Modelling of reverberations and audioconference rooms

Muron, O.; Sikorav, J.

doi:10.1109/icassp.1986.1168848

Cited by 5 publications

(4 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Using similar arguments, we can prove that the maximum number of extrema points of , where is the th-order IIR model given by (4), is , and the corresponding number for the ARMA model is . This means that to faithfully model an energy spectrum whose magnitude exhibits extrema points on the interval , we require no fewer than parameters for the FIR case and for the IIR and ARMA cases.…”

Section: Adequateness Of Iir Models For Acoustic Echo Cancellationmentioning

confidence: 87%

See 1 more Smart Citation

Acoustic echo cancellation: do IIR models offer better modeling capabilities than their FIR counterparts?

Liavas¹,

Regalia²

1998

IEEE Trans. Signal Process.

View full text Add to dashboard Cite

The adequateness of IIR models for acoustic echo cancellation is a long-standing question, and the answers found in the literature are conflicting. We use results from rational Hankel norm and least-squares approximation, and we recall a test that provides a priori performance levels for FIR and IIR models. We apply this test to measured acoustic impulse responses. Upon comparing the performance levels of FIR and IIR models with the same number of free parameters, we do not observe any significant gain from the use of IIR models. We attribute this phenomenon to the shape of the energy spectra of the acoustic impulse responses so tested, which possess many strong and sharp peaks. Faithful modeling of these peaks requires many parameters, irrespective of the type of the model.

show abstract

Section: Adequateness Of Iir Models For Acoustic Echo Cancellationmentioning

confidence: 87%

“…• Conclusions based solely on simulations [4] are not entirely convincing if we cannot guarantee that we have approached the global minimum of the error performance surface for the IIR case. • The use of an equation error model [5] in the (output error) AEC scheme seems questionable on its principles.…”

Section: Introductionmentioning

confidence: 94%

Acoustic echo cancellation: do IIR models offer better modeling capabilities than their FIR counterparts?

Liavas¹,

Regalia²

1998

IEEE Trans. Signal Process.

View full text Add to dashboard Cite

show abstract

“…This pole/zero model corresponds to the following ARMA model equation in the time domain. (2) where ai are AR coefficients and bi are MA coefficients.…”

Section: Pole and Zero Modeling Of An Rtfmentioning

confidence: 99%

ARMA modeling of a room transfer function at low frequencies.

Hori

Makino

Kaneda

et al. 1994

J. Acoust. Soc. Jpn. (E), J Acoust Soc Jpn E

View full text Add to dashboard Cite

An MA model (FIR filter) is widely used for modeling an RTF. The MA model, however, requires a large number of filter taps to model an RTF in a room with a long reverberation time. On the other hand, an ARMA model (pole/ zero model: IIR filter) is expected to reduce the number of filter taps, since poles with few taps can express the resonance characteristics that cause long reverberation.1) There have been many studies on ARMA modeling of an RTF, but they did not report that the number of filter taps was effectively reduced by the ARMA model.2,3) The reason is probably that they studied ARMA modeling of an RTF for a wide frequency band where the mode density of a room is high, i.e., the number of poles is large. In this paper, we discuss the ARMA modeling of an RTF focusing on the low frequency band where the mode density is low.

show abstract

“…The motions of people or objects during the conference cause fast and non-easily predictable variations of the echo path impulse response [1]. Consequently, acoustic echo cancellers must respond quickly to these variations.…”

Section: Introductionmentioning

confidence: 99%