All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=770196

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=9d3989415dd3f7ce3a9b4488e71326bfcb8a49538c58d811731b99445f7d1e8e


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

webrtc-linux-large-tests
webrtc-mac-large-tests
webrtc-win-large-tests

OpenH264 codec was updated to v1.7 in that commit.

The jump of decode_time from 0 to 1ms is not critical if we will consider absolute numbers. I believe that decoding time increased by fraction of millisecond but rounding in windows tests doesn't allow us to see that (no rounding in linux tests - https://chromeperf.appspot.com/report?sid=3a6ee65d7196a1778ebba1253e62b57e1b2a2c59f598a856eef9b9935f0f509d&start_rev=19844&end_rev=20061).

Another degradation is in frame rate/drops. From changes in PSNR and FPS I can conclude that new codec favors frame quality over frame rate: PSNR is higher, FPS is lower. Since frame rate decreased by just 1% and PSNR increased on 0.5-1dB I would consider this as improvement in overall.