QUOTE(mfa682 @ Mar 21 2021, 07:00 AM)
I guess fast.com and youtube work differently under umo mobile data network
Hypothesis:
fast.com will "randomly" select the Netflix server "at the moment" of speedtest is launched. It could show the best or average or worst result "at that particular" moment. If we do the speedtest every five (5) minutes interval, for sure the results will be "significantly" different between each other becoz each test "may" connects to different server(s) independently ...
OTOH, YT will consistently "optimise" the streaming as and when needed so long as the network signal is stable until for some reason the network signal drops and the streaming is buffering every now and then. That's why the readahead value is important to know whether the YT streaming is being auto-optimise for selected network or not...if the readahead value is low (less than 5 seconds), meaning the
network cell tower is poorly optimised or could be affected by congestion issue as well... Just like the example that me showed earlier, even though the signal is full bar/strong, unfortunately, at the cell tower and at that moment, congestions hit the network and thus the speedtest is average. However, due to the good optimisation of cell tower/network (at that particular location), the YT experience is still buffer-less for 1080P@60fps...
This post has been edited by azhar75: Mar 21 2021, 11:25 AM