Three algorithms for noise reduction have been tested in three types of measurements: The investigation of details in the measurements of speech intelligibility has to confirm that there is no advantage of noise reduction as expected for this type of algorithms. In the listening effort scaling the overall result does not show a difference. An advantage of noise reduction would have been considered possible. In the paired comparison measurements the babble noise, processed with the stereo noise reduction, is clearly preferred by the subjects at all sites.
展开▼
