Page 128 - The Master Handbook Of Acoustics
P. 128

103
                                                                            SPEECH, MUSIC, AND NOISE



                                  Number of binary digits  Dynamic range, dB
                                             4                     24
                                             8                     48
                                            12                     72
                                            16                     96
                                            24                    144


                      The theoretical 96-dB dynamic range provided by the 16-bit digital
                      system is staggering to someone steeped in the traditional techniques.
                      At last there is a recording system that handles the concert hall
                      dynamic range reasonably well. The commercial compact disc (CD) is
                      capable of carrying 74 minutes of full-fidelity music with a 96 dB sig-
                      nal-to-noise ratio. The digital audio cassette is another worthy addi-
                      tion to the recording/playback arsenal, as well as digital audio tape
                      (DAT) systems. Digital techniques have transferred dynamic range
                      limitations from the medium to the concert hall on the one hand and
                      the playback environment on the other.
                         However, in recent years the audio community has grown dissatis-
                      fied with the quantization noise and  “graininess” of 16-bit digital
                      audio. New 24-bit formats are on the rise, such as “Super Audio CD”
                      and  “Audio DVD.” In addition, professional audio mastering engi-
                      neers are now working in 24-bit resolution to avoid the audibility of
                      digital artifacts resulting from lower-resolution processing.


                      Power in Speech and Music

                      In learning more about the various signals to be handled, one must
                      consider the peak power of various sources. For speech, the average
                      power is only about 10 microwatts, but peaks might reach a milliwatt.
                      Most of the power of speech is in the low frequencies, with 80 percent
                      below 500 Hz, yet there is very little power below 100 Hz. On the other
                      hand, the small amount of power in the high frequencies determines
                      the intelligibility of speech and thus is very important because that is
                      where the consonants are. The peak power of various musical instru-
                      ments is listed in Table 5-1.
   123   124   125   126   127   128   129   130   131   132   133