ADM::ADM< F > Class Template Reference

Adaptive differential microphone, working for speech frequency range. More...

Public Member Functions
	ADM (F fs, F dist, unsigned lp_order, const F *lp_alphas, unsigned decomb_order, const F *decomb_alphas, F tau_beta=F(50e-3), F mu_beta=F(1e-4))
	Create Adaptive Differential Microphone. More...
F	process (const F &front, const F &back, const F &external_beta=F(-1), bool update_beta=true)
	ADM processes one frame. More...
F	beta () const

Private Attributes
Delay< F >	m_delay_front
Delay< F >	m_delay_back
Linearphase_FIR< F >	m_lp_bf
Linearphase_FIR< F >	m_lp_result
Linearphase_FIR< F >	m_decomb
F	m_beta
F	m_mu_beta
F	m_powerfilter_coeff
F	m_powerfilter_norm
F	m_powerfilter_state

Detailed Description

template<class F>
class ADM::ADM< F >

Adaptive differential microphone, working for speech frequency range.

Constructor & Destructor Documentation

ADM()

template<class F >

ADM::ADM< F >::ADM	(	F	fs,
		F	dist,
		unsigned	lp_order,
		const F *	lp_alphas,
		unsigned	decomb_order,
		const F *	decomb_alphas,
		F	tau_beta = F(50e-3),
		F	mu_beta = F(1e-4)
	)

Create Adaptive Differential Microphone.

Parameters

fs	Sampling rate / Hz
dist	Distance between physical microphones / m
lp_order	Filter order of FIR lowpass filter used for adaptation
lp_alphas	Pointer to array of alpha coefficients for the lowpass filter used for adaptation. Since this class uses linear phase FIR filters only, only the first half (order/2 + 1) of the coefficients will be read (coefficients for linear-phase FIR filters are symmetric).
decomb_order	Filter order of FIR compensation filter (compensates for comb filter characteristic). decomb_order <= 1 deactivates filter.
decomb_alphas	Pointer to array of alpha coefficients for the compensation filter used to compensate for the comb filter characteristic. Since this class uses linear phase FIR filters only, only the first half (order/2 + 1)of the coefficients will be read (coefficients for linear-phase FIR filters are symmetric).
tau_beta	Time constant of the lowpass filter used for averaging the power of the output signal
mu_beta	adaptation speed

Member Function Documentation

process()

template<class F >

F ADM::ADM< F >::process ( const F &  front, const F &  back, const F &  external_beta = F(-1), bool  update_beta = true  )

inline

ADM processes one frame.

Parameters

front	The current front input signal sample
back	The current rear input signal sample
external_beta	If >= 0, this is used as the "beta" parameter for direction to filter out. Else, the beta parameter is adapted to filtered out a direction so that best reduction of signal intensity from the back hemisphere is achieved.
update_beta	Perform the beta adaptation step?

Returns

The computed output sample

beta()

template<class F >

F ADM::ADM< F >::beta ( ) const

inline

Member Data Documentation

m_delay_front

template<class F >

Delay<F> ADM::ADM< F >::m_delay_front

private

m_delay_back

template<class F >

Delay<F> ADM::ADM< F >::m_delay_back

private

m_lp_bf

template<class F >

Linearphase_FIR<F> ADM::ADM< F >::m_lp_bf

private

m_lp_result

template<class F >

Linearphase_FIR<F> ADM::ADM< F >::m_lp_result

private

m_decomb

template<class F >

Linearphase_FIR<F> ADM::ADM< F >::m_decomb

private

m_beta

template<class F >

F ADM::ADM< F >::m_beta

private

m_mu_beta

template<class F >

F ADM::ADM< F >::m_mu_beta

private

m_powerfilter_coeff

template<class F >

F ADM::ADM< F >::m_powerfilter_coeff

private

m_powerfilter_norm

template<class F >

F ADM::ADM< F >::m_powerfilter_norm

private

m_powerfilter_state

template<class F >

F ADM::ADM< F >::m_powerfilter_state

private

---

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

• adm.hh