I get that some scenes are easier to compress than others, so their encoded sizes can vary. How does a stream still play smoothly instead of pausing whenever the data size changes?
How does video streaming avoid jitter when some frames compress much better than others?
Asked by
C9~Gromlet
Answers
The main trick is buffering. A player downloads data ahead of the point where it will be shown, then uses that cushion while network delivery and decoding speed vary. If a segment is small, it does not have to be displayed early. The video’s timestamps still determine when each frame belongs on screen.
Reply to this answer
Frame sizes can vary without the encoded stream becoming wildly unpredictable. Many streaming encoders aim for a target bitrate over time, trading off compression efficiency and visual quality to keep the data rate within a useful range. A difficult, detailed scene may look softer or more blocky than an easy scene, while a simple scene can use fewer bits.
Also, video compression often uses information shared across neighboring frames, not just a separate compressed image for every frame. The amount of data still changes, but the encoder can manage that variation rather than insisting on identical frame sizes.
Reply to this answer
It helps to separate variable frame size from playback timing. The player schedules frames from timestamps, not from how many bytes each one happened to need. Network throughput, buffering, and decoding workload all affect whether playback stays smooth, but a smaller compressed frame by itself does not make the next frame appear sooner.
If delivery slows down long enough to exhaust the buffer, then you can get a stall. Smooth playback is not guaranteed for every network or every video. It is the result of encoding choices and enough data arriving ahead of playback.
Reply to this answer
Compression also does not always make a frame smaller in the same way. Lossless compression cannot shrink every possible input, while typical streaming video uses lossy techniques that can spend fewer bits by discarding detail. A flat-color scene is often cheap to represent, but that mainly changes the amount of data to send, not the intended time each frame is shown.
For live calls or cloud gaming, there is less room to build a large buffer because delay matters. Those systems have to make a different tradeoff, so variations in network delivery can be more noticeable.
Reply to this answer
Add your answer
Answers are reviewed before publication.
C9~Gromlet:
So the player is keeping a little bit of the future ready, rather than showing each frame as soon as it arrives?
j5!Gudge:
Right. The size of that cushion depends on the service and playback situation. A longer buffer smooths out more variation, but adds delay before playback or when seeking.