I am new to Video SDK and doing a tech spike in preparation for a migration away from Twilio Programmable Video (EOL Dec 2024).
We run a SaaS platform for communication analysis and require individual audio recordings for each participant, which are synchronized to at best <100ms, otherwise our analysis is messed up. We achieved this with Twilio via the compositions feature which guarantees synchronization of all streams of all participants.
However, in the Zoom Recordings API:
The recording_start, recording_end values in examples are all accurate to the second, not the millisecond. This means when we cross-compare points in time across two participant recordings there are potentially up to 999ms out of sync.
How can Zoom Video SDK provide us with millisecond level timestamps on individual participant audio?