Your last comment is correct for us.
but when it works…
To the USA and to NZ and to the UK.
Edit - all of the international stuff is non commercial media. A mix of 1080 and 4K. Multichannel audio.
Hiya, @ChuckPa so I have to highlight a discrepancy from previous discussions and I’m unable to PM you directly and I feel the discrepancy is substantial enough to warrant bringing up
The test case with the settings provided would very literally be impossible to achieve a steady playback within the current limitations of plex’s approach with a raw new, previously uncached 100% network stream.
But as an additive to this, there does appear to me some more external factors that can take place internationally, not just raw routing and agreements between ISPs, but also peak hours of congestion for various parts of the world at various times, but all these things can and should be fixed.
To reiterate the topic title here in perhaps a more understandable way (I hope), and the thing a lot of us plex users are experiencing and would like plex to look into fixing is:
The current transfer protocol plex is using to deliver video has a flaw when exposed to certain conditions, one of which is definitely latency, that essentially the amount of data that plex sends to clients is sufficient for a certain amount of playback, and between each chunk the plex server awaits a response from the client to confirm whether the data was received or if it is ready for the next bit of data, in that await lies a problem that can be exacerbated by various external factors which causes playback to exceed the amount of playback data. There’s a sweet spot where this becomes an issue as the amount of data seems to be chosen dynamically by the plex server maybe when negotiating how much RAM the client can use for cache(?)
It’s a real problem and there’s countless posts across the plex forums and reddit asking for this to be looked into. Yes it’s not super widespread (at least to my knowledge) but also it can be difficult to measure as plex by default sets the remote stream quality to 720p, which is usually flawless playback [edit: it looks like the new default is 1080p@12mbps, so perhaps this problem will end up being more mainstream for laymans users in the coming months?] - though I don’t think playing a 4k film from your home if you’re on holiday across the side of the world should be impossible, and honestly I figured that was one of the main purposes of plex.
I had previously given up on this point but I’ve received another 2 messages from users experiencing the same issue traced down the exact same path asking me for a solution for other clients that aren’t windows desktop exclusive.
Please, we just want this fixed.
It’s a shambles that this is still a bug that exists. Latency being a way to induce this behaviour doesn’t exclusively make it a latency issue given that it happens over local network for a large swath of users and alterating the client’s buffer behaviour (windows only still and requires the desktop client to modify MPV config…) seems to alleviate it entirely.
Logs shared, recordings shown and still plex staff just brush it under the rug. Hundreds of users experiencing the issue, many complaints and no solutions.
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.
