Once you have obtained a token from your render request, you can use it to render and retrieve the video. You construct the render URL from the token, a render mode and an extension. There are three render modes which control how the rendered video is transferred to a client. Each mode has distinct traits in respect to responsiveness and latency.
|render||Download||The entire movie is rendered on the server first and then delivered in one piece as fast as possible. This mode has the highest latency but zero probability of pausing in mid-stream.|
|get||Progressive Download||Starts streaming as soon as possible, render and stream as fast as the system and the connection can. Medium latency, lowest probability of pausing. Default render mode,|
|play||Streaming||Starts streaming immediately, render only the frames necessary for the current time position and stream according to user bandwidth with minimal buffer. Dynamic content is fetched as late as possible. Usually the delay in fetching an external resource is not noticeable, because it only affects one frame, and the player has some buffer to compensate. Lowest latency.|
Not all combinations of Render Modes and Extensions are possible or make sense. See the following table for valid combinations:
|Extension (Container Format)|
|Mode||mp4||webm||flv||HLS (m3u8)||DASH (mpd)||jpg, tif, png (Single Frames)|