Digigram IQOYA *LINK User Manual page 91

APPENDIX E: TYPICAL LATENCY VALUES
The back to back latency between two IQOYA*LINK devices depends on the selected audio format, the
network quality, and the enabled functionalities (backup, half/full duplex, FEC).
See the table underneath for maximum latency values in half-duplex, using neither failover configurations nor
FEC, with the jitter buffer size set to 0, and with an optimized network.
Audio Type
PCM
MPEG
AAC-4 LC
AAC-4 LC+SBR
AAC-4
LC+SBR+PS
AAC-4 LD
AAC-4 ELD
Apt-X
All measurements taken on stereo samples at 48 kHz
Impact of the processing granularity
Add about 4 ms to the latency each time the processing granularity is increased of 1ms.
Impact of the network on latency
Latency highly depends on the quality of the network. Network jitter and packets loss typically have a direct
impact on latency.
Network jitter compensation is achieved by buffering audio data on the decoder. A good quality
●
network generally offers a low jitter, then requiring low buffering on the decoder, which means a low
increase of latency. But a network with a high jitter requires increasing the decoder buffering
accordingly, leading to a significant increase of latency.
In case of packets loss on the network, it is necessary to enable an FEC, which allows recovering lost
●
packets thanks to redundant frames. FEC increases the latency.
Impact of features on the latency
The amount of features used in IQOYA directly impacts the latency. For a given audio format, the lowest
latency is obtained in half duplex mode, with no backup defined and no FEC. As soon as one of these
features is used, the latency increases a bit.
Audio format
(processing
granularity set to
Automatic)
24 bit
256 kbps
128 kbps
96 kbps
48 kbps
160 kbps
160 kbps
2 ​ 4 bits
Latency Latency
(processing
granularity set to
1ms)
43 ms
80 ms
100 ms
225 ms
250 ms
70 ms
50 ms
40 ms
90
Description
9ms
54ms Layer II, III
78ms
180ms aka HEv1
220ms aka HEv2
39ms
34ms
12ms