All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=853499

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=d1de8a1abbe2a38e67fa802a0c17f91631f33aebfe25ad48cd71a6f001008fe6


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

chromium-rel-win10

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/149e30d5240000

📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/149e30d5240000

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/12a69543240000

^^^ New run with wider range.

📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/12a69543240000

Over to speed > bisect: graph shoes a clear sustained regression, but pinpoint isn't finding it.

> clear sustained regression

(vs the ref, which held steady)

FYI:  Issue 853498  tracks a very similar regression (same range and bot, different test) which is also not found by bisect

+dtu

Dave, can you help them out?

Similar to  issue 853498 , I'm going to re-run the Pinpoint job with no --story-filter.

+perezju: Soundwave noise analysis
+johnchen, crouleau: benchmark owners
FYI: looks like this benchmark/metric produces very different results when running the full benchmark vs the individual pages.

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/12f7da07240000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/12f7da07240000

Field trial for AudioService experiment by olka@chromium.org
https://chromium.googlesource.com/chromium/src/+/8d891be67dac6be0f01a3c91e56ca9729101782c
197 → 242.2 (+45.21)

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions

Filed issue 859450 to investigate the issue of the benchmark giving different results with/without story filter.