Tasking control for a backgorund DSP operation.

This class provides a receptacle for an audio stream. Since it is its own thread, it can provide limited read-time processing.

Source:

DSP_Thread.h:14

Author:

Green Lizard Staff

Version:

1.01

Entity	Inherited From	Type	Scope	Short Description
~CDSP_Thread		destructor	public	undocumented
CDSP_Thread		constructor	public	automatically generated
CDSP_Thread		constructor	public	undocumented
DSP_BufferSize		static attribute	public	undocumented
GetEvent		method	public	undocumented
GetRecognitionTime		method	public	undocumented
InitInstance		method	public	The run time entry and exit point.
MuLawToLinear		method	public	Mu law, a telephony standard, to linear PCM.
m_AudioBuffer		attribute	public	undocumented
m_Delta		attribute	public	undocumented
m_DownSampleSize		attribute	public	undocumented
m_Index		attribute	public	undocumented
m_MovingAverages		attribute	public	undocumented
m_NoiseFloor		attribute	public	undocumented
m_NumberOfSampleWindows		attribute	public	undocumented
m_RecognitionTime		attribute	public	undocumented
m_SamplesPerSecond		attribute	public	undocumented
m_SampleWindowSize		attribute	public	undocumented
m_State		attribute	public	undocumented
operator =		operator	public	automatically generated
States		enum	public	undocumented

undocumented

Source:

DSP_Thread.h:34

Code:

public ~ CDSP_Thread ( )

automatically generated

Source:

DSP_Thread.h

Code:

public CDSP_Thread (

const CDSP_Thread & )

undocumented

Source:

DSP_Thread.h:33

Code:

public CDSP_Thread (

int index )

undocumented

Source:

DSP_Thread.h:17

Code:

public static const long DSP_BufferSize

undocumented

Source:

DSP_Thread.h:51

Code:

public int GetEvent (

void )

undocumented

Source:

DSP_Thread.h:50

Code:

public float GetRecognitionTime (

void )

The run time entry and exit point.

Source:

DSP_Thread.h:37

Code:

public BOOL InitInstance (

void )

Mu law, a telephony standard, to linear PCM.

Telephony audio compression typically uses mu-law encoding, where four bits are the sample, and three are an exponent, with a sign bit. For Digital Signal Processing, we prefer a linear Pulse Code Modulation with no "companding" (the compression of the signal described above) so we can detect absolute signal levels without respect to the Signal to Noise ratio.

Source:

DSP_Thread.h:48

Code:

public int MuLawToLinear (

unsigned char ulawbyte )

undocumented

Source:

DSP_Thread.h:21

Code:

public char * m_AudioBuffer

undocumented

Source:

DSP_Thread.h:31

Code:

public long m_Delta

undocumented

Source:

DSP_Thread.h:31

Code:

public long m_DownSampleSize

undocumented

Source:

DSP_Thread.h:22

Code:

public int m_Index

undocumented

Source:

DSP_Thread.h:23

Code:

public long * m_MovingAverages

undocumented

Source:

DSP_Thread.h:31

Code:

public long m_NoiseFloor

undocumented

Source:

DSP_Thread.h:26

Code:

public int m_NumberOfSampleWindows

undocumented

Source:

DSP_Thread.h:24

Code:

public float m_RecognitionTime

undocumented

Source:

DSP_Thread.h:31

Code:

public long m_SamplesPerSecond

undocumented

Source:

DSP_Thread.h:31

Code:

public long m_SampleWindowSize

undocumented

Source:

DSP_Thread.h:19

Code:

public States m_State

automatically generated

Source:

DSP_Thread.h

Code:

public CDSP_Thread & operator = (

const CDSP_Thread & )

undocumented

Source:

DSP_Thread.h:18

Code:

public enum States {	Stalled ,
	Scanning ,
	Matching ,
	Silence ,
	Recognized ,
	Stop }