truncated-playback-and-restart__splice_main_splice_ad: 2 observations fail on 3 TVs and only pass on 1

[jpiesing]

From the December test run ..

Failing on 3 TVs and only passing on 1 is enough to justify some form of analysis, if only to check that the OF is correctly reporting failure.

[yanj-github]

I’ve reviewed the OF reporting and confirmed that it accurately reflects the failures.

One particularly interesting failure I came across is the following:
FAIL: First frame found is 10, expected to start from 1. First frame number tolerance is 0. Last frame found is 138, expected to end at 144. Last frame number tolerance is 0. Mid frame number tolerance is 10. Frames out of order 124, 1 in playout 1. Following frames are missing in playout 1: 35 36 1st presentation ending frame found is 37, expected to end from 125 to 144. 2nd presentation starting frame found is 49, expected to start with from 1. Total of missing frame count is 17.

The device plays frames 10 through 124 on A1, then loops back to frame 1 on A1 and continues until frame 37. After that, it switches to AD-A1 from frame 49. When it reached frame 138, the status changed to 'finished,' causing the OF to stop detection prematurely. As a result, it finished to frame 144, but the OF missed it due to the 'finished' status being triggered too early.