cta-wave / device-playback-task-force

9 stars 0 forks source link

definition of "start-up delay" and how to observe it #90

Closed jpiesing closed 1 year ago

jpiesing commented 3 years ago

Most tests in the DPCTF spec have an observation something like this;

The start-up delay should be sufficiently low, i.e., TR [k, 1] – Ti < TSMax.

There is no definition of "start-up delay" other than the formula given.

From https://github.com/cta-wave/device-observation-framework/issues/27#issuecomment-869784306 , @louaybassbouss said "HTML5 players usually show the first frame even if you don't call play().".

What is start-up delay & how should it be measured?

One possibility would be to measure the time between when play() is called and when frame no. 2 is displayed and then correct for one frame interval.

Whatever we do, the DPCTF spec will need changing because using "TR [k, 1]" is clearly wrong regardless.

haudiobe commented 2 years ago

DPC Editing call implemented in v1.0.8

for video 2) The start-up delay should be sufficiently low. As user agents may pre-load the first frame, the time to first frame is not relevant, but what is relevant that once hitting play, the second frame is rendered within the considered start-up delay. Hence, TR [k, 2] – Ti < TSMax + 1/framerate

for audio, nothing is defined yet.

yanj-github commented 2 years ago

Thanks @haudiobe however it does not cover all different cases. We propose the following:

The start-up delay should be sufficiently low. As user agents may pre-load the first frame, the time to first frame is not relevant, but what is relevant is that once hitting play, the second frame is rendered within the considered start-up delay. In addition, there may be missing frames at start up. Hence, TR [k, x] – Ti < TSMax where x is the first detected frame change after the play() event.

haudiobe commented 2 years ago

DPCTF 2021/12/15 ok with this update.

gitwjr commented 1 year ago

@gitwjr Review current version of the spec to ensure start-up delay is correct and is used consistently in the spec.

gitwjr commented 1 year ago

Implemented in v1.38 of spec.