资源说明：Hacked version of SoundStretch which allows STDIN and STDOUT by specifying files as '-'

  
  
  
  
  
  
  
  



SoundTouch audio processing library v1.3.2pre

SoundTouch library Copyright (c) Olli
Parviainen 2002-2008 

1. Introduction 
SoundTouch is an open-source audio
processing library that allows changing the sound tempo, pitch
and playback rate parameters independently from each other, i.e.:

  
Sound tempo can be increased or decreased while
maintaining the original pitch
  
Sound pitch can be increased or decreased while
maintaining the original tempo 
  
Change playback rate that affects both tempo
and pitch at the same time 
  
Choose any combination of tempo/pitch/rate

1.1 Contact information 
Author email: oparviai 'at' iki.fi 
SoundTouch WWW page: http://www.surina.net/soundtouch

2. Compiling SoundTouch
Before compiling, notice that you can choose the sample data format
if it's desirable to use floating point sample
data instead of 16bit integers. See section "sample data format"
for more information.
2.1. Building in Microsoft Windows
Project files for Microsoft Visual C++ 6.0 and Visual C++ .NET are
supplied with the source code package. Please notice that SoundTouch
library uses processor-specific optimiations for Pentium III and AMD
processors that require a processor pack upgrade for
the Visual Studio 6.0 to be installed in order to support these
optimiations. The processor pack upgrade can be downloaded from
Microsoft site at this URL:

http://msdn.microsoft.com/vstudio/downloads/tools/ppack/default.aspx
If the above URL is unavailable or removed, go
to http://msdn.microsoft.com
and perform a search with keywords processor pack. 
Visual Studio .NET supports required
instructions by default and thus doesn't require installing the
processor pack.
To build the binaries with Visual C++ 6.0
compiler, either run "make-win.bat" script or open the
appropriate project files in source code directories with Visual
Studio. The final executable will appear under the "SoundTouch\bin"
directory. If using the Visual Studio IDE instead of the
make-win.bat script, directories bin and
lib have to be created manually to the SoundTouch
package root for the final executables. The make-win.bat script
creates these directories automatically.

Also other C++ compilers than Visual C++ can be
used, but project or makefiles then have to be adapted accordingly.
Performance optimiations are written in Visual C++ compatible
syntax, they may or may not be compatible with other compilers. If
using GCC (Gnu C Compiler) compiler package such as DJGPP or Cygwin,
please see next chapter for instructions. 
2.2. Building in Gnu platforms
The SoundTouch library can be compiled in
practically any platform supporting GNU compiler (GCC) tools.
SoundTouch have been tested with gcc version 3.3.4., but it
shouldn't be very specific about the gcc version. Assembler-level
performance optimiations for GNU platform are currently available in
x86 platforms only, they are automatically disabled and replaced with
standard C routines in other processor platforms.
To build and install the binaries, run the
following commands in SoundTouch/ directory:

  

    

      

      
./configure  -
      
      

      
Configures the SoundTouch package for the local
environment.
      
    
    

      

      
make         -
      
      

      
Builds the SoundTouch library &
SoundStretch utility.
      
    
    

      

      
make install -
      
      

      
Installs the SoundTouch & BPM libraries
to /usr/local/lib and SoundStretch utility to /usr/local/bin.
Please notice that 'root' privileges may be required to install the
binaries to the destination locations.
      
    
  

NOTE: At the time of release the SoundTouch package has been
tested
to compile in GNU/Linux platform. However, in past it's happened that
new
gcc versions aren't necessarily compatible with the assembler setttings
used in the optimied routines. If you have problems getting the
SoundTouch library compiled, try the workaround of disabling the
optimiations by editing the file "include/STTypes.h" and removing
the following definition there:

  
#define ALLOW_OPTIMIZATIONS 1


3. About implementation & Usage tips
3.1. Supported sample data formats
The sample data format can be chosen
between 16bit signed integer and 32bit floating point values, the
default is 32bit floating point. 


In Windows environment, the sample data format is chosen
in file "STTypes.h" by choosing one of the following
defines:

  
#define INTEGER_SAMPLES
for 16bit signed
integer
  
#define FLOAT_SAMPLES for
32bit floating point


In GNU environment, the floating sample format is used by default, but 
integer sample format can be chosen by giving the 
following switch to the configure script:

./configure --enable-integer-samples


The sample data can have either single (mono)
or double (stereo) audio channel. Stereo data is interleaved so
that every other data value is for left channel and every second
for right channel. Notice that while it'd be possible in theory
to process stereo sound as two separate mono channels, this isn't
recommended because processing the channels separately would
result in losing the phase coherency between the channels, which
consequently would ruin the stereo effect.
Sample rates between 8000-48000H are
supported.
3.2. Processing latency
The processing and latency constraints of
the SoundTouch library are:

  
Input/output processing latency for the
SoundTouch processor is around 100 ms. This is when time-stretching is
used. If the rate transposing effect alone is used, the latency
requirement
is much shorter, see section 'About algorithms'.
  
Processing CD-quality sound (16bit stereo
sound with 44100H sample rate) in real-time or faster is possible
starting from processors equivalent to Intel Pentium 133Mh or better,
if using the "quick" processing algorithm. If not using the "quick"
mode or
if floating point sample data are being used, several times more CPU
power is typically required.

3.3. About algorithms
SoundTouch provides three seemingly
independent effects: tempo, pitch and playback rate control.
These three controls are implemented as combination of two primary
effects, sample rate transposing and time-stretching.
Sample rate transposing affects
both the audio stream duration and pitch. It's implemented simply
by converting the original audio sample stream to the  desired
duration by interpolating from the original audio samples. In
SoundTouch, linear interpolation with anti-alias filtering is
used. Theoretically a higher-order interpolation provide better
result than 1st order linear interpolation, but in audio
application linear interpolation together with anti-alias
filtering performs subjectively about as well as higher-order
filtering would.
Time-stretching means changing
the audio stream duration without affecting it's pitch. SoundTouch
uses WSOLA-like time-stretching routines that operate in the time
domain. Compared to sample rate transposing, time-stretching is a
much heavier operation and also requires a longer processing
"window" of sound samples used by the
processing algorithm, thus increasing the algorithm input/output
latency. Typical i/o latency for the SoundTouch
time-stretch algorithm is around 100 ms.
Sample rate transposing and time-stretching
are then used together to produce the tempo, pitch and rate
controls:

  
'Tempo' control is
implemented purely by time-stretching.
  
'Rate' control is implemented
purely by sample rate transposing.
  
'Pitch' control is
implemented as a combination of time-stretching and sample rate
transposing. For example, to increase pitch the audio stream is first
time-stretched to longer duration (without affecting pitch) and then
transposed back to original duration by sample rate transposing, which
simultaneously reduces duration and increases pitch. The result is
original duration but increased pitch.

3.4 Tuning the algorithm parameters
The time-stretch algorithm has few
parameters that can be tuned to optimie sound quality for
certain application. The current default parameters have been
chosen by iterative if-then analysis (read: "trial and error")
to obtain best subjective sound quality in pop/rock music
processing, but in applications processing different kind of
sound the default parameter set may result into a sub-optimal
result.
The time-stretch algorithm default
parameter values are set by these #defines in file "TDStretch.h":

  
#define DEFAULT_SEQUENCE_MS     82
#define DEFAULT_SEEKWINDOW_MS   28
#define DEFAULT_OVERLAP_MS      12

These parameters affect to the time-stretch
algorithm as follows:

  
DEFAULT_SEQUENCE_MS: This is
the default length of a single processing sequence in milliseconds
which determines the how the original sound is chopped in
the time-stretch algorithm. Larger values mean fewer sequences
are used in processing. In principle a larger value sounds better when
slowing down the tempo, but worse when increasing the tempo and vice
versa.

  
  
DEFAULT_SEEKWINDOW_MS: The seeking window
default length in milliseconds is for the algorithm that seeks the best
possible overlapping location. This determines from how
wide a sample "window" the algorithm can use to find an optimal mixing
location when the sound sequences are to be linked back together.

    

The bigger this window setting is, the higher the possibility to find a
better mixing position becomes, but at the same time large values may
cause a "drifting" sound artifact because neighboring sequences can be
chosen at more uneven intervals. If there's a disturbing artifact that
sounds as if a constant frequency was drifting around, try reducing
this setting.

  
  
DEFAULT_OVERLAP_MS: Overlap
length in milliseconds. When the sound sequences are mixed back
together to form again a continuous sound stream, this parameter
defines how much the ends of the consecutive sequences will overlap with each other.

    

    This shouldn't be that critical parameter. If you reduce the
DEFAULT_SEQUENCE_MS setting by a large amount, you might wish to try a
smaller value on this.

Notice that these parameters can also be
set during execution time with functions "TDStretch::setParameters()"
and "SoundTouch::setSetting()".
The table below summaries how the
parameters can be adjusted for different applications:

  

    

      
Parameter name
      
Default value
magnitude
      
Larger value
affects...
      
Smaller value
affects...
      
Music
      
Speech
      
Effect in CPU burden
    
    

      

      
SEQUENCE_MS
      
      
Default value is relatively
large, chosen for slowing down music tempo
      
Larger value is usually
better for slowing down tempo. Growing the value decelerates the
"echoing" artifact when slowing down the tempo.
      
Smaller value might be better
for speeding up tempo. Reducing the value accelerates the "echoing"
artifact when slowing down the tempo 
      
Default value usually good
      
A smaller value than default
might be better
      
Increasing the parameter
value reduces computation burden
    
    

      

      
SEEKWINDOW_MS
      
      
Default value is relatively
large, chosen for slowing down music tempo
      
Larger value eases finding a
good mixing position, but may cause a "drifting" artifact
      
Smaller reduce possibility to
find a good mixing position, but reduce the "drifting" artifact.
      
Default value usually good,
unless a "drifting" artifact is disturbing.
      
Default value usually good
      
Increasing the parameter
value increases computation burden
    
    

      

      
OVERLAP_MS
      
      
Default value is relatively
large, chosen to suit with above parameters.
      
 
      
If you reduce the "sequence
ms" setting, you might wish to try a smaller value.
      
 
      
 
      
Increasing the parameter
value increases computation burden
    
  

3.5 Performance Optimiations 
General optimiations:
The time-stretch routine has a 'quick' mode
that substantially speeds up the algorithm but may degrade the
sound quality by a small amount. This mode is activated by
calling SoundTouch::setSetting() function with parameter  id
of SETTING_USE_QUICKSEEK and value "1", i.e. 

  
setSetting(SETTING_USE_QUICKSEEK, 1);

CPU-specific optimiations:

  
Intel MMX optimied routines are used with
compatible CPUs when 16bit integer sample type is used. MMX
optimiations are available both in Win32 and Gnu/x86 platforms.
Compatible processors are Intel PentiumMMX and later; AMD K6-2, Athlon
and later. 
  
Intel SSE optimied routines are used with
compatible CPUs when floating point sample type is used. SSE
optimiations are currently implemented for Win32 platform only.
Processors compatible with SSE extension are Intel processors starting
from Pentium-III, and AMD processors starting from Athlon XP. 
  
AMD 3DNow! optimied routines are used with
compatible CPUs when floating point sample type is used, but SSE
extension isn't supported . 3DNow! optimiations are currently
implemented for Win32 platform only. These optimiations are used in
AMD K6-2 and Athlon (classic) CPU's; better performing SSE routines are
used with AMD processor starting from Athlon XP. 


4. SoundStretch audio processing utility

SoundStretch audio processing utility

Copyright (c) Olli Parviainen 2002-2005
SoundStretch is a simple command-line
application that can change tempo, pitch and playback rates of
WAV sound files. This program is intended primarily to
demonstrate how the "SoundTouch" library can be used to
process sound in your own program, but it can as well be used for
processing sound files.
4.1. SoundStretch Usage Instructions
SoundStretch Usage syntax:

  
soundstretch infile.wav outfile.wav [switches]

Where: 

  

    

      

      
"infile.wav"
      
      
is the name of the input sound
data file (in .WAV audio file format). 
    
    

      

      
"outfile.wav"
      
      
is the name of the output sound
file where the resulting sound is saved (in .WAV audio file format).
This parameter may be omitted if you  don't want to save the
output
(e.g. when only calculating BPM rate with '-bpm' switch).
    
    

      

      
 [switches]
      
      
Are one or more control
switches.
    
  

Available control switches are:

  

    

      

      
-tempo=n 
      
      
Change the sound tempo by n
percents (n = -95.0 .. +5000.0 %) 
    
    

      

      
-pitch=n
      
      
Change the sound pitch by n
semitones (n = -60.0 .. + 60.0 semitones) 
    
    

      

      
-rate=n
      
      
Change the sound playback rate by
n percents (n = -95.0 .. +5000.0 %) 
    
    

      

      
-bpm=n
      
      
Detect the Beats-Per-Minute
(BPM) rate of the sound and adjust the tempo to meet 'n' BPMs. If this
switch is defined, the "-tempo=n" switch value is ignored. If "=n" is
omitted, i.e. switch "-bpm" is used alone, the program just calculates
and displays the BPM rate but doesn't adjust tempo according to the BPM
value. 
    
    

      

      
-quick
      
      
Use quicker tempo change
algorithm. Gains speed but loses sound quality. 
    
    

      

      
-naa
      
      
Don't use anti-alias
filtering in sample rate transposing. Gains speed but loses sound
quality. 
    
    

      

      
-license
      
      
Displays the program license
text (LGPL)
    
  

Notes:

  
The numerical switch values can be entered
using either integer (e.g. "-tempo=123") or decimal (e.g.
"-tempo=123.45") numbers.
  
The "-naa" and/or "-quick" switches can be
used to reduce CPU usage while compromising some sound quality 
  
The BPM detection algorithm works by detecting
repeating low-frequency (<250H) sound patterns and thus works
mostly with most rock/pop music with bass or drum beat. The BPM
detection doesn't work on pieces such as classical music without
distinct, repeating bass frequency patterns. Also pieces with varying
tempo, varying bass patterns or very complex bass patterns (ja, hiphop) may produce odd BPM readings. 

    

In cases when the bass pattern drifts a bit around a nominal beat rate
(e.g. drummer is again drunken :), the BPM algorithm may report
incorrect harmonic one-halft to one-thirdth of the correct BPM value;
in such case the system could for example report BPM value of 50 or 100
instead of correct BPM value of 150. 

4.2. SoundStretch usage examples 
Example 1
The following command increases tempo of
the sound file "originalfile.wav" by 12.5% and saves
result to file "destinationfile.wav":

  
soundstretch originalfile.wav destinationfile.wav -tempo=12.5

Example 2
The following command decreases the sound
pitch (key) of the sound file "orig.wav" by two
semitones and saves the result to file "dest.wav":

  
soundstretch orig.wav dest.wav -pitch=-2

Example 3
The following command processes the file "orig.wav" by decreasing the sound tempo by 25.3% and
increasing the sound pitch (key) by 1.5 semitones. Result is
saved to file "dest.wav":

  
soundstretch orig.wav dest.wav -tempo=-25.3 -pitch=1.5

Example 4
The following command detects the BPM rate
of the file "orig.wav" and adjusts the tempo to match
100 beats per minute. Result is saved to file "dest.wav":

  
soundstretch orig.wav dest.wav -bpm=100


5. Change History
5.1. SoundTouch library Change History 

v1.3.2:

Bugfixes: Using uninitialied variables, GNU build scripts, compiler errors
  due to 'const' keyword mismatch.
Some source code cleanup



v1.3.1:


Changed static class declaration to GCC 4.x compiler compatible syntax.
Enabled MMX/SSE-optimied routines also for GCC compilers. Earlier
the MMX/SSE-optimied routines were written in compiler-specific inline 
assembler, now these routines are migrated to use compiler intrinsic 
syntax which allows compiling the same MMX/SSE-optimied source code with 
both Visual C++ and GCC compilers. 
Set floating point as the default sample format and added switch to 
the GNU configure script for selecting the other sample format.



v1.3.0:


  
Fixed tempo routine output duration inaccuracy due to rounding
error 
  
Implemented separate processing routines for integer and
floating arithmetic to allow improvements to floating point routines
(earlier used algorithms mostly optimied for integer arithmetic also
for floating point samples) 
  
Fixed a bug that distorts sound if sample rate changes during the
sound stream 
  
Fixed a memory leak that appeared in MMX/SSE/3DNow! optimied
routines 
  
Reduced redundant code pieces in MMX/SSE/3DNow! optimied
routines vs. the standard C routines.
  
MMX routine incompatibility with new gcc compiler versions 
  
Other miscellaneous bug fixes 

v1.2.1: 

  
Added automake/autoconf scripts for GNU
platforms (in courtesy of David Durham)
  
Fixed SCALE overflow bug in rate transposer
routine.
  
Fixed 64bit address space bugs.
  
Created a 'soundtouch' namespace for
SAMPLETYPE definitions.

v1.2.0: 

  
Added support for 32bit floating point sample
data type with SSE/3DNow! optimiations for Win32 platform (SSE/3DNow!
optimiations currently not supported in GCC environment)
  
Replaced 'make-gcc' script for GNU environment
by master Makefile
  
Added time-stretch routine configurability to
SoundTouch main class
  
Bugfixes

v1.1.1: 

  
Moved SoundTouch under lesser GPL license (LGPL). This allows using SoundTouch library in programs that aren't
released under GPL license. 
  
Changed MMX routine organiation so that MMX
optimied routines are now implemented in classes that are derived from
the basic classes having the standard non-mmx routines. 
  
MMX routines to support gcc version 3. 
  
Replaced windows makefiles by script using the .dsw files 

v1.01: 

  
"mmx_gcc.cpp": Added "using namespace std" and
removed "return 0" from a function with void return value to fix
compiler errors when compiling the library in Solaris environment. 
  
Moved file "FIFOSampleBuffer.h" to "include"
directory to allow accessing the FIFOSampleBuffer class from external
files. 

v1.0: 

  
Initial release 

 
5.2. SoundStretch application Change
History 
v1.3.0:

  
Simplified accessing WAV files with floating
point sample format.

  

v1.2.1: 

  
Fixed 64bit address space bugs.

v1.2.0: 

  
Added support for 32bit floating point sample
data type
  
Restructured the BPM routines into separate
library
  
Fixed big-endian conversion bugs in WAV file
routines (hopefully :)

v1.1.1: 

  
Fixed bugs in WAV file reading & added
byte-order conversion for big-endian processors. 
  
Moved SoundStretch source code under 'example'
directory to highlight difference from SoundTouch stuff. 
  
Replaced windows makefiles by script using the .dsw files 
  
Output file name isn't required if output
isn't desired (e.g. if using the switch '-bpm' in plain format only) 

v1.1:

  
Fixed "Release" settings in Microsoft Visual
C++ project file (.dsp) 
  
Added beats-per-minute (BPM) detection routine
and command-line switch "-bpm" 

v1.01: 

  
Initial release 


6. Acknowledgements 
Kudos for these people who have submitted bugfixed since
SoundTouch v1.3.1: 

  
Arthur A.: Bugfix
  
Stanislav Brabec / Takashi Iwai
  
Jason Garland

Moral greetings to all earlier contributors as well!

7. LICENSE 
SoundTouch audio processing library

Copyright (c) Olli Parviainen
This library is free software; you can
redistribute it and/or modify it under the terms of the GNU
Lesser General Public License as published by the Free Software
Foundation; either version 2.1 of the License, or (at your option)
any later version.
This library is distributed in the hope
that it will be useful, but WITHOUT ANY WARRANTY; without even
the implied warranty of MERCHANTABILITY or FITNESS FOR A
PARTICULAR PURPOSE. See the GNU Lesser General Public License for
more details.
You should have received a copy of the GNU
Lesser General Public License along with this library; if not,
write to the Free Software Foundation, Inc., 59 Temple Place,
Suite 330, Boston, MA 02111-1307 USA

联系我们：verysource_com

CopyRight © 2008-2022 verySource.Com All Rights reserved.　京ICP备17048824号-1　京公网安备：11010502034788