Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Dual Mode Block Compound PCM

IP.com Disclosure Number: IPCOM000079959D
Original Publication Date: 1973-Oct-01
Included in the Prior Art Database: 2005-Feb-26
Document File: 2 page(s) / 14K

Publishing Venue

IBM

Related People

Croister, A: AUTHOR [+2]

Abstract

The bit rate of PCM transmission can be substantially reduced by replacing instantaneous companding with block companding. In this approach, the use of a dual-mode encoding permits the achievement of a better coding efficiency. The proposed scheme is compatible with existing instantaneous companding systems from which it can be translated easily.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 53% of the total text.

Page 1 of 2

Dual Mode Block Compound PCM

The bit rate of PCM transmission can be substantially reduced by replacing instantaneous companding with block companding. In this approach, the use of a dual-mode encoding permits the achievement of a better coding efficiency. The proposed scheme is compatible with existing instantaneous companding systems from which it can be translated easily.

The code is based on a floating-point representation where the exponent is kept unchanged during a "block" of 16 samples. When transmitted through CCITT standard multiplex frames, superframe synchronization may be used to identify the groups of 16-sampling periods. The transmission format is shown on Fig. 1: A A A A A A A A A A A A A A A A B B B B B B B B B B B B B B B B C C C C C C C C C C C C C C C C W X Y Z D D D D D D D D D D D D

Fig. 1: Transmission format, each vertical represents one time slot of 125 mus.

In each time slot, bits A, B, C and D represent the value of a sample coded linearly with a scale factor indicated by XYZ. A flag bit, W, indicates whether the samples are coded conventionally or differentially encoded. The four first time slots have no " " bit and are, therefore, coded with a lower accuracy. Upon reconstruction of the original signal, the missing D bits are replaced with pseudo- random data, e.g., from a cyclic code generator.

It is well known that for speech signals, pitched sounds have a relatively large amplitude and a low-frequency spectrum while nonpitched sounds have a relatively small amplitude and a high-frequency spectrum. Moreover, quantization noise is particularly detrimental for pitched sounds, while it is more acceptable for nonpitched sounds where quantization noise is more or less buried within the colored noise structure of the signal. Under these conditions, a differential encoding scheme is more efficient for the transmission of pitched signals, while a conventional encoding is better suited for the transmission of nonpitched signals.

The algorithm for selecting the encoding mode within a frame consists in subtracting each frame sample to the preceding one. This results in a new frame made of differentially encoded samples. The largest differentially encoded sample, S/1/p, is com...