Live video stops feeling live somewhere between half a second and two. A clinician asking a question, a bidder raising a hand, a classroom waiting on an answer — those moments need the picture to arrive while the sentence is still being spoken. This note is a placeholder you can edit or replace from Posts. It shows how a published article sits in the Vasl theme.
Why half a second is the line
Glass-to-glass latency is the time from a camera sensor to a viewer’s screen. Under about 500 milliseconds, a conversation still feels like one room. Past that, people talk over each other, bids land late, and “live” turns into a delayed broadcast. Vasl is built to stay under that line on real networks, not only on a lab switch.
Under about half a second, a conversation still feels like one room.
One server, the protocols you already use
Most stacks split ingest, transcoding, packaging, and delivery across vendors. Vasl keeps that path in one engine: ingest over RTMP, RTSP, SRT, or WebRTC, then play out to WebRTC when latency matters and HLS when you need reach. The same config runs on a laptop and in production.
- WebRTC for sub-second two-way video
- RTMP, RTSP, and SRT for cameras and encoders you already own
- HLS when a CDN and broad device support matter more than the last millisecond
- Recording and DVR when the session has to be reviewed later
Keep the stream on your network
Telehealth, surveillance, and enterprise all-hands often cannot send media to a third-party cloud. A self-hosted license lets you run Vasl on your own hardware or VPC, with signed playback, TLS, and access rules at the edge of your network. Publish the next note from Posts and it appears in this grid automatically.