
GS1582 Data Sheet
40117 - 1 November 2007
38 of 114
4.7.3 Audio Modes of Operation
The audio multiplexer operates in three distinct modes: normal mode, cascade
mode, and audio group replacement mode.
In normal operating mode, up to two audio groups can be added to the output
video. (See
Table 4-3: Audio Multiplexer Modes of Operation
). All existing audio
packets are deleted from the video input, however arbitrary packets, SDTI packets
and SMPTE 352M packets are not deleted.
When the EN_CASCADE host interface bit is set HIGH, the audio multiplexer
operates in cascade mode. Up to two audio groups can be added to the video
output. The added groups will not replace existing embedded audio groups. No
existing packets are deleted from the video input, and the added audio packets are
appended to the last packet in the video input.
In cascade mode, if the audio multiplexer is configured to add group 1 audio data
packets with the same group number as an existing group, MUX_ERRA will be
asserted in the host interface. Similarly, MUX_ERRB will be asserted if the
configuration leads to replacing existing groups with group 2.
When the AGR bit in the host interface is set HIGH, the audio multiplexer operates
in group replacement mode. In this mode, added audio groups can replace existing
embedded audio groups. The embedded audio groups will be sorted by audio
group number. No packets will be deleted from the video input (except for the audio
packets being replaced). SDTI packets and SMPTE 352M packets are placed
before the audio packets, and arbitrary packets are placed after the audio packets
in SD mode.
NOTE 1: When audio group replacement mode is selected, the cascade function
is disabled.
NOTE 2: In normal SD mode, if the incoming stream has a completely full HANC
space, the outgoing HANC space will only get filled with up to 235 words and the
remainder of the space is left blank. This will not occur if there are one or more
words free in the incoming HANC space.
NOTE 3: SD audio in cascade mode may only be inserted if the HANC space of
the incoming stream is not completely full. If the incoming stream has a full HANC
space, audio packets will be inserted beyond the HANC boundary. The space
following the inserted audio packets will then be blanked until the next EAV is
detected in the video stream. This will not occur if there are two or more words free
at the end of the incoming HANC space after the existing packets.
NOTE 4: HD audio in normal mode may only be inserted if the HANC space of the
incoming stream is not completely full. If the incoming stream has a full HANC
space, packets will be inserted beyond the HANC boundary. The space following
the inserted packets will then be blanked until the next EAV is detected in the video
stream. This will not occur if there are two or more words free at the end of the
incoming HANC space after the existing packets.