Signals 1 to 4 are original recordings made for IEEE Std 269™-2010. The recordings were made with a ½” measurement microphone in a quiet and absorbent room. The signals were originally recorded with 48 kHz sampling rate. There is usable signal up to and slightly beyond 20 kHz, so these signals are usable for just about any bandwidth test application.
There has been no equalization of signals 1-4. The spectra of the compressed and uncompressed versions are similar, but not identical.
Signals 5 and 6 were obtained from ITU=T Recommendation P.50, Appendix 1. They are used here with permission of ITU. The original sample rate was 16 kHz. Although they are provided here, with a 44.1 kHz sampling rate, there is no usable signal above about 7500 Hz. These signals are suitable for wideband and narrowband applications.
Equalization has been applied to signals 5 and 6 to closely meet the spectrum specification in P.50. Except for the amplitude difference, the spectra of the compressed and uncompressed versions are very nearly the same.
“IEEE Male mono 48 kHz.wav” consists of 4 sentences spoken by 4 different persons.
Length = 12 s
CF = approx. 19 dB
“IEEE Male Compressed mono 48 kHz.wav” is a compressed version of #1. Pauses have been shortened and/or removed.
Length = 10 s
CF = approx. 14.5 dB
“IEEE Female mono 48kHz.wav” consists of 4 sentences spoken by 3 different persons.
Length = 10 s
CF = approx. 19.5 dB
“IEEE Female Compressed mono 48 kHz.wav” is a compressed version of #3. Pauses have been shortened and/or removed.
Length = 8.5 s
CF = approx. 14 dB
“IEEE P.50-2 mono 44.1 kHz.wav” consists of selected segments of the male artificial voice specified in ITU-T Recommendation P.50. It has been pre-equalized to meet the spectrum tolerance in P.50.
Length = 2 s
CF = approx 20 dB
“IEEE P.50-2 Compressed mono 44.1 kHz.wav” is a compressed version of #5.