Exact semantics of IVS Player getLiveLatency()?

Question

What exactly is getLiveLatency() measuring for a live stream?

https://aws.github.io/amazon-ivs-player-docs/1.7.0/web/interfaces/player.html#getlivelatency

Given:

content recorded (A) --> content received at AWS IVS ingest endpoint (B) ---> presumably some processing in AWS (C) ---> content available for a viewer to view on AWS IVS servers (D) ----> viewer downloads live data, a HLS segment (E) ---> viewer watches segment (F)

getLiveLatency gives the difference between which of the following?

1: A and E?
2: B and E?
3: D and E?
4: Other?

How is that number actually being computed?

Answer

Thank you for using the Amazon Interactive Video Service. The IVS Player's getLiveLatency method (https://aws.github.io/amazon-ivs-player-docs/1.7.0/web/interfaces/player.html#getlivelatency) measures the latency between when the content is received by Amazon IVS (B) and when the viewer watches the segment (F). The delay/latency before the content enters the IVS system is unknown to IVS, thus it cannot be calculated into the final live latency value.

Exact semantics of IVS Player getLiveLatency()?

Relevant questions

is IVS trailing pathname is fixed?

AWS IVS player sample code to display Status, Health, Duration and Viewers

IVS - live streaming issue - Starvation start