Transmission Delay - Polycom SoundStructure C16 Design Manual

All echo cancellers have some upper limit, called the tail time, after which they do not recognize the echo
signal as a version of the reference signal and consequently can not remove the echo from the room. The
tail time is measured in milliseconds (although can be interpreted in terms of distance) and should be greater
than 100 msec for medium sized rooms and greater than 200 msec for larger rooms. As shown in the
following figure, if the room is lively, the length of an echo path may be longer than expected making the
room appear acoustically larger than it is physically.
Microphone
Loudspeaker

Transmission Delay

In all communication networks, there is a combination of processing latency and network latency that delays
when the local signal is heard in the remote room and conversely when the remote signal is heard in the
local room. There have been numerous studies that have shown that when delay is added to an echo, the
perceived communication impairment caused by the echo is increased. Processing latency is due to the
selection of algorithms, the speed of the underlying processors, the implementation of the algorithms, and
how the audio is collected from the analog-to-digital converter and also sent to the digital-to-audio converter
in a digital system. In some cases, such as in video conferencing systems, delay may be intentionally added
to the audio signal to compensate for the delays inherent with video signal in order to maintain audio and
video lip synchronization. Network transmission delay is limited by the speed of electrons (or the speed of
light with satellite and microwave transmission) in a network and any other processing or data handling that
may occur in the data as it is transmitted through the network. It is not uncommon for networks to require a
hundred milliseconds or more to transfer audio from one site to the other.
Polycom, Inc. 726