S3 Server access logging vs CloudTrail logs

Question

Given that both services are enabled (A single S3 bucket with Server Access Logging enabled and CloudTrail with object-level logging enabled for that bucket):
1. What events will initiate logging from both services?
2. In such a case, what data will one service contain that the other will not?
3. What events will result in a log created by only one of the services?

I am having a hard time understanding the logical difference between those two, as both support object level logging.

Answer 1

This is summed up in the AWS documentation here. Note that AWS recommends CloudTrail.