A calibrated noise-data-base, which consists of a calibration signal and an optimal selection of 23 different noises, is available. This data-base was earlier provided on a CD which is not available any-more.
The use of the data-base is described in two papers in:
Steeneken, H.J.M. and Varga, A. (1993) “Assessment for automatic speech recognition I: Comparison of assessment methods”, Speech Communication 12 (1993). nr. 3, pp. 241-246, 1993 P39.
Varga, A. and Steeneken, H.J.M. (1993) “Assessment for automatic speech recognition II: NOISEX-92: A database and an experiment to study the effect of additive noise on speech recognition systems”. Speech Communication 12 (1993), nr. 3, pp. 247-251 1993 P40.
In general each signal has a duration of aprox 4 minutes, sample rate 20 kHz, word-length 16 bit and is stored according the .wav format.
The contents of the database are described in: RSG-10_Noise-data-base. This report has three annexes respectively:
Annex A: APPEND_A_Noise-data-base,
Annex B: APPEND_B_Noise-data-base,
Annex C: APPEND_C_Noise-data-base.
1 SIGNAL001-20kHz 2 SIGNAL002-20kHz 3 SIGNAL003-20kHz
4 SIGNAL004-20kHz 5 SIGNAL005-20kHz 6 SIGNAL006-20kHz
7 SIGNAL007-20kHz 8 SIGNAL008-20kHz 9 SIGNAL009-20kHz
10 SIGNAL010-20kHz 11 SIGNAL011-20kHz 12 SIGNAL012-20kHz
13 SIGNAL013-20kHz 14 SIGNAL014-20kHz 15 SIGNAL015-20kHz
16 SIGNAL016-20kHz 17 SIGNAL017-20kHz 18 SIGNAL018-20kHz
19 SIGNAL019-20kHz 20 SIGNAL020-20kHz 21 SIGNAL021-20kHz
22 SIGNAL022-20kHz 23 SIGNAL023-20kHz 24 SIGNAL024-20kHz