Specifications

fiber at even closer to the speed of light. Hubs and other intermediate devices have their own propagation delay

and slow the data down by no more than one or two microseconds.

The receiving RAVE unit, when it sees a data packet with the identifier of the network channel it is set to, pulls

that packet into its buffer, separates the data into all eight audio channels, and feeds the data directly to the

outputs, if it’s a digital unit, or to digital-to-analog converters if it’s an analog one. And it does this 48,000 times

per second also.

All this happens in 6.3 milliseconds. You might look at the latency specification and wonder how it will affect your

system. It’s a perfectly legitimate question. To determine this, let’s first look at what the time interval equates to.

SOUND IN FREE SPACE

At an ambient temperature of 20° C (68° F), sound travels at a velocity of approximately 343 meters (1125 feet)

per second, or 34.3 cm (13.5 in) per millisecond. Therefore, the delay inherent in a RAVE network is equal to the

time it takes sound to travel 2.16 meters (7 feet 1 inch). If you use a RAVE network in a system for sound

reinforcement, cinema sound, stadium/arena sound, etc., you might need to consider this delay when you design

and set up the system, especially if you use any of the same audio program, routed both via a RAVE system and

by direct wire, to cover any of the same acoustic space.

For example, a large concert sound system uses both a main speaker system for the near audience and a system

of speakers for the outer reaches of the audience, delayed 150 milliseconds so it will be coincident with the

sound from the main speakers. It uses a RAVE network to distribute sound to the delayed system, but not to the

main system. The system engineer, therefore, will set the delay units for the outer system at 143.7 milliseconds,

because the RAVE network itself has 6.3 milliseconds of delay.

SYNCHRONIZATION WITH VIDEO

When using audio and video together, it is vital that they stay synchronized with each other. Otherwise, the

embarassing effects of people’s voices not matching the motions of their mouths, or of sound effects not

coinciding with the visual clues of their events, could occur.

NTSC video uses 525 lines per frame at a frame rate of 29.97 Hz, so each line of video is 63.56 microseconds

long. PAL and SECAM video use 625 lines at a frame rate of 25 Hz, so each line is 64.0 microseconds long. Thus,

the delay caused by a RAVE network’s latency represents only 99.1 lines of NTSC video, or 98.4 lines of PAL or

SECAM video. These are respectively less than 1/5 and 1/6 of a frame, and therefore will not cause noticeable

loss of synchronization between the visual and the audio programming.