Information Technology Reference
In-Depth Information
Analog-to-Digital Signal Conversion
For clear voice communic ations, analog speech is conver ted in to dig ital format. Dig itized
voice can travel longer distances than analog voice. The steps involved to covert voice from
analog into digital format are
Step 1.
Filtering
Step 2.
Sampling
Step 3.
Digitizing
Most of the spoken language ranges from 300 Hz to approximately 3400 Hz. In the first
step, codecs are configured to filter signals over 4000 Hz out of the analog signal.
In the second step, the signal is sampled at 8000 times per second using pulse-ampli-
tude modulation (PAM). It is sampled 8000 times a second because that is twice the
highest frequency of the filtered voice stream at 4000 Hz. This produces a sample every
125 microseconds.
Third, the amplitude samples are converted to a binary code. This is process where PCM
occurs. The difference between PCM and PAM is that PCM does the additional step of
encoding each analog sample into binary.
The digitizing process is divided further into two subprocesses:
Companding:
This term comes from “compressing and expanding.” The analog sam-
ples are compressed into logarithmic segments. Then each segment is quantized and
coded, which is the next subprocess.
■
Quantization and coding:
This process converts the analog value into a distinct
value that is assigned a digital value. The standard word size is 8 bits, which allows for
256 distinct quantization intervals. The rate then becomes the sampling rate times the
size of the codeword (2 * 4kHz * 8 bits = 64 kbps).
■
Codec Standards
Codecs transform analog signals into a digital bit stream and digital signals back into ana-
log signals. Figure 14-14 shows that an analog signal is digitized with a coder for digital
transport. The decoder converts the digital signal into analog form.
Analog Input
Signal
Digital Coded
Signal
Analog Output
Signal
Coder
Decoder
Figure 14-14
Codec
Each codec provides a certain quality of speech. Each codec provides a certain level of fi-
delity to the original audio, or quality of speech. The term
mean opinion score
(MOS) is
used to rate the fidelity for a codec. A MOS score is not a scientific measure. Instead, it is
dio fidelity from 1 (bad) to 5 (best). The scores are then averaged to provide the MOS for
