All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=760551

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=d146159bc9d524fc1fdd38a7277d39b33f5fc0d70dde7010b749e5f0e1e569dc


Bot(s) for this bug's original alert(s):

webrtc-mac-large-tests

Can you please have a look and see if https://codereview.webrtc.org/3003673002 could be the cause for the perf changes. 

It really shouldn't have any real impact on these tests. The change effectively enables content type extension. Even if it had some measurable impact on the metrics, it should've be consistent. But here we have some improvements in sender_time together with regressions in different tests. Also cpu_usage shouldn't go down. 

Finally, either graph's tooltips are off by 1 commit, or changes actually correspond to that commit:
https://chromium.googlesource.com/external/webrtc/+/c3c125662e4cc4776f1d4fb76a030504213011dc