Speech in the VQE device environment

The speech quality over the converged voice and data networks represents a challenge for the network developers and operators. Elaborate signal processing algorithms are developed to improve the speech quality (called voice quality enhancement-VQE-device) on the network. Due to the high complexity of the processing procedures, the VQE device could generate an undesired degradation of the speech quality. Field performance testing tools providing only an objective estimator of the subjective opinion on speech quality have to evolve into devices that can correlate the estimated perceived score with a diagnosis of the VQE environment. The paper is focused on listening quality, analyzing aspects concerning automatic level control ALC and noise reduction NR algorithms that could be developed in a VQE device. An evaluation procedure is developed to analyze the objectively estimated human perceived quality in the VQE environment along with the VQE performance and its impact on speech. The results of the analysis are provided and a quality evaluation metric is proposed.