Method and apparatus for digitally controlling gain in a talking path

A digital control circuit for controlling a gain of analog signals carried on a bidirectional talk path between two sources of audio signals, comprising:

• a first bidirectional communication port associated with a first source of audio signals;
• a second bidirectional communication port associated with a second source of audio signals;
• a first codec for receiving audio signals from the first bidirectional communication port and producing corresponding first digital signals, and for receiving digital signals and producing corresponding analog signals coupled to the first bidirectional communication port;
• a second codec for receiving audio signals from the second bidirectional communication port and producing corresponding second digital signals, and for receiving digital signals and producing corresponding analog signals coupled to the second bidirectional communication port;
• and
• a digital processor for receiving the first digital signals from the first codec and producing different digital signals representing first adjusted amplitudes of the corresponding analog signals and coupling the first adjusted amplitude digital signals to the second codec, and for receiving the second digital signals from the second codec and producing different digital signals representing second adjusted amplitudes of the corresponding analog signals and coupling the second adjusted amplitude digital signals to the first codec, whereby directional gain of the analog signals can be controlled.

A method of controlling gain in a talk path, comprising the steps of:

• receiving plural analog signal samples;.
• converting respective amplitudes of the plural analog signal samples to corresponding plural digital signals;
• integrating the amplitudes represented by the plural digital signals to determine an average composite amplitude;
• comparing the average composite amplitude with a predetermined target amplitude to determine if an amplitude adjustment is needed;
• if the average composite amplitude is less than the target amplitude, incrementally increasing the representative amplitudes of digital signals of subsequently received analog signal samples and converting the digital signals with the increased representative amplitudes to corresponding analog samples;
• if the composite amplitude is greater than the target amplitude, incrementally decreasing the representative amplitudes of the digital signals of subsequently received analog signal samples and converting the digital signals with the decreased representative amplitudes to corresponding analog samples;
• and
• transmitting analog signals corresponding to the converted analog samples to a destination receiver.

A method of carrying out automatic gain control in a digital communication talk path having a channel A carrying digitized voice samples and a channel B carrying other digitized voice samples, comprising the steps of:

• assigning a target voice level to channel A and a different target voice level to channel B;
• receiving digitized voice samples from a channel A codec and converting the digitized samples to a linear format;
• receiving digitized voice samples from a channel B codec and converting the digitized samples to the linear format;
• comparing representative amplitudes of the channel A digitized samples with the channel A target voice level to determine if the representative amplitude thereof should be increased or decreased, and determining a first gain factor to achieve the channel A target voice level;
• comparing representative amplitudes of the channel B digitized samples with the channel B target voice level to determine if the representative amplitude thereof should be increased or decreased, and determining a second gain factor to achieve the channel B target voice level;
• multiplying the first gain factor with the second gain factor to determine if the product is greater than unity;
• and
• modifying at least one of the first gain factor or the second gain factor so that the product of the resultant gain factors is not greater than unity.