Arduino STK 4.6.2
stk::LentPitShift Class Reference

Pitch shifter effect class based on the Lent algorithm. More...

#include <LentPitShift.h>

Inheritance diagram for stk::LentPitShift:
stk::Effect stk::Stk

Public Member Functions

 LentPitShift (StkFloat periodRatio=1.0, int tMax=RT_BUFFER_SIZE)
 Class constructor.
void clear (void)
 Reset and clear all internal state.
void setShift (StkFloat shift)
 Set the pitch shift factor (1.0 produces no shift).
StkFloat tick (StkFloat input)
 Input one sample to the filter and return one output.
StkFrames & tick (StkFrames &frames, unsigned int channel=0)
 Take a channel of the StkFrames object as inputs to the filter and replace with corresponding outputs.
StkFrames & tick (StkFrames &iFrames, StkFrames &oFrames, unsigned int iChannel=0, unsigned int oChannel=0)
 Take a channel of the iFrames object as inputs to the filter and write outputs to the oFrames object.
unsigned int channelsOut (void) const
 Return the number of output channels for the class.
const StkFrames & lastFrame (void) const
 Return an StkFrames reference to the last output sample frame.
virtual void setEffectMix (StkFloat mix)
 Set the mixture of input and "effected" levels in the output (0.0 = input only, 1.0 = effect only).
void ignoreSampleRateChange (bool ignore=true)
 A function to enable/disable the automatic updating of class data when the STK sample rate changes.

Static Public Member Functions

static StkFloat sampleRate (void)
 Static method that returns the current STK sample rate.
static void setSampleRate (StkFloat rate)
 Static method that sets the STK sample rate.
static void clear_alertList ()
 Static method that frees memory from alertList_.
static std::string rawwavePath (void)
 Static method that returns the current rawwave path.
static void setRawwavePath (std::string path)
 Static method that sets the STK rawwave path.
static void swap16 (unsigned char *ptr)
 Static method that byte-swaps a 16-bit data type.
static void swap32 (unsigned char *ptr)
 Static method that byte-swaps a 32-bit data type.
static void swap64 (unsigned char *ptr)
 Static method that byte-swaps a 64-bit data type.
static void sleep (unsigned long milliseconds)
 Static cross-platform method to sleep for a number of milliseconds.
static bool inRange (StkFloat value, StkFloat min, StkFloat max)
 Static method to check whether a value is within a specified range.
static void handleError (const char *message, StkError::Type type)
 Static function for error reporting and handling using c-strings.
static void handleError (std::string message, StkError::Type type)
 Static function for error reporting and handling using c++ strings.
static void showWarnings (bool status)
 Toggle display of WARNING and STATUS messages.
static void printErrors (bool status)
 Toggle display of error messages before throwing exceptions.

Static Public Attributes

static const StkFormat STK_SINT8
static const StkFormat STK_SINT16
static const StkFormat STK_SINT24
static const StkFormat STK_SINT32
static const StkFormat STK_FLOAT32
static const StkFormat STK_FLOAT64

Protected Member Functions

void process ()
 Apply the effect on the input samples and store it.
void handleError (StkError::Type type) const
 Internal function for error reporting that assumes message in oStream_ variable.
virtual void sampleRateChanged (StkFloat newRate, StkFloat oldRate)
 This function should be implemented in subclasses that depend on the sample rate.
void addSampleRateAlert (Stk *ptr)
 Add class pointer to list for sample rate change notification.
void removeSampleRateAlert (Stk *ptr)
 Remove class pointer from list for sample rate change notification.

Detailed Description

Pitch shifter effect class based on the Lent algorithm.

This class implements a pitch shifter using pitch tracking and sample windowing and shifting.

by Francois Germain, 2009.

Member Function Documentation

◆ tick() [1/2]

StkFrames & stk::LentPitShift::tick ( StkFrames & frames,
unsigned int channel = 0 )
inline

Take a channel of the StkFrames object as inputs to the filter and replace with corresponding outputs.

The StkFrames argument reference is returned. The channel argument must be less than the number of channels in the StkFrames argument (the first channel is specified by 0). However, range checking is only performed if STK_DEBUG is defined during compilation, in which case an out-of-range value will trigger an StkError exception.

227{
228#if defined(_STK_DEBUG_)
229 if ( channel >= frames.channels() ) {
230 oStream_ << "LentPitShift::tick(): channel and StkFrames arguments are incompatible!";
231 handleError( StkError::FUNCTION_ARGUMENT );
232 }
233#endif
234
235 StkFloat *samples = &frames[channel];
236 unsigned int hop = frames.channels();
237 for ( unsigned int i=0; i<frames.frames(); i++, samples += hop ) {
238 *samples = tick( *samples );
239 }
240
241 return frames;
242}
StkFloat tick(StkFloat input)
Input one sample to the filter and return one output.
Definition LentPitShift.h:209
static void handleError(const char *message, StkError::Type type)
Static function for error reporting and handling using c-strings.

◆ tick() [2/2]

StkFrames & stk::LentPitShift::tick ( StkFrames & iFrames,
StkFrames & oFrames,
unsigned int iChannel = 0,
unsigned int oChannel = 0 )
inline

Take a channel of the iFrames object as inputs to the filter and write outputs to the oFrames object.

The iFrames object reference is returned. Each channel argument must be less than the number of channels in the corresponding StkFrames argument (the first channel is specified by 0). However, range checking is only performed if STK_DEBUG is defined during compilation, in which case an out-of-range value will trigger an StkError exception.

245{
246#if defined(_STK_DEBUG_)
247 if ( iChannel >= iFrames.channels() || oChannel >= oFrames.channels() ) {
248 oStream_ << "LentPitShift::tick(): channel and StkFrames arguments are incompatible!";
249 handleError( StkError::FUNCTION_ARGUMENT );
250 }
251#endif
252
253 StkFloat *iSamples = &iFrames[iChannel];
254 StkFloat *oSamples = &oFrames[oChannel];
255 unsigned int iHop = iFrames.channels(), oHop = oFrames.channels();
256 for ( unsigned int i=0; i<iFrames.frames(); i++, iSamples += iHop, oSamples += oHop ) {
257 *oSamples = tick( *iSamples );
258 }
259
260 return iFrames;
261}

◆ process()

void stk::LentPitShift::process ( )
inlineprotected

Apply the effect on the input samples and store it.

The samples stored in the input frame vector are processed and the delayed result are stored in the output frame vector.

113{
114 StkFloat x_t; // input coefficient
115 StkFloat x_t_T; // previous input coefficient at T samples
116 StkFloat coeff; // new coefficient for the difference function
117
118 unsigned long alternativePitch = tMax_; // Global minimum storage
119 lastPeriod_ = tMax_+1; // Storage of the lowest local minimum under the threshold
120
121 // Loop variables
122 unsigned long delay_;
123 unsigned int n;
124
125 // Initialization of the dt coefficients. Since the
126 // frames are of tMax_ length, there is no overlapping
127 // between the successive windows where pitch tracking
128 // is performed.
129 for ( delay_=1; delay_<=tMax_; delay_++ )
130 dt[delay_] = 0.;
131
132 // Calculation of the dt coefficients and update of the input delay line.
133 for ( n=0; n<inputFrames.size(); n++ ) {
134 x_t = inputLine_.tick( inputFrames[ n ] );
135 for ( delay_=1; delay_<= tMax_; delay_++ ) {
136 x_t_T = inputLine_.tapOut( delay_ );
137 coeff = x_t - x_t_T;
138 dt[delay_] += coeff * coeff;
139 }
140 }
141
142 // Calculation of the pitch tracking function and test for the minima.
143 for ( delay_=1; delay_<=tMax_; delay_++ ) {
144 cumDt[delay_] = dt[delay_] + cumDt[delay_-1];
145 dpt[delay_] = dt[delay_] * delay_ / cumDt[delay_];
146
147 // Look for a minimum
148 if ( dpt[delay_-1]-dpt[delay_-2] < 0 && dpt[delay_]-dpt[delay_-1] > 0 ) {
149 // Check if the minimum is under the threshold
150 if ( dpt[delay_-1] < threshold_ ){
151 lastPeriod_ = delay_-1;
152 // If a minimum is found, we can stop the loop
153 break;
154 }
155 else if ( dpt[alternativePitch] > dpt[delay_-1] )
156 // Otherwise we store it if it is the current global minimum
157 alternativePitch = delay_-1;
158 }
159 }
160
161 // Test for the last period length.
162 if ( dpt[delay_]-dpt[delay_-1] < 0 ) {
163 if ( dpt[delay_] < threshold_ )
164 lastPeriod_ = delay_;
165 else if ( dpt[alternativePitch] > dpt[delay_] )
166 alternativePitch = delay_;
167 }
168
169 if ( lastPeriod_ == tMax_+1 )
170 // No period has been under the threshold so we used the global minimum
171 lastPeriod_ = alternativePitch;
172
173 // We put the new zero output coefficients in the output delay line and
174 // we get the previous calculated coefficients
175 outputLine_.tick( zeroFrame, outputFrames );
176
177 // Initialization of the Hamming window used in the algorithm
178 for ( int n=-(int)lastPeriod_; n<(int)lastPeriod_; n++ )
179 window[n+lastPeriod_] = (1 + std::cos(STK_PI*n/lastPeriod_)) / 2 ;
180
181 long M; // Index of reading in the input delay line
182 long N; // Index of writing in the output delay line
183 StkFloat sample; // Temporary storage for the new coefficient
184
185 // We loop for all the frames of length lastPeriod_ presents between inputPtr and tMax_
186 for ( ; inputPtr<(int)(tMax_-lastPeriod_); inputPtr+=lastPeriod_ ) {
187 // Test for the decision of compression/expansion
188 while ( outputPtr < inputPtr ) {
189 // Coefficients for the linear interpolation
190 env[1] = std::fmod( outputPtr + tMax_, 1.0 );
191 env[0] = 1.0f - env[1];
192 M = tMax_ - inputPtr + lastPeriod_ - 1; // New reading pointer
193 N = 2*tMax_ - (unsigned long)std::floor(outputPtr + tMax_) + lastPeriod_ - 1; // New writing pointer
194 for ( unsigned int j=0; j<2*lastPeriod_; j++,M--,N-- ) {
195 sample = inputLine_.tapOut(M) * window[j] / 2.f;
196 // Linear interpolation
197 outputLine_.addTo(env[0] * sample, N);
198 outputLine_.addTo(env[1] * sample, N-1);
199 }
200 outputPtr = outputPtr + lastPeriod_ * periodRatio_; // new output pointer
201 }
202 }
203 // Shifting of the pointers waiting for the new frame of length tMax_.
204 outputPtr -= tMax_;
205 inputPtr -= tMax_;
206}

◆ setSampleRate()

void stk::Stk::setSampleRate ( StkFloat rate)
staticinherited

Static method that sets the STK sample rate.

The sample rate set using this method is queried by all STK classes that depend on its value. It is initialized to the default SRATE set in Stk.h. Many STK classes use the sample rate during instantiation. Therefore, if you wish to use a rate that is different from the default rate, it is imperative that it be set BEFORE STK objects are instantiated. A few classes that make use of the global STK sample rate are automatically notified when the rate changes so that internal class data can be appropriately updated. However, this has not been fully implemented. Specifically, classes that appropriately update their own data when either a setFrequency() or noteOn() function is called do not currently receive the automatic notification of rate change. If the user wants a specific class instance to ignore such notifications, perhaps in a multi-rate context, the function Stk::ignoreSampleRateChange() should be called.

◆ ignoreSampleRateChange()

void stk::Stk::ignoreSampleRateChange ( bool ignore = true)
inlineinherited

A function to enable/disable the automatic updating of class data when the STK sample rate changes.

This function allows the user to enable or disable class data updates in response to global sample rate changes on a class by class basis.

184{ ignoreSampleRateChange_ = ignore; };

Member Data Documentation

◆ STK_SINT8

const StkFormat stk::Stk::STK_SINT8
staticinherited

-128 to +127

◆ STK_SINT16

const StkFormat stk::Stk::STK_SINT16
staticinherited

-32768 to +32767

◆ STK_SINT24

const StkFormat stk::Stk::STK_SINT24
staticinherited

Lower 3 bytes of 32-bit signed integer.

◆ STK_SINT32

const StkFormat stk::Stk::STK_SINT32
staticinherited

-2147483648 to +2147483647.

◆ STK_FLOAT32

const StkFormat stk::Stk::STK_FLOAT32
staticinherited

Normalized between plus/minus 1.0.

◆ STK_FLOAT64

const StkFormat stk::Stk::STK_FLOAT64
staticinherited

Normalized between plus/minus 1.0.


The documentation for this class was generated from the following file: