Extron electronics VN-Matrix 300 User Manual page 47

By using both FEC methods together, 2 dimensional FEC, the combined effect can protect
against more loss permutations than either scheme used in isolation.
FEC works by adding additional data to the stream. Using this data it is possible for the
decoder to detect and reconstruct any lost packets.
P1 P2 P3 P4
P5 P6 P7 P8
P12
P9 P10 P11
P16
P13 P14 P15
P17 P18 P19 P20
F 1 FC2 F 3 F 4
C C C
P1 P2 P3 P4
P5 x P7 P8
P12
P9 P10 P11
P13 x x x
P17 P18 P19 P20
F 1 F 2 F 3 F 4
C C C C
Setting the FEC value
The values that are set in the FEC row and column drop down lists will depend on the quality
of the network link that is in use.
From the diagrams above it can be seen that the FEC data can add a large overhead to the
overall bit rate that is transported over the network. Therefore certain settings, whilst they
may provide a high level of reliability, will be inefficient. Settings that add little overhead may
not provide a sufficient level of protection on networks with a poor QoS.
The recommended setting is 5,5. This setting should be used as the starting point for the
error correction for the audio transport and provides rugged operation on uncorrected IP
networks.
NOTE:
FEC is ONLY applied to the ancillary data channel. It is NOT applied to the
•
video channel.
FEC is only effective for RTP transports. It should not be used for TCP
•
transport.
Error injection rate
This setting should be left at 0. It used for test purposes only and is used to drop packets
artificially in order to test the effectiveness of the FEC scheme.
The diagram to the left shows the FEC scheme
F 1
R
applied to a 4 (column) x 5 (row) structure. Note
F 2
that this adds 5 row FEC packets and 4 column FEC
R
packets to every 20 data packets. This scheme can
F 3
R
protect for burst packet losses up 5 packets long.
F 4
R
F 5
R
F 1
If packet 6 is lost or corrupted, this can be corrected
R
for using the F 2 FEC packet.
F 2
R
R
F 3
It is not possible to correct packets 14, 15, and 16
R
using the row FEC (F
F 4
R
F 5
R
To correct these packets, the column FEC data is
used. Packets 15 and 16 may be corrected using F
and F 4 column FEC packets. To correct packet 14,
C
packet 6 must first be corrected using the row FEC.
After this, the column FEC may be used to correct
packet 14.
VN-Matrix 300 • System Configuration
4) as too much data is missing.
R
3
C
41