In order to compare latency achievements of the HESP protocol, we designed a test scenario where HESP is compared with CMAF-CTE, using the same encoding settings, and a selected set of chunk and segment sizes for the CMAF setup (meaning a changing GOP size).
Both encoded signals are sent over RTMP to a packager on AWS. Both signals are delivered into players running on Chrome version 73. The camera was pointed towards a millisecond accurate clock, for which the latency was compared with the video displayed in the player.
Following test setup/settings were compared:
- CMAF-CTE with 1s segments and 1 frame chunks
- CMAF-CTE with 2s segments and 5 frame chunks
- CMAF-CTE with 6s segments and 5 frame chunks
The following results were measured.
As the results of this test indicate, end-to-end latency for the HESP protocol is significantly lower compared to all CMAF-CTE configurations which were tested. The latency measured of 330ms is low enough to enable new interactive formats, where for example viewers can directly interact with the TV broadcaster.