next up previous contents
Next: About this document ... Up: Overflow User's Manual Previous: Troubleshooting   Contents

Node Documentation

This documentation is generated automatically from the comments included in the Overflow C++ source code.

AND (Logic)

Logical AND between two inputs

  Name Type Description
Inputs INPUT1 bool First boolean input
  INPUT2 bool Second boolean input
Outputs OUTPUT bool Boolean output
Parameters PULL_ANYWAY bool Pull on INPUT2 even if INPUT1 is false









Abs (DSP:Base)

Computes the absolute value of each element of a vector

  Name Type Description
Inputs INPUT Vector<float> Input vector
Outputs OUTPUT Vector<float> Output vector
Parameters none    









Accept (Network)

Create a network socket of any type

  Name Type Description
Inputs SOCKET socket The socket to listen to
Outputs SOCKET stream The socket to be used for input/output operations
Parameters none    









Accumulate (General)

Accumulation of objects into a buffer, that is, a vector of Objects References. When the node is in the main network or in a sub-network, his input is packed in the vector only once. However while in iterators, his input is packed (added) in the vector at every iteration. As well, his other input "ACCUM" must be connected to a node: "NewAccumulator"(General).

  Name Type Description
Inputs INPUT any Input object
  ACCUM Vector<ObjectRef> Accumulator where to put the input
Outputs OUTPUT Vector<ObjectRef> The input accumulator
Parameters none    









Action (General)

Pulls in order the inputs: BEFORE, INPUT and AFTER.

  Name Type Description
Inputs INPUT any The input
  BEFORE any To be pulled before
  AFTER any To be pulled after
Outputs OUTPUT any The output = The input
Parameters none    









AdaptMAP (HMM) (require: GMM) Performs MAP adaptation (well, almost!)

  Name Type Description
Inputs FRAMES Vector<float> Frame buffer
  GMM GMM GMM to be adapted
Outputs OUTPUT GMM Adapted GMM
Parameters none    









Add (Operator)

Adds two input values and returns the result

  Name Type Description
Inputs INPUT1 any First value
  INPUT2 any Second value
Outputs OUTPUT any Result of the addition
Parameters none    









Amplitude (ZDeprecated)

Deprecated, use RMS instead

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters none    









ArgMax (DSP:Base)

Finds the maximum value in a vector

  Name Type Description
Inputs INPUT Vector<float> Input vector
Outputs OUTPUT Vector<float> Index 0 contains the maximum value, index 1 contains the index where the maximum is found
Parameters START int Index where search is started
  END int Index where search ends









AudioStream (DSP:Audio)

Reads an audio stream and outputs frames

  Name Type Description
Inputs INPUT Stream An audio input stream (IStream, fd or FILE *)
Outputs AUDIO Vector<float> Frames read
  NOT_EOF bool True if we haven't reach the end of file yet
Parameters LENGTH int Length of the frames (in samples)
  ADVANCE int Offset beween frames (in samples)
  ENCODING string Type of encoding (LIN16, ULAW, ALAW, LIN8, SPHERE)
  STREAM_TYPE string Type of stream (stream, fd, FILE)
  REWIND bool If true, the stream rewinds to the beginning of the file when EOF is met









Autocor (DSP:Misc)

Computes the autocorrelation of an input vector with (START <= lag <= END)

  Name Type Description
Inputs INPUT Vector<float> Input vector
Outputs OUTPUT Vector<float> Autocorrelation vector
Parameters START int Smallest lag offset (included)
  END int Largest lag offset (included)
  CONTINUOUS bool Use the previous frame also (cross-correlation) (default false)
  NORMALIZE bool Energy normalization (default false)
  NORMALIZE2 bool Normalize by subtracting the value at lag/2 (default false)









BWExpan (DSP:Adaptive)

Performs bandwidth expansion on an LPC filter, that is, multiplying the radius of the poles by GAMMA

  Name Type Description
Inputs INPUT Vector<float> Original LPC filter
Outputs OUTPUT Vector<float> New "bandwidth expanded" LPC filter
Parameters GAMMA float Pole radius factor









BroadcastLoad (IO)

Load an object from file (registered type)

  Name Type Description
Inputs SOCKET Stream The stream we are loading from
Outputs OUTPUT any The loaded object
Parameters none    









BroadcastSave (IO)

Takes an object and saves it using a stream, returns the input object

  Name Type Description
Inputs OBJECT any The object that will be saved
  SOCKET Stream The output stream where to save
Outputs OUTPUT any The input object
Parameters none    









BuildDoc (General) (require: UIClasses) Builds a network from a document

  Name Type Description
Inputs INPUT UIDocument Loaded document
Outputs OUTPUT Network built network
Parameters none    









CAllPole (ZDeprecated) No description available

  Name Type Description
Inputs INPUT any No description available
  FILTER any No description available
Outputs OUTPUT any No description available
Parameters none    









CGain (DSP:Base)

No description available

  Name Type Description
Inputs INPUT any No description available
  GAIN any No description available
Outputs OUTPUT any No description available
Parameters none    









CMCalc (VQ) (require: CMap) No description available

  Name Type Description
Inputs INPUT any No description available
  CM any No description available
Outputs OUTPUT any No description available
Parameters OUTPUTLENGTH any No description available









CMS (DSP:TimeFreq)

Window-type Cepstram Mean Subtraction (CMS)

  Name Type Description
Inputs INPUT Vector<float> Input frames (cepstrum)
Outputs OUTPUT Vector<float> CMS output frames
Parameters LENGTH int Frame length (number of features)
  LOOKBACK int CMS window look-back (number of frames)
  LOOKAHEAD int CMS window look-ahead (number of frames)









CMTrain (VQ) (require: CMap) Trains a codebook map

  Name Type Description
Inputs TRAIN_IN Vector<ObjectRef> Input feature accumulator
  TRAIN_OUT Vector<ObjectRef> Output feature accumulator
  VQ KMeans Already trained vector quantizer
Outputs OUTPUT CodebookMap Resulting codebook map
Parameters none    









Catch (Flow)

Catches an exception

  Name Type Description
Inputs INPUT any Normal flow
  CATCH any Flow to follow if an exception is caught
Outputs OUTPUT any Flow output
  EXCEPTION any The exception caught (use only as feedback link)
Parameters none    









Collector (Flow)

Pass through with unlimited number of input/output pairs

  Name Type Description
Inputs INPUT any The input
Outputs OUTPUT any The output = The input (same name)
Parameters none    









Concat (Operator)

Concatenates two input values and returns the result

  Name Type Description
Inputs INPUT1 any First value
  INPUT2 any Second value
Outputs OUTPUT any Result of the concatenation
Parameters none    









Concatenate (Vector)

Concatenates two vectors together (deprecated, use Concat instead)

  Name Type Description
Inputs INPUT1 Vector<float> First input vector
  INPUT2 Vector<float> Second input vector
Outputs OUTPUT Vector<float> Concatenated vector
Parameters none    









Conj (DSP:Base)

Computes the complex conjugate of a vector

  Name Type Description
Inputs INPUT Vector<complex> Input vector
Outputs OUTPUT Vector<complex> Conjugate vector
Parameters none    









Connect (Network)

Create a network socket of any type

  Name Type Description
Inputs SOCKET socket The socket to listen to
  HOST any The host we want to connect to.
Outputs SOCKET stream The socket to be used for input/output operations
Parameters none    









Constant (General)

Defines a constant in terms of type and value. The different types are: int, float, bool, string, objects and subnet_param.

  Name Type Description
Inputs none    
Outputs VALUE any The value (parameter)
Parameters VALUE any The value









ConstantVector (ZDeprecated)

Creates a Constant vector

  Name Type Description
Inputs none    
Outputs OUTPUT Vector<float> The vector
Parameters VALUE string The string representation of the vector









CovarianceAccum (Matrix)

Updates (accumulate) a covariance matrix with an observation vector

  Name Type Description
Inputs INPUT Vector<float> Input (observation) vector
  MATRIX Matrix<float> Input (covariance) matrix
Outputs OUTPUT Matrix<float> Updated matrix (same object as input)
Parameters none    









DCMatrix (Matrix)

Creates a matrix of identical values

  Name Type Description
Inputs none    
Outputs OUTPUT Matrix<float> The matrix
Parameters ROWS int Number of rows
  COLUMNS int Number of columns
  VALUE float Value of each element









DCT (DSP:TimeFreq) (require: FFT) Fast implementation of the discrete cosine transform (DCT) using an FFT

  Name Type Description
Inputs INPUT Vector<float> The input vector
Outputs OUTPUT Vector<float> The result of the DCT
Parameters LENGTH int Length of the DCT
  FAST bool If true, the DCT is implemented using an FFT
  OUTPUTLENGTH int Number of coefficients to calculate (only if FAST=false)









DCVector (Vector)

Creates a vector of identical values

  Name Type Description
Inputs none    
Outputs OUTPUT Vector<float> The vector
Parameters LENGTH int The vector length
  VALUE float Value of each element









DTMF (DSP:Audio)

Generates a DTMF signal

  Name Type Description
Inputs INPUT Vector<int> DTMF vectors (line/column, starting at 0)
Outputs OUTPUT Vector<float> DTMF frames
Parameters LENGTH int Frame length
  SAMPLING int Sampling
  GAIN float Value of the gain









Delay (Flow)

Delay the input of DELAY iterations. Therefore, it can only be used in iterators.

  Name Type Description
Inputs INPUT any The input object
Outputs OUTPUT any The output object = input object with a delay
Parameters DELAY int The delay









DiagGMMScore (HMM) (require: DGMM) Scores a DiagGMM

  Name Type Description
Inputs INPUT Vector<float> Input vector
  GMM DiagGMM Input GMM (pdf)
Outputs OUTPUT Float GMM score
Parameters none    









DiagGMMSetScore (HMM) (require: DGMM) Scores a DiagGMM

  Name Type Description
Inputs INPUT Vector<float> Input vector
  GMM DiagGMM Input GMM set (pdf's)
Outputs OUTPUT Vector<float> GMM scores
Parameters none    









Discard (General)

Discards the object pulled

  Name Type Description
Inputs INPUT any The input object
Outputs OUTPUT NilObject Always return a NilObject
Parameters none    









Dist (DSP:Misc)

Calculates the distance between two vectors

  Name Type Description
Inputs INPUT1 Vector<float> First input vector
  INPUT2 Vector<float> Second input vector
Outputs OUTPUT Vector<float> Distance between INPUT1 and INPUT2
Parameters none    









Div (Operator)

Divides a numerator by a denominator

  Name Type Description
Inputs NUM any The numerator
  DEN any The denominator
Outputs OUTPUT any The result of the division
Parameters none    









DownSample (DSP:Base)

Downsamples a signal by outputing one sample for every N input samples

  Name Type Description
Inputs INPUT Vector<float> Downsampling input
Outputs OUTPUT Vector<float> Downsampled (by N) output
Parameters FACTOR int Downsampling factor N









Entropy (DSP:Misc)

Calculates the entropy of a vector

  Name Type Description
Inputs INPUT Vector<float> Input vector
Outputs OUTPUT Vector<float> Entropy value (vector of 1)
Parameters none    









Equal (Operator)

Returns true if two input values are equal, false otherwise

  Name Type Description
Inputs INPUT1 any First value
  INPUT2 any Second value
Outputs OUTPUT bool True or false
Parameters none    









ExecStream (IO)

A command to be executed (stdout is streamed)

  Name Type Description
Inputs INPUT string The command arg
Outputs OUTPUT Stream The stream
Parameters COMMAND string The command









Exp (DSP:Base)

Computes the exponential (base-e) of a vector

  Name Type Description
Inputs INPUT Vector<float> The input of the exponential
Outputs OUTPUT Vector<float> Result of the exponential
Parameters FAST bool Should we use exponential approximation









FDSaveFrame (ZDeprecated)

Writes audio frames to the sound card (or any other) file descriptor (deplaced by WriteAudio)

  Name Type Description
Inputs OBJECT Vector<float> Audio frames
  FD FILEDES (Sound card) File descriptor
Outputs OUTPUT Vector<float> Returning the input audio frames
Parameters LEAD_IN int Number of zero frames to send before starting (for synchronization)









FFT (DSP:TimeFreq) (require: FFT) Computes the real FFT of a float vector

  Name Type Description
Inputs INPUT Vector<float> The input vector
Outputs OUTPUT Vector<float> The FFT resuls as [r(0), r(1),..., r(N/2), i(N/2-1), ..., i(2), i(1)]
Parameters none    









FFTFlip (DSP:Base)

Flips a half-spectrum to produce a symetric spectrum

  Name Type Description
Inputs INPUT Vector<float> Half spectrum (real)
Outputs OUTPUT Vector<float> Symetric spectrum
Parameters none    









FIR (DSP:Filter)

Finite Impulse Response (FIR) filter

  Name Type Description
Inputs INPUT Vector<float> Input frame
  FILTER Vector<float> Filter coefficients
Outputs OUTPUT Vector<float> Filtered output
Parameters CONTINUOUS bool Should the frames be considered continuous (filter with memory). Default is true
  NONCAUSAL int Non-causality in number of samples. Default is causal filter









FLog (DSP:Base)

Computes the natural logarithm of a vector using a *rough* appriximation (only 17 MSB used)

  Name Type Description
Inputs INPUT Vector<float> The input of the log
Outputs OUTPUT Vector<float> Result of the log
Parameters none    









FMapCalc (VQ) (require: FeatureMap) Calculates the result of an hetero-associative map (trained by FMapTrain) for an input vector

  Name Type Description
Inputs INPUT Vector<float> Input vector
  FMAP FeatureMap The feature map that will be used
Outputs OUTPUT Vector<float> Output features
Parameters INPUTLENGTH int Number of input features
  OUTPUTLENGTH int Number of output features









FMapTrain (VQ) (require: FeatureMap) Trains an hetero-associative map based on a decision tree.

  Name Type Description
Inputs TRAIN_IN Vector<ObjectRef> An accumulator with input features
  TRAIN_OUT Vector<ObjectRef> An accumulator with input features
Outputs OUTPUT FeatureMap The trained 'feature map'
Parameters LEVELS any Number of levels to the decision tree









Feedback (Flow)

Feedback objects with a delay of n iterations.

  Name Type Description
Inputs INPUT any The input object
  BEFORE any When count < delay, pull the input from here
Outputs OUTPUT any The output object = input object
  DELAY any The delayed output of DELAY iteration
Parameters DELAY int Number of iteration for the delay
  BEFORE_LIMIT int When count - DELAY is smaller or equal to BEFORE_LIMIT, the input is pulled from BEFORE at (DELAY - count + BEFORE_LIMIT)









Filter (DSP:Filter) No description available

  Name Type Description
Inputs INPUT any No description available
  FIR any No description available
  IIR any No description available
Outputs OUTPUT any No description available
Parameters none    









Float2Vect (Vector)

Converts float values to a vector of elements in past and future

  Name Type Description
Inputs INPUT float The input float
Outputs OUTPUT Vector<float> The vector
Parameters LOOKAHEAD int Number of elements in the future
  LOOKBACK int Number of elements in the past









Floor (DSP:Base)

Floors vector values

  Name Type Description
Inputs INPUT Vector<float> Input vector
Outputs OUTPUT Vector<float> Output vector (after flooring)
Parameters THRESH float Threshold









FrameLabel (DSP:Audio)

Applies a gain to a vector

  Name Type Description
Inputs INPUT Stream Input stream
Outputs OUTPUT String Frame label
Parameters FRAME_ADVANCE int Frame advance to use









FuzzyModelExec (Fuzzy)

FuzzyModelExec takes a FuzzyModel and find its output according the the specified intput.

  Name Type Description
Inputs MODEL any The model to use
  INPUT any The input values to calculate
Outputs OUTPUT any The output of the fuzzy model
Parameters none    









FuzzyRule (Fuzzy)

A Rule containing ANTECEDANTS (IF) and CONSEQUENTS(THEN)

  Name Type Description
Inputs none    
Outputs RULE Vector<ObjectRef> The FuzzyRule Object
Parameters IF string Antecedant of the rule seperated by spaces
  THEN string Consequent of the rule seperated by spaces









FuzzySet (Fuzzy)

A FuzzySet containing functions associated with names

  Name Type Description
Inputs FUNCTIONS Vector<ObjectRef> The Fuzzy Functions
Outputs SET Vector<ObjectRef> The FuzzySet with multiple Fuzzy Functions
Parameters NAME string The name of the set









GCMS (DSP:TimeFreq)

Growing-Window Cepstral Mean Subtraction

  Name Type Description
Inputs INPUT Vector<float> Input frames
Outputs OUTPUT Vector<float> CMS output
Parameters LENGTH int Frame lentgh (features)









GCMS2 (DSP:TimeFreq)

Growing-Window Cepstral Mean Subtraction, counting only speech

  Name Type Description
Inputs INPUT Vector<float> Input frames
  IS_SPEECH bool Whether the frame is speech
Outputs OUTPUT Vector<float> CMS output
Parameters LENGTH int Frame lentgh (features)









GMMScore (HMM)

Scores a GMM for a given frame

  Name Type Description
Inputs FRAMES Vector<float> Frames that will be scored
  GMM GMM GMM used as dpf
Outputs OUTPUT Vector<float> Log-score (as a vector of 1)
Parameters none    









GMMTrain (HMM)

Trains a GMM using an accumulator of frames

  Name Type Description
Inputs FRAMES Vector<ObjectRef> Frame Accumulator
Outputs OUTPUT GMM The trained GMM
Parameters SPLIT_LEVELS int Number of times to perform the split = log2 (number of gaussians)









Gain (DSP:Base)

Applies a gain to a vector

  Name Type Description
Inputs INPUT Vector<float> Input vector
Outputs OUTPUT Vector<float> Output vector (after gain)
Parameters GAIN float Value of the gain









GenericModel (Fuzzy)

A generic Fuzzy controller

  Name Type Description
Inputs RULES Vector<ObjectRef> The Rules to use
  ANTECEDENT_SETS Vector<ObjectRef> The Sets to use
  CONSEQUENT_SETS Vector<ObjectRef> The Sets to use
  INPUT Vector<float> The input value of the variables
Outputs MODEL Model The model (cloned)
  OUTPUT Vector<float> The defuzzified values
Parameters none    









GetComposite (General)

Split up a composite object. This node makes just the opposite of the node "MakeComposite", that is, split up his compressed input (the composite object) into several outputs. However, the outputs must be added manually by by users. To add outputs to the node: double-click on it and click on the tab "Input/Outputs". Give a name to the output and press "Add". Repeat as long as you wish. Therefore, you can regroup inputs with "MakeComposite", send them in one output (the composite object) and get them back with "GetComposite". However, if you want to do so, inputs of "MakeComposite" and outputs of "GetComposite" must have corresponding names.

  Name Type Description
Inputs INPUT CompositeType Composite object
Outputs none    
Parameters none    









Greater (Logic)

Verifies if INPUT1 > INPUT2

  Name Type Description
Inputs INPUT1 any The first operand
  INPUT2 any The second operand
Outputs OUTPUT bool bool value
Parameters none    









GtkPlotProbe (Probe)

Plots a vector using GtkPlot

  Name Type Description
Inputs INPUT Vector<float> Input vector
Outputs OUTPUT Vector<float> Same as input
Parameters BREAK_AT int If set, the probe runs until (count = BREAK_AT)
  SHOW bool Whether or not to show the the data by default
  SKIP int Count increment for each "Next"
  PROBE_NAME string Name (title) of the probe









HP (DSP:Filter) No description available

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters DELAY int No description available
  FILTER_LENGTH int No description available
  FREQ float No description available









HistoVect (DSP:Manip)

No description available

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters LENGTH any No description available









IDCT (DSP:TimeFreq) (require: FFT) Fast implementation of the inverse discrete cosine transform (IDCT) using an FFT

  Name Type Description
Inputs INPUT Vector<float> The input vector
Outputs OUTPUT Vector<float> The result of the DCT
Parameters LENGTH int Length of the input and output vectors









IF (Logic)

Takes a branch or another depending on a condition (Bool value).

  Name Type Description
Inputs COND bool The condition for the if statement
  THEN any What to do if the condition is true
  ELSE any What to do if the condition is false
Outputs OUTPUT any The object from THEN or ELSE depending on COND
Parameters PULL_ANYWAY bool If true, the IF statement pulls also on the branch not taken









IIR (DSP:Filter)

All-pole (IIR) filter

  Name Type Description
Inputs INPUT Vector<float> Input signal (frames)
  FILTER Vector<float> Filter coefficients (denominator)
Outputs OUTPUT Vector<float> Filtered signal
Parameters none    









ILTF (DSP:Filter)

Inverse (all-pole) long-term (comb) filter

  Name Type Description
Inputs INPUT Vector<float> Input frame
  FILTER Vector<float> Filter params as [gain, period]
Outputs OUTPUT Vector<float> Filtered signal
Parameters none    









IRFFT (DSP:TimeFreq) (require: FFT) Inverse FFT, half complex to real vector

  Name Type Description
Inputs INPUT Vector<complex> Half complex vector
Outputs OUTPUT Vector<float> Real inverse FFT output
Parameters none    









Index (Vector)

Returns a float value from a vector

  Name Type Description
Inputs INPUT Vector<float> The input vector
  INDEX int Index value (if not specified in parameter)
Outputs OUTPUT float Float at a certain index
Parameters INDEX int Index value









Index2Vector (DSP:Manip)

No description available

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters LENGTH any No description available









InferenceModel (Fuzzy)

A generic Fuzzy controller

  Name Type Description
Inputs RULES Vector<ObjectRef> The Rules to use
  ANTECEDENT_SETS Vector<ObjectRef> The Sets to use
  CONSEQUENT_SETS Vector<ObjectRef> The Sets to use
  INPUT Vector<float> The input value of the variables
Outputs MODEL Model The model (cloned)
  OUTPUT Vector<float> The defuzzified values
  OUTPUT_SETS Vector<ObjectRef> The copied consequent set(s)
Parameters none    









InputStream (IO)

Creates a read-only stream from a filename

  Name Type Description
Inputs INPUT string The file name
Outputs OUTPUT Stream The Stream
Parameters TYPE String Type of stream: stream, fd, or FILE (default stream)
  RETRY int If set to N, InputStream will retry N times on open fail









IterCount (Logic)

Get the iterator count (iteration number)

  Name Type Description
Inputs none    
Outputs OUTPUT int The iteration count
Parameters none    









IterWall (Flow)

Get the input object only once, compute the result and always give the same output afterwards.

  Name Type Description
Inputs INPUT any The input object
Outputs OUTPUT any The output object = the input object (calculated once)
Parameters none    









Iterate (Flow)

Specify the number of iteration to do (max). Therefore, it can only be used in iterators and the output must be set to COND (left click on the node output and hold the control).

  Name Type Description
Inputs none    
Outputs OUTPUT bool Return true if count < MAX_ITER
Parameters MAX_ITER int Number of iteration to do (max)









KeyPad (Probe)

No description available

  Name Type Description
Inputs none    
Outputs KEYPAD Vector<int> A vector of size 2 representing the tuple Line/column of the pressed key.
  KEYPAD_ID int The Id of the key that is pressed
  KEYPAD_NAME Char The Char description of the key that is pressed
  ACTIVATED bool True if the user is pressing a button, else false.
Parameters none    









LP (DSP:Filter) No description available

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters DELAY int No description available
  FILTER_LENGTH int No description available
  FREQ float No description available









LPC (DSP:Adaptive)

Performs LPC (Linear predictive coefficient) analysis

  Name Type Description
Inputs INPUT Vector<float> Input (audio) vector
Outputs OUTPUT Vector<float> LPC coefficients (including a[0]=1)
Parameters OUTPUTLENGTH int Number of LPC coefficients (order = OUTPUTLENGTH-1)
  RADIUS float Maximum radius of the poles (used for bandwidth expansion)
  LAG_THETA float Minimum resonnance bandwidth allowed (with lag-windowing, approximative)









LPC2PS (DSP:Adaptive) (require: FFT) Calculates the spectral envelope corresponding to an all-pole filter (LPC coefficients)

  Name Type Description
Inputs INPUT Vector<float> LPC coefficients (including the '1' as first coefficient)
Outputs OUTPUT Vector<float> Points of the spectral envelope
Parameters OUTPUTLENGTH int Number of points for the spectral envelope









LPC_DECOMP (DSP:Adaptive) No description available

  Name Type Description
Inputs INPUT any No description available
Outputs LPC any No description available
  EXC any No description available
Parameters LPC_SIZE int No description available
  LAG_THETA float No description available
  FRAME_SIZE2 int No description available









LPFilter (DSP:Filter)

No description available

  Name Type Description
Inputs none    
Outputs OUTPUT any No description available
Parameters LENGTH any No description available
  THETA any No description available
  HP any No description available









LTF (DSP:Filter)

Long-term (comb) filter

  Name Type Description
Inputs INPUT any No description available
  FILTER any No description available
Outputs OUTPUT any No description available
Parameters none    









LTP (DSP:Adaptive)

Long-term predictor, finds best correlation (pitch) within (START <= sample delay <= END)

  Name Type Description
Inputs INPUT Vector<float> Input frame
Outputs OUTPUT Vector<float> [pitch gain, pitch period]
Parameters START int Smallest pitch allowed
  END int Largest pitch allowed









Length (Vector)

Get the length of a vector

  Name Type Description
Inputs INPUT any The vector input
Outputs OUTPUT int The length of the vector
Parameters none    









List (General)

Load a string from a file (seperated into chunks of 256 bytes)

  Name Type Description
Inputs STREAM Stream The stream to load from
Outputs OUTPUT Vector<ObjectRef> The vector output
Parameters none    









Listen (Network)

Create a network socket of any type

  Name Type Description
Inputs SOCKET socket The socket to listen to
Outputs SOCKET socket The socket to be used for input/output operations
Parameters BACKLOG int Number of incoming connections allowed
  BLOCKING bool Blocking call to accept.









Load (IO)

Load an object from file (registered type)

  Name Type Description
Inputs STREAM Stream The stream we are loading from
Outputs OUTPUT any The loaded object
Parameters none    









LoadDoc (General) (require: UIClasses) Loads an Overflow XML document

  Name Type Description
Inputs INPUT string Document name
Outputs OUTPUT UIDocument loaded document
Parameters none    









LoadFile (IO) No description available

  Name Type Description
Inputs none    
Outputs OUTPUT any No description available
Parameters FILENAME string No description available









Log (DSP:Base)

Computes the natural logarithm of a vector

  Name Type Description
Inputs INPUT Vector<float> The input of the log
Outputs OUTPUT Vector<float> Result of the log
Parameters FAST bool Should we use fast log approximation









MDCT (DSP:TimeFreq) (require: MDCT) MDCT implementation (taken from Vorbis)

  Name Type Description
Inputs INPUT Vector<float> Input frame
Outputs OUTPUT Vector<float> MDCT result
Parameters LENGTH int Frame (not window) size









MFCC (ZDeprecated)

Calculates MFCC coefficients from an audio frame (all in one)

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters INPUTLENGTH any No description available
  OUTPUTLENGTH any No description available
  WINDOW any No description available
  SAMPLING any No description available
  LOW any No description available
  HIGH any No description available









MMIScore (VQ)

No description available

  Name Type Description
Inputs FRAMES any No description available
  MMI any No description available
Outputs OUTPUT any No description available
Parameters none    









MMITrain (VQ)

Train Maximum Mutual Information (MMI) Tree

  Name Type Description
Inputs FRAMES Vector<ObjectRef> No description available
Outputs OUTPUT Cell MMI tree
Parameters LEVELS int Number of levels for the tree









MSVQTrain (VQ) (require: MSVQ) Training of a multi-stage vector quantizer

  Name Type Description
Inputs FRAMES any No description available
Outputs OUTPUT any No description available
Parameters STAGES any No description available
  BINARY any No description available









MakeComposite (General)

Creates a composite object. A composite object is somewhere like a structure in C++. Indeed, a composite object is a regrouping of inputs like a structure is a regrouping of variables. You can chose the name and the number of inputs that you want your node to containt. To add inputs in the node: double-click on it and click on the tab "Input/Outputs". Give a name to the input and press "Add". Repeat as long as you wish. Then, the node will regoup all of these inputs together in only one output (a composite object).

  Name Type Description
Inputs none    
Outputs OUTPUT any Create a new composite object.
Parameters none    









MakeDiagGMM (HMM) (require: GMM) Transforms a GMM into a DiagGMM

  Name Type Description
Inputs INPUT GMM Input GMM
Outputs OUTPUT DiagGMM Output DiagGMM
Parameters none    









MarkovProb (HMM)

Calculates the Markov chain probability

  Name Type Description
Inputs INPUT any State probability
  MATRIX any Transition probability matrix
Outputs OUTPUT any A posteriori probability
Parameters none    









MatProduct (Matrix)

Matrix x vector product

  Name Type Description
Inputs INPUT Vector<float> Input vector
  MATRIX Matrix<float> Matrix
Outputs OUTPUT Vector<float> Result
Parameters none    









Max (Operator)

The maximum value

  Name Type Description
Inputs INPUT1 any First value
  INPUT2 any Second value
Outputs OUTPUT any The maximum value between INPUT1 and INPUT2
Parameters none    









Mel (DSP:TimeFreq)

calculates Mel-scale channel energies from power-spectrum

  Name Type Description
Inputs INPUT Vector<float> Input power-spectrum
Outputs OUTPUT Vector<float> Mel-scale channel energies
Parameters INPUTLENGTH int Power-spectrum size
  OUTPUTLENGTH int Number of channel energies
  SAMPLING int Sampling rate used (used for power-spectrum range)
  LOW int Lowest frequency
  HIGH int Highest frequency









MergeChannels (DSP:Audio)

No description available

  Name Type Description
Inputs LEFT any No description available
  RIGHT any No description available
Outputs OUTPUT any No description available
Parameters none    









Min (Operator)

Selects the minimum between two values

  Name Type Description
Inputs INPUT1 any The first value
  INPUT2 any The second value
Outputs OUTPUT any The minimum value between INPUT1 and INPUT2
Parameters none    









Mul (Operator)

Multiplication between two values, vectors, objects (operator* must be defined)

  Name Type Description
Inputs INPUT1 any The first operand
  INPUT2 any The second operand
Outputs OUTPUT any The result of INPUT1 * INPUT2
Parameters none    









MultiPlotProbe (Probe)

Plots multiple vectors using GtkPlot

  Name Type Description
Inputs INPUT Vector<ObjectRef> Input vectors
Outputs OUTPUT Vector<float> Same as input
Parameters BREAK_AT int If set, the probe runs until (count = BREAK_AT)
  SHOW bool Whether or not to show the the data by default
  SKIP int Count increment for each "Next"
  PROBE_NAME string Name (title) of the probe









NLMS (DSP:Adaptive)

Normalized LMS algorithm

  Name Type Description
Inputs INPUT any The input of the adaptive FIR filter
  REF any The signal being tracked
Outputs OUTPUT any The output of the adaptive FIR filter (not the residue)
Parameters FILTER_LENGTH any Length of the adaptive FIR filter
  ALPHA any Adaptation rate of the filter coefficients
  BETA any Adaptation rate of the normalization energy estimate
  POWER any Normalization power









NNetExec (NNet) (require: FFNet) No description available

  Name Type Description
Inputs INPUT any No description available
  NNET any No description available
Outputs OUTPUT any No description available
Parameters OUTPUTLENGTH any No description available









NNetExecRecurrent (NNet)

No description available

  Name Type Description
Inputs INPUT any No description available
  NNET any No description available
Outputs OUTPUT any No description available
Parameters OUTPUTLENGTH any No description available









NNetInit (NNet) (require: FFNet) Initialized the neural network weights to fit the input/output set

  Name Type Description
Inputs TRAIN_IN Vector<ObjectRef> Training input data
  TRAIN_OUT Vector<ObjectRef> Training output data
Outputs OUTPUT FFNet Initialized feed-forward neural network
Parameters TOPO string Number of units on each layer (including input and output layers)
  FUNCTIONS string Activation functions for each layer (except the input layer)
  RAND_SEED int Sets to random seed to RAND_SEED before initialization









NNetNew (NNet)

Returns a new (MLP) neural network

  Name Type Description
Inputs none    
Outputs OUTPUT NNet New (MLP) neural network
Parameters TOPO Vector<string> No description available









NNetSetCalc (NNet)

No description available

  Name Type Description
Inputs INPUT any No description available
  ID any No description available
  NNET any No description available
Outputs OUTPUT any No description available
Parameters OUTPUTLENGTH any No description available









NNetSetChooseBest (NNet)

Initialized the neural network weights to fit the input/output set

  Name Type Description
Inputs TRAIN_IN any No description available
  TRAIN_OUT any No description available
  TRAIN_ID any No description available
  NET1 any No description available
  NET2 any No description available
Outputs OUTPUT any No description available
Parameters none    









NNetSetInit (NNet)

Initialized the neural network weights to fit the input/output set

  Name Type Description
Inputs TRAIN_IN any No description available
  TRAIN_OUT any No description available
  TRAIN_ID any No description available
Outputs OUTPUT any No description available
Parameters NB_NETS any No description available
  TOPO any No description available
  FUNCTIONS any No description available
  RAND_SEED any No description available









NNetSetTrain (NNet)

No description available

  Name Type Description
Inputs TRAIN_IN any No description available
  TRAIN_OUT any No description available
  TRAIN_ID any No description available
  NNET any No description available
Outputs OUTPUT any No description available
Parameters MAX_EPOCH any No description available
  LEARN_RATE any No description available
  MOMENTUM any No description available
  INCREASE any No description available
  DECREASE any No description available
  ERR_RATIO any No description available
  BATCH_SETS any No description available









NNetSetTrainCGB (NNet)

No description available

  Name Type Description
Inputs TRAIN_IN any No description available
  TRAIN_OUT any No description available
  TRAIN_ID any No description available
  NNET any No description available
Outputs OUTPUT any No description available
Parameters MAX_EPOCH any No description available
  SIGMA any No description available
  LAMBDA any No description available









NNetSetTrainDBD (NNet)

No description available

  Name Type Description
Inputs TRAIN_IN any No description available
  TRAIN_OUT any No description available
  TRAIN_ID any No description available
  NNET any No description available
Outputs OUTPUT any No description available
Parameters MAX_EPOCH any No description available
  LEARN_RATE any No description available
  INCREASE any No description available
  DECREASE any No description available









NNetTrain (NNet) (require: FFNetTrain) No description available

  Name Type Description
Inputs TRAIN_IN any No description available
  TRAIN_OUT any No description available
  NNET any No description available
Outputs OUTPUT any No description available
Parameters MAX_EPOCH any No description available
  LEARN_RATE any No description available
  MOMENTUM any No description available
  INCREASE any No description available
  DECREASE any No description available
  ERR_RATIO any No description available
  BATCH_SETS any No description available









NNetTrainCGB (NNet) (require: FFNetTrain) No description available

  Name Type Description
Inputs TRAIN_IN any No description available
  TRAIN_OUT any No description available
  NNET any No description available
Outputs OUTPUT any No description available
Parameters MAX_EPOCH any No description available
  SIGMA any No description available
  LAMBDA any No description available









NNetTrainDBD (NNet) (require: FFNetTrain) Neural network (MLP) training unsing the Delta-bar-delta algorithm

  Name Type Description
Inputs TRAIN_IN Vector<ObjectRef> Input data accumulator
  TRAIN_OUT Vector<ObjectRef> Output data accumulator
  NNET FFNet Neural network that will be trained
Outputs OUTPUT FFNet Trained network
Parameters MAX_EPOCH int Number of training epoch (default 2000)
  LEARN_RATE float Initial learning rate (default 0.000001)
  INCREASE float Learning rate increment (> 1.0) factor (default 1.04)
  DECREASE float Learning rate decrement (< 1.0) factor (default 0.6)
  NB_SETS int Number of batch subsets for accelerated training (default 1)
  ALLOC_CHUNK bool If true, a big vector is allocated to store all the inputs (default false)
  RPROP bool If true, use the RProp variant of delta-bar-delta (default false)









NNetTrainQProp (NNet) (require: FFNetTrain) Neural network (MLP) training unsing the Quickprop algorithm

  Name Type Description
Inputs TRAIN_IN Vector<ObjectRef> Input data accumulator
  TRAIN_OUT Vector<ObjectRef> Output data accumulator
  NNET FFNet Neural network that will be trained
Outputs OUTPUT FFNet Trained network
Parameters MAX_EPOCH int Number of training epoch (default 2000)
  LEARN_RATE float Initial learning rate (default 0.000001)
  INCREASE float Learning rate increment (> 1.0) factor (default 1.04)
  DECREASE float Learning rate decrement (< 1.0) factor (default 0.6)









NNetTrainRecurDBD (NNet) (require: FFNetTrain) No description available

  Name Type Description
Inputs TRAIN_IN any No description available
  TRAIN_OUT any No description available
  NNET any No description available
Outputs OUTPUT any No description available
Parameters MAX_EPOCH any No description available
  LEARN_RATE any No description available
  MOMENTUM any No description available
  INCREASE any No description available
  DECREASE any No description available
  BATCH_SETS any No description available









NNetTrainRecurrent (NNet) (require: FFNetTrain) No description available

  Name Type Description
Inputs TRAIN_IN any No description available
  TRAIN_OUT any No description available
  NNET any No description available
Outputs OUTPUT any No description available
Parameters MAX_EPOCH any No description available
  LEARN_RATE any No description available
  MOMENTUM any No description available
  INCREASE any No description available
  DECREASE any No description available









NNetTrainSCG (NNet) (require: FFNetTrain) Neural network (MLP) training unsing the scaled conjugate gradient algorithm

  Name Type Description
Inputs TRAIN_IN Vector<ObjectRef> Input data accumulator
  TRAIN_OUT Vector<ObjectRef> Output data accumulator
  NNET FFNet Neural network that will be trained
Outputs OUTPUT FFNet Trained network
Parameters MAX_EPOCH int Number of training epoch (default 2000)
  SIGMA float Sigma parameter
  LAMBDA float Lambda parameter









NNetTrainWeightDBD (NNet) (require: FFNetTrain) Neural network (MLP) training unsing the Delta-bar-delta algorithm

  Name Type Description
Inputs TRAIN_IN Vector<ObjectRef> Input data accumulator
  TRAIN_OUT Vector<ObjectRef> Output data accumulator
  TRAIN_WEIGHT Vector<ObjectRef> Error weights for training
  NNET FFNet Neural network that will be trained
Outputs OUTPUT FFNet Trained network
Parameters MAX_EPOCH int Number of training epoch (default 2000)
  LEARN_RATE float Initial learning rate (default 0.000001)
  INCREASE float Learning rate increment (> 1.0) factor (default 1.04)
  DECREASE float Learning rate decrement (< 1.0) factor (default 0.6)









NOP (General)

Pass Through (no operation)

  Name Type Description
Inputs INPUT any The input
Outputs OUTPUT any The output = The input
Parameters none    









NOT (Logic)

Logical NOT of an input

  Name Type Description
Inputs INPUT bool Boolean input
Outputs OUTPUT bool Boolean output
Parameters none    









NewAccumulator (General)

Creates a new Accumulator, that is a vector of Objects References. Accumulators are often used as the input "ACCUM" of the node "Accumulate".

  Name Type Description
Inputs none    
Outputs OUTPUT Vector<ObjectRef> Empty accumulator
Parameters none    









Noise (DSP:Misc)

Noise generator

  Name Type Description
Inputs none    
Outputs OUTPUT Vector<float> Noise signal (uncorrelated)
Parameters LENGTH int Length of the generated noise signal (frame length)
  TYPE string Noise type (UNIFORM, GAUSS, TRIANGLE)
  SD float Noise standard deviation









Normalize (DSP:Base)

Normalizes a vector by dividing each element by the sum of all components

  Name Type Description
Inputs INPUT Vector<float> Input vector
Outputs OUTPUT Vector<float> Normalized vector
Parameters none    









OR (Logic)

Logical OR between two inputs

  Name Type Description
Inputs INPUT1 bool First boolean input
  INPUT2 bool Second boolean input
Outputs OUTPUT bool Boolean output
Parameters PULL_ANYWAY bool Pull on INPUT2 even if INPUT1 is true









OffsetMatrix (DSP:Manip)

Returns a matrix of frames with offset

  Name Type Description
Inputs INPUT Vector<float> Input frame
Outputs OUTPUT Matrix<float> Matrix (ready for SVD, ...)
Parameters COLUMNS int Number of columns (subframe length)
  ROWS int Number of rows (number of offsets)









OutputStream (IO)

Creates a write-only stream from a filename

  Name Type Description
Inputs INPUT string The file name
Outputs OUTPUT Stream The Stream
Parameters TYPE string Type of stream: stream, fd, or FILE (default stream)









Overlap (DSP:Manip)

Outputs overlapping frames from non-overlapping ones

  Name Type Description
Inputs INPUT Vector<float> (Non-overlapped) input frames
Outputs OUTPUT Vector<float> Overlapped output frames
Parameters OUTPUTLENGTH int Frame length for output overlapped frames









OverlapAndAdd (DSP:Audio)

No description available

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters none    









PS (DSP:TimeFreq)

Converts the output of the FFT (not RFFT) node to a power spectrum

  Name Type Description
Inputs INPUT Vector<float> Spectrum output from FFT
Outputs OUTPUT Vector<float> Power spectrum (half the input length)
Parameters none    









PS2LPC (DSP:Adaptive) (require: FFT) Computes LPC coefficients from the spectral envelope of the all-pole filter

  Name Type Description
Inputs INPUT Vector<float> Spectral envelope
Outputs OUTPUT Vector<float> LPC coefficients
Parameters INPUTLENGTH int Number of points in the spectral envelope
  OUTPUTLENGTH int Number of LPC coefficients (order + 1)
  LAG_THETA float Lag-windowing parameter (roughly the minimum bandwidth of resonances)









Pack (Flow)

Pack Data into a vector of Objects references. When the node is in the main network or in a sub-network, his input is packed in the vector only once. However while in iterators, his input is packed (added) in the vector at every iteration.

  Name Type Description
Inputs INPUT any Objects to be packed (until processCount reached)
Outputs OUTPUT Vector<ObjectRef> A vector of ObjectRef(s)
Parameters none    









PackFrames (DSP:Base)

No description available

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters LENGTH any No description available
  BACK any No description available
  FRONT any No description available









ParallelThread (Flow)

Provides parallelism-type threading. When asked for an input, it computes both inputs at the same time and caches the other.

  Name Type Description
Inputs INPUT1 any First parallelized input
  INPUT2 any Second parallelized input
Outputs OUTPUT1 any First parallelized output
  OUTPUT2 any Second parallelized output
Parameters none    









PlotProbe (Probe)

No description available

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters BREAK_AT any No description available
  SHOW any No description available
  SKIP any No description available









Poly (DSP:Base)

No description available

  Name Type Description
Inputs INPUT any No description available
  COEF any No description available
Outputs OUTPUT any No description available
Parameters none    









Pow (DSP:Base)

Raises the input vector to a certain power

  Name Type Description
Inputs INPUT Vector<float> Input vector
Outputs OUTPUT Vector<float> Result: INPUT.
Parameters EXP float Exponent









Print (IO) No description available

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters none    









Probe (Probe)

Helps debugging by allowing to "trace" programs

  Name Type Description
Inputs INPUT any Any data
Outputs OUTPUT any Pass through
Parameters BREAK_AT int If set, the probe runs until (count = BREAK_AT)
  SHOW bool Whether or not to show the the data by default
  SKIP int Count increment for each "Next"
  PROBE_NAME string Name (title) of the probe









RBFTrain (VQ)

No description available

  Name Type Description
Inputs FRAMES any No description available
Outputs OUTPUT any No description available
Parameters NB_GAUSSIANS any No description available









RFFT (DSP:TimeFreq) (require: FFT) Computes the FFT of a real input vector and output a complex result

  Name Type Description
Inputs INPUT Vector<float> Real vector
Outputs OUTPUT Vector<complex> Complex FFT output
Parameters none    









RMS (DSP:Misc)

Root mean squared (RMS) value of a signal

  Name Type Description
Inputs INPUT Vector<float> The input signal
Outputs OUTPUT Vector<float> The RMS value
Parameters none    









ReadInt (IO)

ReadInt an integer from file

  Name Type Description
Inputs STREAM Stream The stream we are loading from
Outputs OUTPUT int The (next) integer in the stream
Parameters none    









Receive (IO)

Receive data from a TCP/IP network (Not working yet)

  Name Type Description
Inputs none    
Outputs OUTPUT any No description available
Parameters VALUE any No description available









Recover (Flow)

Recovers from an error (BaseException)

  Name Type Description
Inputs INPUT any Normal flow
  CATCH any Flow to follow if an error (exception) is caught
Outputs OUTPUT any Flow output
  EXCEPTION String The error message caught (use only as feedback link)
Parameters none    









Reframe (DSP:Manip)

Applies a window on a frame

  Name Type Description
Inputs INPUT Vector<float> Input frame
Outputs OUTPUT Vector<float> Reframeed frame
Parameters LENGTH int Length of the frames
  ADVANCE int Frame advance (offset)









Reverb (DSP:Effects) (require: Reverb) Stereo reverb effect

  Name Type Description
Inputs LEFT Vector<float> Right Input Channel
  RIGHT Vector<float> Left Input Channel
Outputs LEFT Vector<float> Right Output Channel
  RIGHT Vector<float> Left Output Channel
Parameters ROOMSIZE float 0 < Room Size < 1 (default = .5)
  DAMP float 0 < Damp < 1 (default = .5)
  WET float 0 < Wet < 1 (default = 1/3)
  DRY float 0 < Dry < 1 (default = 0)
  WIDTH float 0 < Width < 1 (default = 1)









Round (Found)

Rounds a float values to the nearest integer

  Name Type Description
Inputs INPUT float The input float
Outputs OUTPUT int Nearest integer
Parameters none    









SampleAndHold (Flow)

Downsamples in the "count" domain

  Name Type Description
Inputs INPUT any The input x[count]
Outputs OUTPUT any x[count - (count "modulo" FACTOR)]
Parameters DOWNSAMPLING int The downsampling factor









SampleDelay (DSP:Base)

No description available

  Name Type Description
Inputs INPUT any No description available
  DELAY any No description available
Outputs OUTPUT any No description available
Parameters LENGTH any No description available
  DELAY any No description available
  LOOKBACK any No description available
  LOOKAHEAD any No description available









Saturate (DSP:Effects)

Saturation (distortion) effect

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters SATURATION any No description available
  THRESHOLD any No description available









Save (IO)

Takes an object and saves it using a stream, returns the input object

  Name Type Description
Inputs OBJECT any The object that will be saved
  STREAM Stream The output stream where to save
Outputs OUTPUT any The input object
Parameters PRETTY_PRINT bool If we want to print human readable output (and Matlab)









SaveAs (IO) No description available

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters FILENAME string No description available









Select (Vector)

Selects an index range in an input vector

  Name Type Description
Inputs INPUT Vector<float> Input vector
Outputs OUTPUT Vector<float> Output vector (size = END-START+1)
Parameters START int Start index (inculded)
  END int End index (included)









Send (General)

Send data through network (Not Working Yet)

  Name Type Description
Inputs none    
Outputs OUTPUT any No description available
Parameters VALUE string No description available









SeparChannels (DSP:Audio)

No description available

  Name Type Description
Inputs INPUT Vector<float> Stero frame (encoded as left, right, left, right, ...)
Outputs LEFT Vector<float> Frame for the left channel
  RIGHT Vector<float> Frame for the right channel
Parameters none    









SerialThread (Flow)

Provides a pipeline-type multi-threading. A thread is started and computes inputs before the are needed by the output node.

  Name Type Description
Inputs INPUT any Flow input (asynchronous flow)
Outputs OUTPUT any Output flow (synchronous)
Parameters LOOKAHEAD int Pipeline look-ahead









Serialize (IO)

Takes an object and saves it using a stream, returns the input object

  Name Type Description
Inputs OBJECT any The object that will be saved
  STREAM Stream The output stream where to save
Outputs OUTPUT any The input object
Parameters none    









SerializeAs (IO) No description available

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters FILENAME string No description available









Sleep (Flow)

Sleep a certain amount of time.

  Name Type Description
Inputs none    
Outputs VALUE any Always return TRUE.
Parameters SECONDS float Sleep x seconds.









Smaller (Logic)

Verifies if INPUT1 is smaller than INPUT2

  Name Type Description
Inputs INPUT1 any The first operand
  INPUT2 any The second operand
Outputs OUTPUT any Boolean output
Parameters none    









SmoothAdd (DSP:Manip)

No description available

  Name Type Description
Inputs LEFT any No description available
  CENTER any No description available
  RIGHT any No description available
Outputs OUTPUT any No description available
Parameters LENGTH any No description available









Socket (Network)

Create a network socket of any type

  Name Type Description
Inputs none    
Outputs OUTPUT socket The socket to be used for input/output operations
Parameters TYPE string Type of socket : BROADCAST, TCP_STREAM, etc.
  PORT int Communication port









Sort (DSP:Base)

Sorts an input vector

  Name Type Description
Inputs INPUT Vector<float> Input vector
Outputs OUTPUT Vector<float> Sorted output vector
Parameters none    









Sound (DSP:Audio)

Opens a sound device

  Name Type Description
Inputs none    
Outputs OUTPUT any A file descriptor to the sound device
Parameters DEVICE string Path to the sound device
  RATE int Sampling rate
  STEREO int 1 for stereo, 0 for mono
  MODE string R for sound input, W for sound output, RW for full-duplex mode
  BUFFER int Length of the audio buffer to allocate (not reliable)
  DUMMY any Put something here to output to a file









SpectrumProbe (Probe) No description available

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters BREAK_AT int No description available
  SHOW bool No description available
  SKIP int No description available
  SQRT bool No description available
  LOG bool No description available









Sqrt (DSP:Base)

Square root of a vector

  Name Type Description
Inputs INPUT Vector<float> Input vector
Outputs OUTPUT Vector<float> Result vector of square root
Parameters none    









Stderr (IO)

Returns the stderr stream (cerr)

  Name Type Description
Inputs none    
Outputs OUTPUT Stream Stderr stream
Parameters none    









Stdin (IO)

Returns the stdin stream (cin)

  Name Type Description
Inputs none    
Outputs OUTPUT Stream Stdin stream
Parameters none    









Stdout (IO)

Returns the stdout stream (cout)

  Name Type Description
Inputs none    
Outputs OUTPUT Stream Stdout stream
Parameters none    









StopRecord (DSP:Misc)

For SV

  Name Type Description
Inputs INPUT bool frame by frame
Outputs OUTPUT bool false when should stop
Parameters START int Number of true frames before start
  TIMEOUT int Number of false frames before end









StrCat (ZDeprecated)

Concatenates two strings together (deprecated, use Concat instead)

  Name Type Description
Inputs INPUT1 String First input string
  INPUT2 String Second input string
Outputs OUTPUT String Concatenated strings
Parameters none    









Sub (Operator)

Subtracts two values, Vectors, Objects

  Name Type Description
Inputs INPUT1 any The value to subtract from
  INPUT2 any The subtracted value
Outputs OUTPUT any The result of the subtraction
Parameters none    









Sum (Vector)

Sum of all the elements of a vector

  Name Type Description
Inputs INPUT Vector<float> The input vector
Outputs OUTPUT float The sum
Parameters none    









Sync (Flow)

No-op node for which count ratio (getInput/getOutput) = RATIO

  Name Type Description
Inputs INPUT any Input
Outputs OUTPUT any Output (no-op) same as input with different count
Parameters RATIO any (input/output) count ratio









TextProbe (Probe)

Prints the data as text

  Name Type Description
Inputs INPUT any Any data
Outputs OUTPUT any Pass through
Parameters BREAK_AT int If set, the probe runs until (count = BREAK_AT)
  SHOW bool Whether or not to show the the data by default
  SKIP int Count increment for each "Next"
  PROBE_NAME string Name (title) of the probe









ThreadJoin (Flow)

Acts like a mutex and prevents two Overflow threads from accessing the same (input) node at the same time.

  Name Type Description
Inputs INPUT any The input
Outputs OUTPUT any The output = The input
Parameters none    









Throw (Flow)

Throw a FlowException

  Name Type Description
Inputs INPUT any The Object included in the FlowException
Outputs OUTPUT any Will automatically throw a FlowException if pulled
Parameters none    









TimeAutocorr (DSP:Misc)

Autocorrelation across vectors (frames)

  Name Type Description
Inputs INPUT Vector<float> Input vectors (frames)
Outputs OUTPUT Vector<float> Autocorrelations (summed) for each delay
Parameters INPUTLENGTH int Length ov input vectors
  LOOKAHEAD int Maximum forward (non-causal) delay
  LOOKBACK int Maximum backward (causal) delay









TimeEntropy (DSP:Misc)

Non-stationnarity (pseudo-entropy) measure across vectors (frames)

  Name Type Description
Inputs INPUT Vector<float> Input vectors (frames)
Outputs OUTPUT Vector<float> Value of the non-stationnarity measure (as a vector of 1 component)
Parameters LOOKAHEAD int Maximum forward (non-causal) delay
  LOOKBACK int Maximum backward (causal) delay









TimeFilter (DSP:Filter)

Filters across vectors (frames)

  Name Type Description
Inputs INPUT Vector<float> Input vectors (frames)
Outputs OUTPUT Vector<float> Filtered vectors
Parameters LENGTH int Vector length
  FIR string FIR part as <Vector<float> ... >
  IIR string IIR part as <Vector<float> ... >
  LOOKAHEAD int Non-causality (in frames)









TimeMedian (DSP:Filter)

Performs median filtering across vectors (frames)

  Name Type Description
Inputs INPUT Vector<float> Input vectors (frames)
Outputs OUTPUT Vector<float> Median-filtered vectors
Parameters LENGTH int Vector size
  LOOKAHEAD int Median look back (number of frames)
  LOOKBACK int Median look ahead (number of frames)









Trace (Probe)

Pass Through, tracing initialization, requests, inputs and exceptions

  Name Type Description
Inputs INPUT any The input
Outputs OUTPUT any The output = The input
Parameters TAG string Tag to put on the lines









TransMatrix (HMM)

No description available

  Name Type Description
Inputs INPUT any state numbers in a frame buffer
Outputs OUTPUT any No description available
Parameters NB_STATES any Number of HMM states
  THRESHOLD any The minimum transition probability allowed









TrapezoidalFunction (Fuzzy)

A Fuzzy Function to be included in a FuzzySet

  Name Type Description
Inputs none    
Outputs FUNCTION Vector<ObjectRef> The FuzzyFunction object
Parameters A float A value
  B float B value
  C float C value
  D float D value
  NAME string The name of the function









TriangularFunction (Fuzzy)

No description available

  Name Type Description
Inputs none    
Outputs FUNCTION Vector<ObjectRef> The FuzzyFunction object
Parameters A float A value
  B float B value
  C float C value
  NAME string The name of the function









UnPack (Flow)

Unpack data already packed. This node makes just the opposite of "Pack" and is often used with it.

  Name Type Description
Inputs INPUT any The packed vector
Outputs OUTPUT any The single unpacked Object
  NOT_END any True if there's still data
Parameters none    









UpSample (DSP:Base)

Upsamples a signal by inserting zeros at regular interval

  Name Type Description
Inputs INPUT Vector<float> Input frames
Outputs OUTPUT Vector<float> Upsampled output
Parameters FACTOR int Upsampling factor









VMethod (General)

Applies a certain method on an object int or float. The name of the method to call can be: log, exp, sin or cos.

  Name Type Description
Inputs INPUT int or float Object on wich the method will be applied
Outputs OUTPUT int or float Return value of the method
Parameters METHOD string The name of the method to call









VQClass (VQ) (require: VQ) No description available

  Name Type Description
Inputs INPUT any No description available
  VQ any No description available
Outputs OUTPUT any No description available
Parameters none    









VQCloseness (VQ)

No description available

  Name Type Description
Inputs INPUT any No description available
  VQ any No description available
Outputs OUTPUT any No description available
Parameters none    









VQTrain (VQ) (require: VQ) No description available

  Name Type Description
Inputs FRAMES any No description available
Outputs OUTPUT any No description available
Parameters MEANS any No description available
  BINARY any No description available









VQWeightMeans (VQ)

No description available

  Name Type Description
Inputs INPUT any No description available
  VQ any No description available
Outputs OUTPUT any No description available
Parameters OUTPUTLENGTH any No description available









VQuantize (VQ) (require: VQ) No description available

  Name Type Description
Inputs INPUT any No description available
  VQ any No description available
Outputs OUTPUT any No description available
Parameters none    









VQuantizeDiff (VQ)

No description available

  Name Type Description
Inputs INPUT any No description available
  VQ any No description available
Outputs OUTPUT any No description available
Parameters LENGTH any No description available









VarLoad (General)

Load a variable. The variable is pulled from a node of type: " VarStore "(General) which has the same given name. The node: " VarStore " can be declared in a different Overflow file, but must be declared before.

  Name Type Description
Inputs none    
Outputs OUTPUT any The variable value
Parameters VARIABLE string The name of the variable t be loaded









VarStore (General)

Store a variable under a specified name. The variable may be used in other Overflow files by the node: " VarLoad "(General).

  Name Type Description
Inputs INPUT any The value of the variable
Outputs OUTPUT any The value of the variable
Parameters VARIABLE string The variable name









VectorCode (Vector)

Modifies a vector using C++ code

  Name Type Description
Inputs INPUT Vector<float> Input vector
Outputs OUTPUT Vector<float> Output vector (after gain)
Parameters CODE string C++ code inside function [void func(const float *x, float *y, int length)]









Window (DSP:Base)

Applies a window on a frame

  Name Type Description
Inputs INPUT Vector<float> Input frame
Outputs OUTPUT Vector<float> Windowed frame
Parameters LENGTH int Length of the frames/window
  WINDOW string Window type (HANNING, HAMMING, HALF_HANNING)
  SYMETRIC bool Symetric window, uses (length-1) for normalization









WriteAudio (DSP:Audio)

Writes audio frames to the sound card (or any other) stream

  Name Type Description
Inputs OBJECT Vector<float> Audio frames
  DEVICE Stream (Sound card) stream
Outputs OUTPUT Vector<float> Returning the input audio frames
Parameters LEAD_IN int Number of zero frames to send before starting (for synchronization)









ZCrossing (DSP:Misc)

Number of zero-crossing in a vector: count(v[i]*v[i+1]<0)

  Name Type Description
Inputs INPUT Vector<float> The input vector
Outputs OUTPUT float Number of zero-crossing
Parameters none    









dB (DSP:Base) No description available

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters none    









undB (DSP:Base) No description available

  Name Type Description
Inputs INPUT any No description available
Outputs OUTPUT any No description available
Parameters none    










next up previous contents
Next: About this document ... Up: Overflow User's Manual Previous: Troubleshooting   Contents
Jean-Marc Valin 2002-06-17