0

I am trying to understand deeply the pseudostreaming mechanism. There are two ways for requesting the partial content of a video:

  • Use a byte range request.

  • Use the start parameter. This second method can use a byte o a time
    range.

My understanding is that both the range request and the start parameter using byte requests work the same. The server does not need to read and understand the metadata on the video, whereas with start parameter using a time range the player does not need to fully load the metadata because it will just request a time to the server. The server should do all the dirty work and process the metadata to find the closest key frame.

The thing is that I have read some comments that lead me think this might not be like this. So I would appreciate if somebody with experience could briefly compare both techniques and the undergoing processes that happen in the web player and server.

And also why some techniques are used with MP4 whereas other happen for FLV.

Thanks in advance,

jay.end
  • 69
  • 1
  • 10

1 Answers1

0

Flash and HTTP streaming are a bit different whn it comes to seeking. I recommend reading this basic Wikipedia article to understand the behavior of seeking in both cases: https://en.wikipedia.org/wiki/Progressive_download

Yury
  • 671
  • 3
  • 12
  • Thanks for the reply. But I want to understand why a Flash player requests MP4 videos using time offsets and a HTML uses bytes. Flash players requests FLV using bytes... why do they use time stamps for MP4? – jay.end Oct 15 '14 at 08:50