ViewVC logotype

Annotation of /trunk/3rdparty/SoundTouch/README.html

Parent Directory Parent Directory | Revision Log Revision Log

Revision 15 - (hide annotations) (download) (as text)
Mon Sep 6 13:44:19 2010 UTC (10 years, 4 months ago) by william
File MIME type: text/html
File size: 31002 byte(s)
initial checkout of r3113 from upstream repository
1 william 15 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
2     <html>
3     <head>
4     <meta http-equiv="Content-Type"
5     content="text/html; charset=windows-1252">
6     <meta http-equiv="Content-Language" content="en-us">
7     <meta name="author" content="Olli Parviainen">
8     <meta name="description"
9     content="Readme file for SoundTouch audio processing library">
10     <meta name="GENERATOR" content="Microsoft FrontPage 4.0">
11     <meta name="ProgId" content="FrontPage.Editor.Document">
12     <title>SoundTouch library README</title>
13     <style>
14     <!--
15     .normal { font-family: Arial }
16     -->
17     </style>
18     </head>
19     <body class="normal">
20     <hr>
21     <h1>SoundTouch audio processing library v1.5.0
22     </h1>
23     <p class="normal">SoundTouch library Copyright (c) Olli
24     Parviainen 2002-2009 </p>
25     <hr>
26     <h2>1. Introduction </h2>
27     <p>SoundTouch is an open-source audio
28     processing library that allows changing the sound tempo, pitch
29     and playback rate parameters independently from each other, i.e.:</p>
30     <ul>
31     <li>Sound tempo can be increased or decreased while
32     maintaining the original pitch</li>
33     <li>Sound pitch can be increased or decreased while
34     maintaining the original tempo </li>
35     <li>Change playback rate that affects both tempo
36     and pitch at the same time </li>
37     <li>Choose any combination of tempo/pitch/rate</li>
38     </ul>
39     <h3>1.1 Contact information </h3>
40     <p>Author email: oparviai 'at' iki.fi </p>
41     <p>SoundTouch WWW page: <a href="http://www.surina.net/soundtouch">http://www.surina.net/soundtouch</a></p>
42     <hr>
43     <h2>2. Compiling SoundTouch</h2>
44     <p>Before compiling, notice that you can choose the sample data format
45     if it's desirable to use floating point sample
46     data instead of 16bit integers. See section "sample data format"
47     for more information.</p>
48     <h3>2.1. Building in Microsoft Windows</h3>
49     <p>Project files for Microsoft Visual C++ 6.0 and Visual C++ .NET are
50     supplied with the source code package.&nbsp;</p>
51     <p> Please notice that SoundTouch
52     library uses processor-specific optimizations for Pentium III and AMD
53     processors. Visual Studio .NET and later versions supports the required
54     instructions by default, but Visual Studio 6.0 requires a processor pack upgrade
55     to be installed in order to support these optimizations. The processor pack upgrade can be downloaded from
56     Microsoft site at this URL:</p>
57     <p><a href="http://msdn.microsoft.com/en-us/vstudio/aa718349.aspx">http://msdn.microsoft.com/en-us/vstudio/aa718349.aspx</a></p>
58     <p>If the above URL is unavailable or removed, go
59     to <a href="http://msdn.microsoft.com/">http://msdn.microsoft.com</a>
60     and perform a search with keywords &quot;processor pack&quot;. </p>
61     <p>To build the binaries with Visual C++
62     compiler, either run &quot;make-win.bat&quot; script, or open the
63     appropriate project files in source code directories with Visual
64     Studio. The final executable will appear under the &quot;SoundTouch\bin&quot;
65     directory. If using the Visual Studio IDE instead of the make-win.bat script, directories bin and
66     lib may need to be created manually to the SoundTouch
67     package root for the final executables. The make-win.bat script
68     creates these directories automatically.
69     </p>
70     <h3>2.2. Building in Gnu platforms</h3>
71     <p>The SoundTouch library can be compiled in
72     practically any platform supporting GNU compiler (GCC) tools.
73     SoundTouch have been tested with gcc version 3.3.4., but it
74     shouldn't be very specific about the gcc version. Assembler-level
75     performance optimizations for GNU platform are currently available in
76     x86 platforms only, they are automatically disabled and replaced with
77     standard C routines in other processor platforms.</p>
78     <p>To build and install the binaries, run the
79     following commands in the SoundTouch/ directory:</p>
80     <table border="0" cellpadding="0" cellspacing="4">
81     <tbody>
82     <tr valign="top">
83     <td>
84     <pre>./configure -</pre>
85     </td>
86     <td>
87     <p>Configures the SoundTouch package for the local
88     environment.</p>
89     </td>
90     </tr>
91     <tr valign="top">
92     <td>
93     <pre>make -</pre>
94     </td>
95     <td>
96     <p>Builds the SoundTouch library &amp;
97     SoundStretch utility.</p>
98     </td>
99     </tr>
100     <tr valign="top">
101     <td>
102     <pre>make install -</pre>
103     </td>
104     <td>
105     <p>Installs the SoundTouch &amp; BPM libraries
106     to <b>/usr/local/lib</b> and SoundStretch utility to <b>/usr/local/bin</b>.
107     Please notice that 'root' privileges may be required to install the
108     binaries to the destination locations.</p>
109     </td>
110     </tr>
111     </tbody>
112     </table>
113     <h4><b>2.2.1 Required GNU tools</b>&nbsp;</h4>
114     <p> Bash shell, GNU C++ compiler, libtool, autoconf and automake tools are required
115     for compiling
116     the SoundTouch library. These are usually included with the GNU/Linux distribution, but if
117     not, install these packages first. For example, in Ubuntu Linux these can be acquired and
118     installed with the following command:</p>
119     <pre><b>sudo apt-get install <font SIZE="2">automake autoconf libtool build-essential</font></b></pre>
120     <h4><b>2.2.2 Problems with GCC compiler compatibility</b></h4>
121     <p>At the release time the SoundTouch package has been tested to compile in
122     GNU/Linux platform. However, in past it's happened that new gcc versions aren't
123     necessarily compatible with the assembler settings used in the optimized
124     routines. <b>If you have problems getting the
125     SoundTouch library compiled, try the workaround of disabling the optimizations</b>
126     by editing the file &quot;include/STTypes.h&quot; and removing the following
127     definition there:</p>
128     <blockquote>
129     <pre>#define ALLOW_OPTIMIZATIONS 1</pre>
130     </blockquote>
131     <h4><b>2.2.3 Problems with configure script or build process</b>&nbsp;</h4>
132     <p>Incompatibilities between various GNU toolchain versions may cause errors when running the &quot;configure&quot; script or building the source
133     codes, if your GNU tool versions are not compatible with the versions used for
134     preparing the SoundTouch kit.&nbsp;</p>
135     <p>To resolve the issue, regenerate the configure scripts with your local tool
136     set by running
137     the &quot;<b>./bootstrap</b>&quot; script included in the SoundTouch source code
138     kit. After that, run the <b>configure</b> script and <b>make</b> as usually.</p>
139     <h4><b>2.2.4 Compiler issues with non-x86 processors</b></h4>
140     <p>SoundTouch library works also on non-x86 processors.</p>
141     <p>However, in case that you get compiler errors when trying to compile for non-Intel processor, edit the file
142     &quot;<b>source\SoundTouch\Makefile.am</b>&quot; and remove the &quot;<b>-msse2</b>&quot;
143     flag on the <b>AM_CXXFLAGS </b>line:</p>
144     <pre><b>AM_CXXFLAGS=-O3 -fcheck-new -I../../include&nbsp;&nbsp;&nbsp; # Note: -msse2 flag removed!</b></pre>
145     <p>After that, run &quot;<b>./bootstrap</b>&quot; script, and then run <b>configure</b>
146     and <b>make</b> again.</p>
147     <hr>
148     <h2>3. About implementation &amp; Usage tips</h2>
149     <h3>3.1. Supported sample data formats</h3>
150     <p>The sample data format can be chosen
151     between 16bit signed integer and 32bit floating point values, the
152     default is 32bit floating point. </p>
154     <p>
155     In Windows environment, the sample data format is chosen
156     in file &quot;STTypes.h&quot; by choosing one of the following
157     defines:</p>
158     <ul>
159     <li><span style="font-weight: bold;">#define INTEGER_SAMPLES</span>
160     for 16bit signed
161     integer</li>
162     <li><span style="font-weight: bold;">#define FLOAT_SAMPLES</span> for
163     32bit floating point</li>
164     </ul>
165     <p>
166     In GNU environment, the floating sample format is used by default, but
167     integer sample format can be chosen by giving the
168     following switch to the configure script:
169     <blockquote>
170     <pre>./configure --enable-integer-samples</pre>
171     </blockquote>
173     <p>The sample data can have either single (mono)
174     or double (stereo) audio channel. Stereo data is interleaved so
175     that every other data value is for left channel and every second
176     for right channel. Notice that while it'd be possible in theory
177     to process stereo sound as two separate mono channels, this isn't
178     recommended because processing the channels separately would
179     result in losing the phase coherency between the channels, which
180     consequently would ruin the stereo effect.</p>
181     <p>Sample rates between 8000-48000H are
182     supported.</p>
183     <h3>3.2. Processing latency</h3>
184     <p>The processing and latency constraints of
185     the SoundTouch library are:</p>
186     <ul>
187     <li>Input/output processing latency for the
188     SoundTouch processor is around 100 ms. This is when time-stretching is
189     used. If the rate transposing effect alone is used, the latency
190     requirement
191     is much shorter, see section 'About algorithms'.</li>
192     <li>Processing CD-quality sound (16bit stereo
193     sound with 44100H sample rate) in real-time or faster is possible
194     starting from processors equivalent to Intel Pentium 133Mh or better,
195     if using the "quick" processing algorithm. If not using the "quick"
196     mode or
197     if floating point sample data are being used, several times more CPU
198     power is typically required.</li>
199     </ul>
200     <h3>3.3. About algorithms</h3>
201     <p>SoundTouch provides three seemingly
202     independent effects: tempo, pitch and playback rate control.
203     These three controls are implemented as combination of two primary
204     effects, <em>sample rate transposing</em> and <em>time-stretching</em>.</p>
205     <p><em>Sample rate transposing</em> affects
206     both the audio stream duration and pitch. It's implemented simply
207     by converting the original audio sample stream to the&nbsp; desired
208     duration by interpolating from the original audio samples. In SoundTouch, linear interpolation with anti-alias filtering is
209     used. Theoretically a higher-order interpolation provide better
210     result than 1st order linear interpolation, but in audio
211     application linear interpolation together with anti-alias
212     filtering performs subjectively about as well as higher-order
213     filtering would.</p>
214     <p><em>Time-stretching </em>means changing
215     the audio stream duration without affecting it's pitch. SoundTouch
216     uses WSOLA-like time-stretching routines that operate in the time
217     domain. Compared to sample rate transposing, time-stretching is a
218     much heavier operation and also requires a longer processing
219     "window" of sound samples used by the
220     processing algorithm, thus increasing the algorithm input/output
221     latency. Typical i/o latency for the SoundTouch
222     time-stretch algorithm is around 100 ms.</p>
223     <p>Sample rate transposing and time-stretching
224     are then used together to produce the tempo, pitch and rate
225     controls:</p>
226     <ul>
227     <li><strong>'Tempo'</strong> control is
228     implemented purely by time-stretching.</li>
229     <li><strong>'Rate</strong>' control is implemented
230     purely by sample rate transposing.</li>
231     <li><strong>'Pitch</strong>' control is
232     implemented as a combination of time-stretching and sample rate
233     transposing. For example, to increase pitch the audio stream is first
234     time-stretched to longer duration (without affecting pitch) and then
235     transposed back to original duration by sample rate transposing, which
236     simultaneously reduces duration and increases pitch. The result is
237     original duration but increased pitch.</li>
238     </ul>
239     <h3>3.4 Tuning the algorithm parameters</h3>
240     <p>The time-stretch algorithm has few
241     parameters that can be tuned to optimize sound quality for
242     certain application. The current default parameters have been
243     chosen by iterative if-then analysis (read: "trial and error")
244     to obtain best subjective sound quality in pop/rock music
245     processing, but in applications processing different kind of
246     sound the default parameter set may result into a sub-optimal
247     result.</p>
248     <p>The time-stretch algorithm default
249     parameter values are set by the following #defines in file &quot;TDStretch.h&quot;:</p>
250     <blockquote>
253     #define DEFAULT_OVERLAP_MS 8</pre>
254     </blockquote>
255     <p>These parameters affect to the time-stretch
256     algorithm as follows:</p>
257     <ul>
258     <li><strong>DEFAULT_SEQUENCE_MS</strong>: This is
259     the default length of a single processing sequence in milliseconds
260     which determines the how the original sound is chopped in
261     the time-stretch algorithm. Larger values mean fewer sequences
262     are used in processing. In principle a larger value sounds better when
263     slowing down the tempo, but worse when increasing the tempo and vice
264     versa.&nbsp;<br>
265     <br>
266     By default, this setting value is calculated automatically according to
267     tempo value.<br>
268     </li>
269     <li><strong>DEFAULT_SEEKWINDOW_MS</strong>: The seeking window
270     default length in milliseconds is for the algorithm that seeks the best
271     possible overlapping location. This determines from how
272     wide a sample "window" the algorithm can use to find an optimal mixing
273     location when the sound sequences are to be linked back together.&nbsp;<br>
274     <br>
275     The bigger this window setting is, the higher the possibility to find a
276     better mixing position becomes, but at the same time large values may
277     cause a "drifting" sound artifact because neighboring sequences can be
278     chosen at more uneven intervals. If there's a disturbing artifact that
279     sounds as if a constant frequency was drifting around, try reducing
280     this setting.<br>
281     <br>
282     By default, this setting value is calculated automatically according to
283     tempo value.<br>
284     </li>
285     <li><strong>DEFAULT_OVERLAP_MS</strong>: Overlap
286     length in milliseconds. When the sound sequences are mixed back
287     together to form again a continuous sound stream, this parameter
288     defines how much the ends of the consecutive sequences will overlap with each other.<br>
289     <br>
290     This shouldn't be that critical parameter. If you reduce the
291     DEFAULT_SEQUENCE_MS setting by a large amount, you might wish to try a
292     smaller value on this.</li>
293     </ul>
294     <p>Notice that these parameters can also be
295     set during execution time with functions "<strong>TDStretch::setParameters()</strong>"
296     and "<strong>SoundTouch::setSetting()</strong>".</p>
297     <p>The table below summaries how the
298     parameters can be adjusted for different applications:</p>
299     <table border="1">
300     <tbody>
301     <tr>
302     <td valign="top"><strong>Parameter name</strong></td>
303     <td valign="top"><strong>Default value
304     magnitude</strong></td>
305     <td valign="top"><strong>Larger value
306     affects...</strong></td>
307     <td valign="top"><strong>Smaller value
308     affects...</strong></td>
309     <td valign="top"><strong>Effect to CPU burden</strong></td>
310     </tr>
311     <tr>
312     <td valign="top">
313     <pre>SEQUENCE_MS</pre>
314     </td>
315     <td valign="top">Default value is relatively
316     large, chosen for slowing down music tempo</td>
317     <td valign="top">Larger value is usually
318     better for slowing down tempo. Growing the value decelerates the
319     "echoing" artifact when slowing down the tempo.</td>
320     <td valign="top">Smaller value might be better
321     for speeding up tempo. Reducing the value accelerates the "echoing"
322     artifact when slowing down the tempo </td>
323     <td valign="top">Increasing the parameter
324     value reduces computation burden</td>
325     </tr>
326     <tr>
327     <td valign="top">
328     <pre>SEEKWINDOW_MS</pre>
329     </td>
330     <td valign="top">Default value is relatively
331     large, chosen for slowing down music tempo</td>
332     <td valign="top">Larger value eases finding a
333     good mixing position, but may cause a "drifting" artifact</td>
334     <td valign="top">Smaller reduce possibility to
335     find a good mixing position, but reduce the "drifting" artifact.</td>
336     <td valign="top">Increasing the parameter
337     value increases computation burden</td>
338     </tr>
339     <tr>
340     <td valign="top">
341     <pre>OVERLAP_MS</pre>
342     </td>
343     <td valign="top">Default value is relatively
344     large, chosen to suit with above parameters.</td>
345     <td valign="top">&nbsp;</td>
346     <td valign="top">If you reduce the "sequence
347     ms" setting, you might wish to try a smaller value.</td>
348     <td valign="top">Increasing the parameter
349     value increases computation burden</td>
350     </tr>
351     </tbody>
352     </table>
353     <h3>3.5 Performance Optimizations </h3>
354     <p><strong>General optimizations:</strong></p>
355     <p>The time-stretch routine has a 'quick' mode
356     that substantially speeds up the algorithm but may degrade the
357     sound quality by a small amount. This mode is activated by
358     calling SoundTouch::setSetting() function with parameter&nbsp; id
359     of SETTING_USE_QUICKSEEK and value "1", i.e. </p>
360     <blockquote>
361     <p>setSetting(SETTING_USE_QUICKSEEK, 1);</p>
362     </blockquote>
363     <p><strong>CPU-specific optimizations:</strong></p>
364     <ul>
365     <li>Intel MMX optimized routines are used with
366     compatible CPUs when 16bit integer sample type is used. MMX optimizations are available both in Win32 and Gnu/x86 platforms.
367     Compatible processors are Intel PentiumMMX and later; AMD K6-2, Athlon
368     and later. </li>
369     <li>Intel SSE optimized routines are used with
370     compatible CPUs when floating point sample type is used. SSE optimizations are currently implemented for Win32 platform only.
371     Processors compatible with SSE extension are Intel processors starting
372     from Pentium-III, and AMD processors starting from Athlon XP. </li>
373     <li>AMD 3DNow! optimized routines are used with
374     compatible CPUs when floating point sample type is used, but SSE
375     extension isn't supported . 3DNow! optimizations are currently
376     implemented for Win32 platform only. These optimizations are used in
377     AMD K6-2 and Athlon (classic) CPU's; better performing SSE routines are
378     used with AMD processor starting from Athlon XP. </li>
379     </ul>
380     <hr>
381     <h2><a name="SoundStretch"></a>4. SoundStretch audio processing utility
382     </h2>
383     <p>SoundStretch audio processing utility<br>
384     Copyright (c) Olli Parviainen 2002-2009</p>
385     <p>SoundStretch is a simple command-line
386     application that can change tempo, pitch and playback rates of
387     WAV sound files. This program is intended primarily to
388     demonstrate how the &quot;SoundTouch&quot; library can be used to
389     process sound in your own program, but it can as well be used for
390     processing sound files.</p>
391     <h3>4.1. SoundStretch Usage Instructions</h3>
392     <p>SoundStretch Usage syntax:</p>
393     <blockquote>
394     <pre>soundstretch infilename outfilename [switches]</pre>
395     </blockquote>
396     <p>Where: </p>
397     <table border="0" cellpadding="2" width="100%">
398     <tbody>
399     <tr>
400     <td valign="top">
401     <pre>&quot;infilename&quot;</pre>
402     </td>
403     <td valign="top">Name of the input sound
404     data file (in .WAV audio file format). Give &quot;stdin&quot; as filename to use
405     standard input pipe. </td>
406     </tr>
407     <tr>
408     <td valign="top">
409     <pre>&quot;outfilename&quot;</pre>
410     </td>
411     <td valign="top">Name of the output sound
412     file where the resulting sound is saved (in .WAV audio file format).
413     This parameter may be omitted if you&nbsp; don't want to save the
414     output
415     (e.g. when only calculating BPM rate with '-bpm' switch). Give &quot;stdout&quot;
416     as filename to use standard output pipe.</td>
417     </tr>
418     <tr>
419     <td valign="top">
420     <pre>&nbsp;[switches]</pre>
421     </td>
422     <td valign="top">Are one or more control
423     switches.</td>
424     </tr>
425     </tbody>
426     </table>
427     <p>Available control switches are:</p>
428     <table border="0" cellpadding="2" width="100%">
429     <tbody>
430     <tr>
431     <td valign="top">
432     <pre>-tempo=n </pre>
433     </td>
434     <td valign="top">Change the sound tempo by n
435     percents (n = -95.0 .. +5000.0 %) </td>
436     </tr>
437     <tr>
438     <td valign="top">
439     <pre>-pitch=n</pre>
440     </td>
441     <td valign="top">Change the sound pitch by n
442     semitones (n = -60.0 .. + 60.0 semitones) </td>
443     </tr>
444     <tr>
445     <td valign="top">
446     <pre>-rate=n</pre>
447     </td>
448     <td valign="top">Change the sound playback rate by
449     n percents (n = -95.0 .. +5000.0 %) </td>
450     </tr>
451     <tr>
452     <td valign="top">
453     <pre>-bpm=n</pre>
454     </td>
455     <td valign="top">Detect the Beats-Per-Minute (BPM) rate of the sound and adjust the tempo to meet 'n'
456     BPMs. When this switch is
457     applied, the &quot;-tempo&quot; switch is ignored. If "=n" is
458     omitted, i.e. switch &quot;-bpm&quot; is used alone, then the BPM rate is
459     estimated and displayed, but tempo not adjusted according to the BPM
460     value. </td>
461     </tr>
462     <tr>
463     <td valign="top">
464     <pre>-quick</pre>
465     </td>
466     <td valign="top">Use quicker tempo change
467     algorithm. Gains speed but loses sound quality. </td>
468     </tr>
469     <tr>
470     <td valign="top">
471     <pre>-naa</pre>
472     </td>
473     <td valign="top">Don't use anti-alias
474     filtering in sample rate transposing. Gains speed but loses sound
475     quality. </td>
476     </tr>
477     <tr>
478     <td valign="top">
479     <pre>-license</pre>
480     </td>
481     <td valign="top">Displays the program license
482     text (LGPL)</td>
483     </tr>
484     </tbody>
485     </table>
486     <p>Notes:</p>
487     <ul>
488     <li>To use standard input/output pipes for processing, give &quot;stdin&quot;
489     and &quot;stdout&quot; as input/output filenames correspondingly. The
490     standard input/output pipes will still carry the audio data in .wav audio
491     file format.</li>
492     <li>The numerical switches allow both integer (e.g. "-tempo=123") and decimal (e.g.
493     "-tempo=123.45") numbers.</li>
494     <li>The &quot;-naa&quot; and/or "-quick" switches can be
495     used to reduce CPU usage while compromising some sound quality </li>
496     <li>The BPM detection algorithm works by detecting
497     repeating bass or drum patterns at low frequencies of &lt;250Hz. A
498     lower-than-expected BPM figure may be reported for music with uneven or
499     complex bass patterns. </li>
500     </ul>
501     <h3>4.2. SoundStretch usage examples </h3>
502     <p><strong>Example 1</strong></p>
503     <p>The following command increases tempo of
504     the sound file &quot;originalfile.wav&quot; by 12.5% and stores result to file &quot;destinationfile.wav&quot;:</p>
505     <blockquote>
506     <pre>soundstretch originalfile.wav destinationfile.wav -tempo=12.5</pre>
507     </blockquote>
508     <p><strong>Example 2</strong></p>
509     <p>The following command decreases the sound
510     pitch (key) of the sound file &quot;orig.wav&quot; by two
511     semitones and stores the result to file &quot;dest.wav&quot;:</p>
512     <blockquote>
513     <pre>soundstretch orig.wav dest.wav -pitch=-2</pre>
514     </blockquote>
515     <p><strong>Example 3</strong></p>
516     <p>The following command processes the file &quot;orig.wav&quot; by decreasing the sound tempo by 25.3% and
517     increasing the sound pitch (key) by 1.5 semitones. Resulting .wav audio data is
518     directed to standard output pipe:</p>
519     <blockquote>
520     <pre>soundstretch orig.wav stdout -tempo=-25.3 -pitch=1.5</pre>
521     </blockquote>
522     <p><strong>Example 4</strong></p>
523     <p>The following command detects the BPM rate
524     of the file &quot;orig.wav&quot; and adjusts the tempo to match
525     100 beats per minute. Result is stored to file &quot;dest.wav&quot;:</p>
526     <blockquote>
527     <pre>soundstretch orig.wav dest.wav -bpm=100</pre>
528     </blockquote>
529     <p><strong>Example 5</strong></p>
530     <p>The following command reads .wav sound data from standard input pipe and
531     estimates the BPM rate:</p>
532     <blockquote>
533     <pre>soundstretch stdin -bpm</pre>
534     </blockquote>
535     <hr>
536     <h2>5. Change History</h2>
537     <h3>5.1. SoundTouch library Change History </h3>
539     <p><strong>1.5.0:</strong></p>
540     <ul>
541     <li>Added normalization to correlation calculation and improvement automatic seek/sequence parameter calculation to improve sound quality</li>
543     <li>Bugfixes:&nbsp;
544     <ul>
545     <li>Fixed negative array indexing in quick seek algorithm</li>
546     <li>FIR autoalias filter running too far in processing buffer</li>
547     <li>Check against zero sample count in rate transposing</li>
548     <li>Fix for x86-64 support: Removed pop/push instructions from the cpu detection algorithm.&nbsp;</li>
549     <li>Check against empty buffers in FIFOSampleBuffer</li>
550     <li>Other minor fixes &amp; code cleanup</li>
551     </ul>
552     </li>
554     <li>Fixes in compilation scripts for non-Intel platforms</li>
555     <li>Added Dynamic-Link-Library (DLL) version of SoundTouch library build,
556     provided with Delphi/Pascal wrapper for calling the dll routines</li>
557     <li>Added #define PREVENT_CLICK_AT_RATE_CROSSOVER that prevents a click artifact
558     when crossing the nominal pitch from either positive to negative side or vice
559     versa</li>
561     </ul>
563     <p><strong>1.4.1:</strong></p>
564     <ul>
565     <li>Fixed a buffer overflow bug in BPM detect algorithm routines if processing
566     more than 2048 samples at one call&nbsp;</li>
568     </ul>
570     <p><strong>1.4.0:</strong></p>
571     <ul>
572     <li>Improved sound quality by automatic calculation of time stretch algorithm
573     processing parameters according to tempo setting</li>
574     <li>Moved BPM detection routines from SoundStretch application into SoundTouch
575     library</li>
576     <li>Bugfixes: Usage of uninitialied variables, GNU build scripts, compiler errors
577     due to 'const' keyword mismatch.</li>
578     <li>Source code cleanup</li>
580     </ul>
582     <p><strong>v1.3.1:
583     </strong></p>
584     <ul>
585     <li>Changed static class declaration to GCC 4.x compiler compatible syntax.</li>
586     <li>Enabled MMX/SSE-optimized routines also for GCC compilers. Earlier
587     the MMX/SSE-optimized routines were written in compiler-specific inline
588     assembler, now these routines are migrated to use compiler intrinsic
589     syntax which allows compiling the same MMX/SSE-optimized source code with
590     both Visual C++ and GCC compilers. </li>
591     <li>Set floating point as the default sample format and added switch to
592     the GNU configure script for selecting the other sample format.</li>
594     </ul>
596     <p><strong>v1.3.0:
597     </strong></p>
598     <ul>
599     <li>Fixed tempo routine output duration inaccuracy due to rounding
600     error </li>
601     <li>Implemented separate processing routines for integer and
602     floating arithmetic to allow improvements to floating point routines
603     (earlier used algorithms mostly optimized for integer arithmetic also
604     for floating point samples) </li>
605     <li>Fixed a bug that distorts sound if sample rate changes during the
606     sound stream </li>
607     <li>Fixed a memory leak that appeared in MMX/SSE/3DNow! optimized
608     routines </li>
609     <li>Reduced redundant code pieces in MMX/SSE/3DNow! optimized
610     routines vs. the standard C routines.</li>
611     <li>MMX routine incompatibility with new gcc compiler versions </li>
612     <li>Other miscellaneous bug fixes </li>
613     </ul>
614     <p><strong>v1.2.1: </strong></p>
615     <ul>
616     <li>Added automake/autoconf scripts for GNU
617     platforms (in courtesy of David Durham)</li>
618     <li>Fixed SCALE overflow bug in rate transposer
619     routine.</li>
620     <li>Fixed 64bit address space bugs.</li>
621     <li>Created a 'soundtouch' namespace for
622     SAMPLETYPE definitions.</li>
623     </ul>
624     <p><strong>v1.2.0: </strong></p>
625     <ul>
626     <li>Added support for 32bit floating point sample
627     data type with SSE/3DNow! optimizations for Win32 platform (SSE/3DNow! optimizations currently not supported in GCC environment)</li>
628     <li>Replaced 'make-gcc' script for GNU environment
629     by master Makefile</li>
630     <li>Added time-stretch routine configurability to
631     SoundTouch main class</li>
632     <li>Bugfixes</li>
633     </ul>
634     <p><strong>v1.1.1: </strong></p>
635     <ul>
636     <li>Moved SoundTouch under lesser GPL license (LGPL). This allows using SoundTouch library in programs that aren't
637     released under GPL license. </li>
638     <li>Changed MMX routine organiation so that MMX optimized routines are now implemented in classes that are derived from
639     the basic classes having the standard non-mmx routines. </li>
640     <li>MMX routines to support gcc version 3. </li>
641     <li>Replaced windows makefiles by script using the .dsw files </li>
642     </ul>
643     <p><strong>v1.01: </strong></p>
644     <ul>
645     <li>&quot;mmx_gcc.cpp&quot;: Added "using namespace std" and
646     removed "return 0" from a function with void return value to fix
647     compiler errors when compiling the library in Solaris environment. </li>
648     <li>Moved file &quot;FIFOSampleBuffer.h&quot; to "include"
649     directory to allow accessing the FIFOSampleBuffer class from external
650     files. </li>
651     </ul>
652     <p><strong>v1.0: </strong></p>
653     <ul>
654     <li>Initial release </li>
655     </ul>
656     <p>&nbsp;</p>
657     <h3>5.2. SoundStretch application Change
658     History </h3>
660     <p><strong>1.4.0:</strong></p>
661     <ul>
662     <li>Moved BPM detection routines from SoundStretch application into SoundTouch
663     library</li>
664     <li>Allow using standard input/output pipes as audio processing input/output
665     streams</li>
667     </ul>
669     <p><strong>v1.3.0:</strong></p>
670     <ul>
671     <li>Simplified accessing WAV files with floating
672     point sample format.
673     </li>
674     </ul>
675     <p><strong>v1.2.1: </strong></p>
676     <ul>
677     <li>Fixed 64bit address space bugs.</li>
678     </ul>
679     <p><strong>v1.2.0: </strong></p>
680     <ul>
681     <li>Added support for 32bit floating point sample
682     data type</li>
683     <li>Restructured the BPM routines into separate
684     library</li>
685     <li>Fixed big-endian conversion bugs in WAV file
686     routines (hopefully :)</li>
687     </ul>
688     <p><strong>v1.1.1: </strong></p>
689     <ul>
690     <li>Fixed bugs in WAV file reading &amp; added
691     byte-order conversion for big-endian processors. </li>
692     <li>Moved SoundStretch source code under 'example'
693     directory to highlight difference from SoundTouch stuff. </li>
694     <li>Replaced windows makefiles by script using the .dsw files </li>
695     <li>Output file name isn't required if output
696     isn't desired (e.g. if using the switch '-bpm' in plain format only) </li>
697     </ul>
698     <p><strong>v1.1:</strong></p>
699     <ul>
700     <li>Fixed "Release" settings in Microsoft Visual
701     C++ project file (.dsp) </li>
702     <li>Added beats-per-minute (BPM) detection routine
703     and command-line switch &quot;-bpm&quot; </li>
704     </ul>
705     <p><strong>v1.01: </strong></p>
706     <ul>
707     <li>Initial release </li>
708     </ul>
709     <hr>
710     <h2 >6. Acknowledgements </h2>
711     <p >Kudos for these people who have contributed to development or submitted
712     bugfixes since
713     SoundTouch v1.3.1: </p>
714     <ul>
715     <li>Arthur A</li>
716     <li>Richard Ash</li>
717     <li>Stanislav Brabec</li>
718     <li>Christian Budde</li>
719     <li>Brian Cameron</li>
720     <li>Jason Champion</li>
721     <li>Patrick Colis</li>
722     <li>Justin Frankel</li>
723     <li>Jason Garland</li>
724     <li>Takashi Iwai</li>
725     <li>Paulo Pizarro</li>
726     <li>RJ Ryan</li>
727     <li>John Sheehy</li>
728     </ul>
729     <p >Moral greetings to all other contributors and users also!</p>
730     <hr>
731     <h2 >7. LICENSE </h2>
732     <p>SoundTouch audio processing library<br>
733     Copyright (c) Olli Parviainen</p>
734     <p>This library is free software; you can
735     redistribute it and/or modify it under the terms of the GNU
736     Lesser General Public License version 2.1 as published by the Free Software
737     Foundation.</p>
738     <p>This library is distributed in the hope
739     that it will be useful, but WITHOUT ANY WARRANTY; without even
740     the implied warranty of MERCHANTABILITY or FITNESS FOR A
741     PARTICULAR PURPOSE. See the GNU Lesser General Public License for
742     more details.</p>
743     <p>You should have received a copy of the GNU
744     Lesser General Public License along with this library; if not,
745     write to the Free Software Foundation, Inc., 59 Temple Place,
746     Suite 330, Boston, MA 02111-1307 USA</p>
747     <hr>
748     <!--
749     $Id: README.html 81 2009-12-28 20:51:18Z oparviai $
750     -->
751     </body>
752     </html>

  ViewVC Help
Powered by ViewVC 1.1.22