding WavPack files this metadata item is
     <b>not</b> simply copied from source to destination, but is recreated with the
     appropriate value.
				</p>

				<p>
					<b><tt>-w "Field=[@]Value" = write specified metadata to APEv2 tag</tt></b>
				</p>
				<p>
					Write specified information to APEv2 tag appended to WavPack file(s). May be used
					multiple times for multiple fields. APEv2 tags are the preferred tag format for
					WavPack files and are read by all the standard WavPack playback plugins. If the
					specified value begins with a '@', then the value is assumed to be a
					filename which is used to obtain the item's actual value. This is handy for
					including the CUESHEET field for use with images files + cuesheets and foobar2000.
				 The filename may contain wildcards if it matches exactly one file. Also, if the
				 file cannot be found in the current directory then the source and destination
				 directories (if specified) are also checked.
				</p>

				<p>
					<b><tt>--write-binary-tag "Field=@file.ext" = write specified binary metadata to APEv2 tag</tt></b>
				</p>
				<p>
					Write specified binary file to APEv2 tag appended to WavPack file(s). This is most
     commonly used to embed album cover art into WavPack files (with the field name
     <b>"Cover Art (Front)"</b>), but could be used for anything desired. A file must be
     specified (the data cannot come from the command-line) and it may contain wildcards
     if it matches exactly one file. Also, if the file cannot be found in the current directory
     then the source and destination directories (if specified) are also checked.
				</p>

				<p>
					<b><tt>-x[n] = extra encode processing (optional n = 0-6, 1 = default)</tt></b>
				</p>
				<p>
					Like pre-4.0 versions of WavPack (and many other compressors), WavPack 5.9.0 normally
					works "symmetrically" in that encoding and decoding operate at about the same rate
					(regardless of the mode used). However, WavPack has an option to work
					"asymmetrically", so that extra processing can be done during encoding to
					provide better compression, but with NO corresponding cost to decoding performance!
				</p>
				<p>
					This is enabled with the -x option and provides an average improvement in CD music 
					compression of about 1% in "fast" mode, about 0.5% in the normal mode, and still less in 
					the higher modes. Because the standard compression parameters are optimized for 
					"normal" CD music audio, this option works best with "non-standard" audio 
					(synthesized sounds, non-standard sampling rates, etc.) where it can often achieve 
					enormous gains. The default level (n=1) provides a decent improvement with little
     cost in encoding speed and is recommended for all but the most time critical encoding.
     Higher levels provide some marginal improvement with an increasing cost of encoding
     speed. The highest levels (n = 4-6) are extremely slow but can provide significant
     improvement in special situations (i.e. synthesized sounds). Use -x0 to specify no
     extra processing.
				</p>
    <p>
     This option is not applicable to DSD audio and is simply ignored.
    </p>

				<p>
					<b><tt>-y = yes to all warnings (use with caution!)</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

				<p>
					<b><tt>-z[n] = don't (n=0 or omit) or do (n=1) write to console title bar to show overall progress</tt></b>
				</p>
				<p>
					If enabled leaves "WavPack Completed" on the title bar.
				</p>
				</div>
			</div>

			<br/>
			<div class="box" id="wvunpack">
				<div class="boxHeader">WvUnpack Options</div>
				<div class="boxContent">
				<p>
     <b><tt>--aif or --aif-le = force output to Apple AIFF Audio format</tt></b>
				</p>
				<p>
     Output an Apple AIFF Audio file, either legacy AIFF or AIFF-C/sowt (little-endian),
     regardless of the input file format. All extra information in the original file's
     header and trailer will be lost and a "fresh" AIFF Audio header will be generated.
     Note that DSD audio files will be decimated 8X and output as 24-bit PCM.
				</p>
				<p>
     <b><tt>-b  = blindly decode all stream blocks &amp; ignore length info</tt></b>
				</p>
				<p>
				 This option uses the "streaming" mode of the WavPack decoder. Normally this
     doesn't make any difference except that the percentage complete
				 display does not work. However, it causes the decoder to ignore all information
				 contained in the WavPack blocks concerned with duration and location and blindly
				 unpacks every valid WavPack block it encounters, and will do so indefinitely
				 until it receives an EOF. This may be of some use to recover corrupted or partial
     WavPack files, or it could be used to decode a sequence of WavPack files that had been
				 concatenated. If this is used to write a .wav file to stdout then note that the
     resulting .wav file will not indicate the correct length because it can't back up
     to fix it once it knows the actual length.
				</p>
				<p>
     <b><tt>-c  = extract cuesheet only to stdout (no audio decode)</tt></b>
				</p>
				<p>
					If the specified WavPack file contains an APEv2 tag, and that tag contains a CUESHEET
     field, then only dump that text to stdout without decoding any audio. If no
     cuesheet is found then an error is generated. This is equivalent to "-x cuesheet".
				</p>
				<p>
     <b><tt>-cc = extract cuesheet file (.cue) in addition to audio file</tt></b>
				</p>
				<p>
					If the specified WavPack file contains an APEv2 tag, and that tag contains a CUESHEET
     field, then that text will be extracted into an appropriately named .cue file
     located in the same folder as the .wav file. Note that the cuesheet
     is not automatically modified to point to the correct .wav file; this must be
     handled either when the cuesheet is embedded or (if the .wv file is renamed) by
     the user after extraction. This option has no effect if there's no cuesheet and
     is equivalent to "-xx cuesheet=%a.cue".
				</p>

				<p>
     <b><tt>--caf-be or --caf-le = force output to Apple Core Audio format</tt></b>
				</p>
				<p>
     Output an Apple Core Audio file (either big- or little-endian) regardless of the input
     file format. All extra information in the original file's header and trailer will be
     lost and a "fresh" Core Audio header will be generated. Note that DSD audio files will
     be decimated 8X and output as 24-bit PCM.
				</p>

				<p>
					<b><tt>-d = delete source file if successful (use with caution!)</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

				<p>
     <b><tt>--dff or --dsdiff = force output to Philips DSDIFF format</tt></b>
				</p>
				<p>
     Output a Philips DSDIFF file (extension .dff) audio file regardless of the input
     file format. All extra information in the original file's header and trailer will be
     lost and a "fresh" DSDIFF header will be generated. Note that only DSD audio files can
     be exported as DSDIFF files (WvUnpack does not contain a DSD encoder).
				</p>

				<p>
					<b><tt>--drop = accept multiple files, no output specification allowed</tt></b>
				</p>
				<p>
					This option should be specified before any source files and alters the syntax to allow multiple
     input files instead of just one. This is useful for using the WvUnpack executable icon as the target
     of a "drag and drop" operation in which case the option is added to the executable filename using
     braces (e.g., <b>wvunpack{--drop}.exe</b>). Because this is mostly intended for drap and drop support,
     the specification of output names or folders is not supported.
				</p>

				<p>
     <b><tt>--dsf = force output to Sony DSD format</tt></b>
				</p>
				<p>
     Output a Sony DSD Stream File (extension .dsf) regardless of the input
     file format. All extra information in the original file's header and trailer will be
     lost and a "fresh" DSF header will be generated. Note that only DSD audio files can
     be exported as DSF files (WvUnpack does not contain a DSD encoder).
				</p>

				<p>
					<b><tt>-f[n] = dump summary file information only to stdout in machine-parsable format (no decode)</tt></b>
				</p>
				<p>
					This command would normally be used by another application to obtain information about
     WavPack files. It outputs a single, semicolon delimited line to stdout for each input file specified. Here is
     a sample output line and a description of the 10 items of information provided:
				</p>
				<p>
     &nbsp;&nbsp;44100;16;int;2;0x3;9878400;023066a6345773674c0755ee6be54d87;4;0x18a2;Track01.wv<br/><br/>
     &nbsp;&nbsp;1. sampling rate<br/>
     &nbsp;&nbsp;2. bit-depth (1-32)<br/>
     &nbsp;&nbsp;3. format ("int" or "float")<br/>
     &nbsp;&nbsp;4. number of channels<br/>
     &nbsp;&nbsp;5. channel mask (in hex because it's a mask, always prefixed with "0x")<br/>
     &nbsp;&nbsp;6. number of samples (missing if unknown)<br/>
     &nbsp;&nbsp;7. md5sum (technically is hex, but not prefixed with "0x", might be missing)<br/>
     &nbsp;&nbsp;8. encoder version (this will be 4 or 5 as older files require a specially built decoder)<br/>
     &nbsp;&nbsp;9. encoding mode (in hex because it's a bitfield, always prefixed with "0x") <br/>
     &nbsp;&nbsp;10. filename (if available)<br/>
				</p>
    <p>
     The optional parameter can be used to output a single item from the 10 available.
    </p>

				<p>
					<b><tt>--help = this help display</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>
				<p>
					<b><tt>-i = ignore .wvc file (forces hybrid lossy decompression)</tt></b>
				</p>
				<p>
					This option forces WvUnpack to <i>not</i> look for a corresponding correction file when 
					unpacking hybrid mode files. This could be used to compare the effects of the hybrid 
					lossy mode or in the case that the .wvc had become corrupt and was preventing 
					decompression (although this is unlikely).
				</p>
				<p>
					<b><tt>-l = run at low priority (for smoother multitasking)</tt></b>
				</p>
				<p>
					This option can be used (in Windows only) to force WavPack to run at a low priority 
     and is handy for doing large unpack or verification batch conversions in the background.
				</p>
				<p>
					<b><tt>-m = calculate and display MD5 signature; verify if lossless</tt></b>
				</p>
				<p>
					Enables the calculation and display of the MD5 checksum on the uncompressed audio 
					data. If an MD5 sum for the original audio data is stored in the WavPack file then 
					this will be displayed also. For lossless operation these numbers should match (and 
					WvUnpack will verify that they match). For lossy operation these numbers should not 
					match.
				</p>
				<p>
					<b><tt>-n = no audio decoding</tt></b>
				</p>
				<p>
					Use this option with the <b>-xx</b> and <b>-cc</b> options to perform the tag extraction only.
				</p>

    <p>
     <b><tt>--no-overwrite = don't overwrite existing files, and don't ask to</tt></b>
    </p>
    <p>
     Normally WvUnpack checks that a file it is about to create already exists, and if it
     does will prompt the user before overwriting it. When this option is specified
     WvUnpack will skip this prompt and simply go to the next file (if there are any).
     This might be useful to resume a long, multi-file WvUnpack operation that was cancelled
     because in that case all the previously encoded files are quickly skipped.
    </p>

    <p>
     <b><tt>--normalize-floats = normalize floating-point audio if it isn't already</tt></b>
    </p>
    <p>
	    Floating-point audio is considered "normalized" when the fullscale range is +/-1.0 (although
					peaks may exceed this range, which is one of its advantages). Most PCM file formats (WAV, CAF, etc)
					require normalized floating-point audio, but it's possible to have WavPack floating-point audio
					files that are normalized to a different range, specifically those created from certain Cool Edit
					or Adobe Audition versions. If it is desired to export such files to a standard PCM format (or a
					normalized raw file) then include this option. Note that since the audio will be scaled, the
					operation will no longer be strictly lossless and so the option for verifying the MD5 checksum
					should not be used.
    </p>

    <p>
     <b><tt>--no-threads = do not use multiple threads for faster operation</tt></b>
    </p>
    <p>
     Normally, if WvUnpack can determine that it is running on a multicore system, it will
     use multiple threads to speed up the operation, sometimes significantly. Use this
     option to force single-threaded operation, for example in cases where multithreading
     is being done elsewhere. This is equivalent to specifing <b>--threads=1</b>.
    </p>

    <p>
     <b><tt>--no-utf8-convert = leave tag items in UTF-8 on extract or display</tt></b>
    </p>
    <p>
     The text fields of APEv2 tags are encoded in the UTF-8 variant of Unicode, so when
     tag information is displayed or extracted, it is converted to local multibyte encoding
     before being displayed or written to files. If your system already expects UTF-8, use this
     option to prevent the conversion.
    </p>
				<p>
					<b><tt>--pause</tt></b>
				</p>
				<p>
     Pause before exiting the console program (Windows only), allowing the user to press
     any key to continue. This might be useful to include in situations where the console
     window disappears before the completion status can be seen (like the
     WavPack FrontEnd) or when using the program icon as "drag and drop".
				</p>
				<p>
					<b><tt>-q = quiet (keep console output to a minimum)</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>
				<p>
					<b><tt>-r or --raw = force raw audio decode (results in .raw extension)</tt></b>
				</p>
				<p>
					This option causes WvUnpack to strip off any file header (and any 
					trailer) and only output the uncompressed audio data. If a file is created then it 
					will be given the .raw extension instead of the extension on the original file. Note that
     the raw data still conforms to standard WAV audio data (i.e., little-endian and signed,
     except for 8-bit data which is unsigned). DSD audio data is output in Philips DSDIFF
     format (i.e., channels interleaved by bytes, MSB first temporally) unless the
     <b>--raw-pcm</b> option is used instead.
				</p>
				<p>
					<b><tt>--raw-pcm = force raw PCM audio decode (results in .raw extension)</tt></b>
				</p>
				<p>
					This option is identical to the <b>-r</b> and <b>--raw</b> option above except in the
     case of DSD-encoded WavPack files which are decimated 8x to 24-bit PCM instead of being
     output in raw DSD. Note that the decimated PCM audio still contains large amounts of
     DSD quantization noise above the audible range and should be filtered or downsampled
     appropriately.
				</p>
				<p>
					<b><tt>-s = display summary information only to stdout (no decode)</tt></b>
				</p>
				<p>
					This option causes WvUnpack to simply display general information about the 
					specified WavPack files including source information, encoder version and options 
					and original MD5 checksum. This information can be directed into a file with the 
					"&gt;" operator. No file decoding or verification can occur with this option.
				</p>
				<p>
					<b><tt>-ss = display super summary (including tags) to stdout (no decode)</tt></b>
				</p>
				<p>
					This option is similar to -s except that information from any valid tag found
				 (either APEv2 or ID3v1) is also listed. For binary or multi-line text items, only
     the size of the tag data is displayed, but you can use the -x command to view this
     data (or redirect it to a file).
				</p>

				<p>
     <b><tt>--skip=[-][sample|hh:mm:ss.ss] = start decoding at specified sample/time</tt></b>
				</p>
				<p>
     Specifies an alternate start position for decoding, as either an integer sample index
     or as a time in hours, minutes, and seconds (with fraction). A minus ('-') sign indicates
     that the time or sample index is relative to the end of the file. The WavPack file must be seekable
     (i.e. not a pipe). This option can be used with the --until option to decode a
     specific region of a track.
				</p>

				<p>
					<b><tt>-t = copy input file's time stamp to output file(s)</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

				<p>
     <b><tt>--threads[=n] = enable multithreaded operation (optional n = 1 for no threading to 12 threads maximum)</tt></b>
				</p>
				<p>
     On modern multicore CPUs WvUnPack can get a significant performance boost by utilizing
     multiple threads. This is applicable to all modes, including verify. If the parameter
     <b>n</b> is omitted then the number of threads is determined automatically as a balance
     between operational speed and energy consumption. Using multiple threads is now the
     default operation on multicore systems, so this option is really only needed to fine-tune
     the number of threads.
				</p>

				<p>
     <b><tt>--until=[+|-][sample|hh:mm:ss.ss] = stop decoding at specified sample/time</tt></b>
				</p>
				<p>
     Specifies an alternate stop position for decoding, as either an integer sample index
     or as a time in hours, minutes, and seconds (with fraction). If a plus ('+') or minus ('-')
     sign is inserted before the specified sample (or time) then it becomes a relative amount,
     either from the position specified by a --start option (if plus) or from the end of the
     file (if minus).
				</p>

				<p>
					<b><tt>-v = verify source data only (no output file created)</tt></b>
				</p>
				<p>
					This option allows the user to verify the integrity of WavPack files (including 
					any correction file). Note that this option not only verifies that a WavPack file 
					has not been corrupted since creation, but it also verifies that the audio data 
					is being unpacked exactly as intended (even in the lossy mode). Therefore, it can 
					also detect algorithm errors, incompatible implementations of the encoder or 
					decoder, and even faulty processor hardware.
				</p>

				<p>
					<b><tt>-vv = quick verify (no output, version 5+ files only)</tt></b>
				</p>
				<p>
				 WavPack blocks have always contained a checksum for the audio data that is
					used during decoding (or verifying) to make sure the block is not corrupt.
				 Starting with version 5 however, WavPack blocks contain an additional checksum
					for the <i>encoded</i> data block which is used to increase robustness during decode
					by rejecting corrupt blocks before they are even parsed.
				</p>
				<p>
				 This option allows the user to verify the blocks in a WavPack file using just these
					additional checksums, and because the audio is not decoded this operation is <i>much</i>
					faster than the regular verify option (-v). It is not as complete as the full verify
					for detecting all possible issues, but should easily detect the case where a WavPack
					file has been corrupted due to a transfer or storage problem. This operation will
					automatically fall back to the regular (slow) verify mode if older files that do not
					have the new checksum are encountered.
				</p>

				<p>
					<b><tt>-vvv = quick verify verbose</tt></b>
				</p>
				<p>
				 This is similar to the -vv option above but also displays additional information that
     may be useful or interesting, like the channel count or chunks of unknown data between
     the WavPack blocks (which is valid up to 1 MB). This will also display whether the total
     length of the WavPack file is missing from the first block (which is valid, but means
     that the file requires an extra seek to the end to determine the length).
				</p>

				<p>
					<b><tt>--version = display program version to stdout</tt></b>
				</p>
				<p>
					Both the version of the command-line program and the WavPack library are displayed.
				</p>

				<p>
     <b><tt>-w or --wav = force output to Microsoft WAV format</tt></b>
				</p>
				<p>
     Output a Microsoft WAV file regardless of the input file format. All extra information in
     the original file's header and trailer will be lost and a "fresh" WAV header will be
     generated. For multichannel files, a WAVEFORMATEXTENSIBLE header is written, and for files
     over 4 GB, an RF64 file will be written (still with .wav extension).
     Note that DSD audio files will be decimated 8X and output as 24-bit PCM.
				</p>

				<p>
     <b><tt>--w64 = force output to Sony Wave64 format</tt></b>
				</p>
				<p>
     Output a Sony Wave64 file regardless of the input file format. All extra information in
     the original file's header and trailer will be lost and a "fresh" Wave64 header will be
     generated. Note that DSD audio files will be decimated 8X and output as 24-bit PCM.
				</p>

				<p>
     <b><tt>-x "Field" = extract specified tag field only to stdout (no audio decode)</tt></b>
				</p>
				<p>
					If the specified WavPack file contains a tag, and that tag contains the specified
     field, then only dump that field to stdout without decoding any audio. If the tag field
     is not found then an error is generated. Of course, the data may be redirected to a file
     or pipe. Only one field may be specified, and -c can be used as a shortcut to extract
     the CUESHEET field. 
				</p>
				<p>
     <b><tt>-xx "field[=file]" = extract specified tag field to file (with audio decode)</tt></b>
				</p>
				<p>
					If the specified WavPack file contains a tag, and that tag contains the specified 
     field, then that field will be extracted into an appropriately named file
     located in the same folder as the target .wav file. The name of the file will be
     the name of the tag field with a .txt extension for text items or, in the case of
     binary items, the name comes from the tag itself (although the convention is for this
     to be the tag field name with the extension from the original source file).
    </p>
    <p>
     If the automatically generated name is not acceptable, then a new name can be specified,
     and this new name specification may contain the following replacement codes:
				</p>
				<p>
					&nbsp;&nbsp;%a = audio output filename<br/>
					&nbsp;&nbsp;%t = tag field name (note: comes from data for binary tags)<br/>
					&nbsp;&nbsp;%e = extension from binary tag source file, or 'txt' for text tag<br/>
				</p>
				<p>
					<b><tt>-y = yes to overwrite warning (use with caution!)</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

				<p>
					<b><tt>-z[n] = don't (n=0 or omitted) or do (n=1) write to console title bar to show overall progress</tt></b>
				</p>
				<p>
					If enabled leaves "WvUnpack Completed" on the title bar.
				</p>
				</div>
			</div>

			<br/>
			<div class="box" id="wvgain">
				<div class="boxHeader">WvGain Options</div>
				<div class="boxContent">
				<p>
					<b><tt>-a = album mode (all files scanned are considered an album)</tt></b>
				</p>
				<p>
					The default mode for WvGain is to operate in <i>track</i> mode where each WavPack
				 file is independently analyzed for perceived volume (and peak level). This option
				 causes WvGain to additionally operate in <i>album</i> mode where all the specified
				 files are analyzed together as an "album" so that a set of composite values are
				 also generated. Note that when this mode is specified, WvGain must wait until the
				 end of scanning before making another pass through all the files to append the
				 calculated values.
				</p>
				<p>
					<b><tt>-c = clean ReplayGain values from all files (no analysis)</tt></b>
				</p>
				<p>
					This option does not perform any analysis but simply removes all ReplayGain values
				 from the specified files. 
				</p>
				<p>
					<b><tt>-d = display calculated values only (no files are modified)</tt></b>
				</p>
				<p>
					This option causes WvGain to do the specified analysis (with or with the -a switch
				 for <i>album</i> mode) and display the calculated values, but no values are actually
				 written to the files.
				</p>
				<p>
					<b><tt>-i = ignore .wvc file (forces hybrid lossy decompression)</tt></b>
				</p>
				<p>
					This option forces WvGain to not look for a corresponding correction file when 
					analyzing hybrid lossless mode files. This is probably a good idea in most cases
     because it's faster and the lossy versions tend to have slightly higher peak
     amplitudes than the lossless versions (although the overall levels will be
     almost identical).
				</p>
				<p>
					<b><tt>-l = run at low priority (for smoother multitasking)</tt></b>
				</p>
				<p>
					This option can be used (in Windows only) to force WvGain to run at a low priority 
     and is handy for doing large batch operations in the background.
				</p>
				<p>
					<b><tt>-n = new files only</tt></b>
				</p>
				<p>
					This option instructs WvGain to check the tags of the files first to make sure that they
     don't already have ReplayGain information. Files that already have ReplayGain information
     are skipped; in "album" mode a file encountered with ReplayGain information will abort the
     whole operation. 
				</p>
				<p>
					<b><tt>-q = quiet (keep console output to a minimum)</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>
				<p>
					<b><tt>-s = show stored values only (no analysis)</tt></b>
				</p>
				<p>
					This option does not perform any analysis but simply displays (to stdout) any
				 ReplayGain values stored in the specified WavPack files.
				</p>

				<p>
					<b><tt>-t = use multiple threads for faster operation</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

				<p>
					<b><tt>-1 = don't use multiple threads for faster operation</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

				<p>
					<b><tt>-v = display program version to stdout</tt></b>
				</p>
				<p>
					Both the version of the command-line program and the WavPack library are displayed.
				</p>

				<p>
					<b><tt>-z[n] = don't (n=0 or omit) or do (n=1) write to console title bar to show overall progress</tt></b>
				</p>
				<p>
					If enabled leaves "WvGain Completed" on the title bar.
				</p>
				</div>
			</div>

			<br/>
			<div class="box" id="wvtag">
				<div class="boxHeader">WvTag Options</div>
				<div class="boxContent">

				<p>
					<b><tt>--allow-huge-tags = allow tag data up to 16 MB (otherwise it's 1 MB)</tt></b>
				</p>
				<p>
     Normally WavPack allows the APEv2 tags to contain up to 1 MB of data. This limit was
     implemented to allow for their use on portable devices which may have limited memory or
     processing resources. However, in some situations it may be desirable to place more
     data in the tags (for high resolution cover art scans, for example) and this option
     permits that. Note that these files are not fully WavPack compliant and may not work
     in all situations or with older versions of WavPack programs and plugins.
				</p>

				<p>
     <b><tt>-c  = extract cuesheet only to stdout</tt></b>
				</p>
				<p>
					If the specified WavPack file contains an APEv2 tag, and that tag contains a CUESHEET
     field, then dump that text to stdout. If no cuesheet is found then an error is generated.
     This is equivalent to "-x cuesheet".
				</p>
				<p>
     <b><tt>-cc = extract cuesheet file (.cue)</tt></b>
				</p>
				<p>
					If the specified WavPack file contains an APEv2 tag, and that tag contains a CUESHEET
     field, then that text will be extracted into an appropriately named .cue file
     located in the same folder as the source file. This option has no effect if there's
     no cuesheet and is equivalent to "-xx cuesheet=%a.cue".
				</p>

				<p>
					<b><tt>--clean or --clear = clean all tag items from APEv2 tag (done first)</tt></b>
				</p>
				<p>
					This option, if specified, is performed first and results in a fresh tag to start with.
				</p>

				<p>
					<b><tt>-d "Field" = delete specified metadata item (either text or binary) from APEv2 tag</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

				<p>
					<b><tt>-h or --help = display usage information</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

				<p>
					<b><tt>--import-file file.wv = import all tag items from specified WavPack file</tt></b>
				</p>
				<p>
     This option is used to copy all the tag items from one WavPack file to one (or more) other
     WavPack files. Note that the "encoder" and "version" tags are excluded because those might no
     longer be valid. The tag items are <i>added</i> to any existing tags on the the target file(s),
     so include the <b>--clean</b> option if you want to start fresh.
    </p>
    <p>
     Note that if over 1 MB of image data is present, then the <b>--allow-huge-tags</b> option must be included
				</p>

				<p>
					<b><tt>--import-id3 = import applicable tag items from ID3v2.3 or ID3v2.4 tag on DSF and other files</tt></b>
				</p>
				<p>
					Sony's DSF file format specifies that these files may contain an ID3v2 tag at the end. WavPack
     considers this a trailing "wrapper" and stores it in the WavPack file as such so that the DSF
     file can be restored verbatim. However, stored this way it is not easily accessible for
     reading (and it is certainly not writable) because WavPack uses APEv2 (or, sometimes, ID3v1) tags
     for metadata. This option causes any trailing ID3v2.3 or ID3v2.4 tag in the DSF file (or other
     file type) to be scanned and all applicable items imported into the APEv2 tag, including cover art.
    </p>
    <p>
     Note that if over 1 MB of image data is present, then the <b>--allow-huge-tags</b> option must be included
				</p>

				<p>
					<b><tt>-l or --list = list all tag items (done last)</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

    <p>
     <b><tt>--no-utf8-convert = don't recode passed tags to UTF-8, assume they are UTF-8 already</tt></b>
    </p>
    <p>
     The text fields of APEv2 tags are encoded in the UTF-8 variant of Unicode, so when
     tag information is passed in on the command-line they are converted to UTF-8 before
     being stored. If your system is already passing the strings in UTF-8, use this option
     to prevent double conversion.
    </p>

				<p>
					<b><tt>-q = quiet (keep console output to a minimum)</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

				<p>
					<b><tt>-v or --version = display program version to stdout</tt></b>
				</p>
				<p>
					Both the version of the command-line program and the WavPack library are displayed.
				</p>

				<p>
					<b><tt>-w "Field=[@]Value" = write specified metadata to APEv2 tag</tt></b>
				</p>
				<p>
					Write specified information to APEv2 tag appended to WavPack file(s). May be used
					multiple times for multiple fields. APEv2 tags are the preferred tag format for
					WavPack files and are read by all the standard WavPack playback plugins. If the
					specified value begins with a '@', then the value is assumed to be a
					filename which is used to obtain the item's actual value. This is handy for
					including the CUESHEET field for use with images files + cuesheets and foobar2000.
				 The filename may contain wildcards if it matches exactly one file. Also, if the
				 file cannot be found in the current directory then the source directory (if
     specified) is also checked.
				</p>

				<p>
					<b><tt>--write-binary-tag "Field=@file.ext" = write specified binary metadata to APEv2 tag</tt></b>
				</p>
				<p>
					Write specified binary file to APEv2 tag appended to WavPack file(s). This is most
     commonly used to embed album cover art into WavPack files (with the field name
     <b>"Cover Art (Front)"</b>), but could be used for anything desired. A file must be
     specified (the data cannot come from the command-line) and it may contain wildcards
     if it matches exactly one file. Also, if the file cannot be found in the current directory
     then the source directory (if specified) is also checked.
				</p>

				<p>
     <b><tt>-x "Field" = extract specified tag field only to stdout (no audio decode)</tt></b>
				</p>
				<p>
					If the specified WavPack file contains a tag, and that tag contains the specified
     field, then dump that field to stdout. If the tag field is not found then an error is
     generated. Of course, the data may be redirected to a file or pipe. Only one field may
     be specified, and -c can be used as a shortcut to extract the CUESHEET field. 
				</p>

				<p>
     <b><tt>-xx "field[=file]" = extract specified tag field to file (with audio decode)</tt></b>
				</p>
				<p>
					If the specified WavPack file contains a tag, and that tag contains the specified 
     field, then that field will be extracted into an appropriately named file
     located in the same folder as the source file. The name of the file will be
     the name of the tag field with a .txt extension for text items or, in the case of
     binary items, the name comes from the tag itself (although the convention is for this
     to be the tag field name with the extension from the original source file).
    </p>
    <p>
     If the automatically generated name is not acceptable, then a new name can be specified,
     and this new name specification may contain the following replacement codes:
				</p>
				<p>
					&nbsp;&nbsp;%a = audio output filename<br/>
					&nbsp;&nbsp;%t = tag field name (note: comes from data for binary tags)<br/>
					&nbsp;&nbsp;%e = extension from binary tag source file, or 'txt' for text tag<br/>
				</p>

				<p>
					<b><tt>-y = yes to overwrite warning (use with caution!)</tt></b>
				</p>
				<p>
					Self explanatory.
				</p>

				</div>
			</div>

			<br/>
			<div class="box" id="usage">
				<div class="boxHeader">Usage Guide</div>
				<div class="boxContent">
				<p>
					The options of WavPack can be a little confusing at first, so here's a little 
					tutorial on what to try first if you're not sure. First of all, if you are only 
					interested in lossless compression, try the default operation with no options. This 
					will give a decent compression ratio at a very good speed for both packing and 
					unpacking. Also, it is usually recommended to add the -x switch because it can
     provide a decent improvement in compression with only a minor cost in encoding
     time (about 2x) and <i>no</i> cost in decoding speed.
    </p>
				<p>
     If speed is less of an issue try the "high" mode (-h), this will increase both 
					the packing and unpacking time by about 50% (although still not nearly as slow as many 
					other lossless compressors) while improving the compression ratio somewhat. If speed 
					is paramount use the "fast" (-f) option. This gives the fastest packing and unpacking 
					speed possible with WavPack, while still providing a very reasonable compression 
					ratio. Again, the -x switch works well with these two modes.
				</p>
				<p>
					If decoding speed is important, but the time to encode is not (because it happens 
					only once), try using numeric parameters with the -x mode (valid values are 2-6).
     This will slow the encoding way down (depending on the level) but still have no
     effect on decoding speed. In some music files, this option can actually bump the
     compression ratio a whole mode (i.e. the "fast" mode can match the default mode,
     etc.). For the best lossless compression WavPack can offer, use -hhx6 (but be
     prepared to make yourself a sandwich!).
				</p>
				<p>
					If you want to try the hybrid mode, all the above applies but you also have to 
					choose a bitrate. The quality of WavPack's lossy mode cannot match the conventional 
					lossy codecs like MP3 and WMA at similar bitrates, and in fact it won't even 
					operate at the most common bitrate of 128 kbps (with CD audio, at least). The 
					lowest bitrate that I recommend for WavPack lossy is 256 kbps which can provide 
					transparent reproduction for most non-critical listening situations and is roughly
					equivalent to LAME MP3 encoding somewhere between 160 kbps and 192 kbps (CBR). Above
     256 kbps the quality of WavPack's lossy mode 
					increases rapidly, with added quantization noise (which is the only artifact) 
					dropping by about 1 dB for every 15 kbps. At 320 kbps the quality is difficult 
					for even critical listeners to distinguish from the original, and at 384 kbps 
					WavPack becomes essentially transparent.
				</p>
				<p>
					The various modes and options that apply to lossless mode also apply to hybrid mode. 
					However the "extra" mode (-x) deserves a special mention. Most of the time, the 
					"extra" mode makes only modest improvements because WavPack has already been finely 
					tuned for ordinary music. However, sometimes there is an instrument or sound 
					situation that does not compress well with the standard settings and the "extra" 
					mode can make a significant improvement. In lossless mode this would slightly improve 
					the overall compression ratio and would probably go unnoticed. However, in lossy 
					mode that difficult section might trigger clearly audible noise to be added, and 
					in this case the "extra" mode would save the day by greatly reducing the noise in 
					the exact spot where it might have been audible.
				</p>
				<p>
					Here is a chart of recommended settings for the various useful bitrates:
				</p>
				<pre>
	 Bitrate     High quality   Faster Encode   Faster Decode 
 	 -------     ------------   -------------   ------------- 
 	 256 kbps    -hb256x3       -hb256x         -b256x3
 	 320 kbps    -hb320x3       -hb320x         -b320x3
 	 384 kbps    -b384x3        -hb384x         -fb384x3
				</pre>
				</div>
			</div>

			<br/>
			<div class="box" id="noncd">
				<div class="boxHeader">Non-CD</div>
				<div class="boxContent">
    <p>
     In addition to regular CD audio data, WavPack hybrid mode is also well suited to other
     sampling rates, bitdepths, and channel configurations. For generating bitrates lower
     than the minimum of 196 kbps that applies to CD audio, it is necessary to resample the
     audio to lower sampling rates and/or convert to mono using an external program. Note that
     conventional lossy codecs accomplish the same thing by lowpass filtering the source when
     used for low bitrates.
    </p>
    <p>
     For high-resolution audio (e.g., 24-bit/96-kHz) WavPack hybrid makes a particularly logical
     choice. The reasoning here is that standard lossless encoding achieves relatively poor compression
     on this type of material because the additional 8 bits are essentially noise and cannot be
     further compressed. In addition, conventional lossy codecs like MP3 or AAC are based on
     psychoacoustic models of hearing that presume CD audio quality (16-bit/44-kHz) to be
     "perfect", and therefore discard everything outside of that range.
    </p>
    <p>
     When used at around 1024 kbps, WavPack hybrid preserves the full bandwidth and dynamic
     range of 24/96 sources. It accomplishes this by moving most of the quantization noise up
     above the audible audio band using first-order noise shaping in much the same way that DSD
     does. In fact, at normally loud listening levels, the added quantization noise remains
     completely inaudible, <i>even without the masking effect of the material itself!</i> While
     1024 kbps may seem high at first, keep in mind that this is right in the range of some
     losslessly compressed CD-quality music, and is approximately 1/3 of the bitrate required to
     losslessly compress the same 24/96 material! 
    </p> 
    <p>
     WavPack also handles multichannel audio like 5.1 and 7.1 channel configurations, and these
     can be stored in lossless and hybrid modes (including hybrid lossless). With all the
     combinations of sampling rates and numbers of channels it might become confusing to
     determine a desirable bitrate. This is an ideal application of the alternate form of the
     <b>-b</b> option which accepts a number of bits per sample rather than kilobits per second.
     When using this form a reasonable starting point of experimentation is 4 bits per sample
     (-b4). Except for extreme cases this will produce transparent results (it is equivalent to
     about 350 kbps for CD audio). Note that the compression algorithms are identical when using
     this form; it is simply an alternate way of specifying the bitrate.
    </p>
    <p>
     Representative samples ranging from 24 kbps to 1024 kbps are found in the "hybrid_bitrates"
     directory of the WavPack Test Suite which is currently hosted
     <a href="http://www.rarewares.org/wavpack/test_suite.zip">here</a>, but which can always
     be found on the WavPack website.
    </p>
				</div>
			</div>

			<br/>
			<div class="box" id="plugins">
				<div class="boxHeader">Plugins</div>
				<div class="boxContent">
					<p>The documentation for the various WavPack plugins is provided as readme.txt 
					files included with those plugins.</p>
				</div>
			</div>
			
			<br/>
			<div class="box" id="disclaimer">
				<div class="boxContent">
					<p>
						WavPack and its associated utilities are free programs; feel free to give them 
						to anyone who may find them useful. There is no warranty provided and you agree 
						to use them completely at your own risk. Be sure to visit <a 
						href="http://www.wavpack.com">www.wavpack.com</a> for the latest version of WavPack.
					</p>
				</div>
			</div>
			<br/>
			<center>
				<a href="index.html">home</a><br/><br/>
				<a href="http://validator.w3.org/check?uri=referer" target="_blank">XHTML 1.0</a><br/>
				<a href="http://jigsaw.w3.org/css-validator/check/referer" target="_blank">CSS 2.1</a><br/>
			</center><br/>
			<div class="hidden">Website design by <a href="mailto:ariakis@comcast.net">Ariakis</a></div>
		</div>
	</body>
</html>