All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=625629

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgkoPPuAoM


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

webrtc-mac-large-tests

Hey aluebs, just FYI, seems like a mac perfbot flagged your change in https://codereview.webrtc.org/2053773002 as improving performance a bit.

Unclear why this metric was improved.

This is actually expected, since after this fix only the necessary channels are merged after the beamformer reduces the number of channels. Here is the description of the CL:

"Before this change the ChannelBuffer had a fixed number of channels. This meant for example that when the Beamformer would reduce the number of channels to one, the merging filter bank was still merging all the channels, which was unnecessary since they were not processed and just discarded later. This change doesn't change the signal at all. It just reflects the number of channels in the ChannelBuffer, reducing the complexity."