You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
apparently the RTMP server is setup to create 20s fragments so i am going to assume that's where most of the latency comes from. is there a good reason not to go lower with the fragments length?
It's the minimum we can go based on how fast IPFS can pin. If we go much lower we build up a queue IPFS can't consume fast enough. 20s is what we found acceptable.
We have touched on the topic of how real-time this stream is, so I want to start documenting here.
Playing from
ipfs-server
m3u8 (origin):Playing from m3u8 on
ipfs-mirror
:Playing from m3u8 on
ipfs-mirror
but ts chunks from ipfs.infura.io:In all cases observed latency is between 1 to 1.5 min, although I have seen as much as 2 min occasionally.
The text was updated successfully, but these errors were encountered: