Interleaving...not video but audio... ???

[rasputin1963]

Playing with video is my fascination du jour. And naturally audio plays a big part in this...

 

YouTube is perhaps my immediate goal; I want to publish videos which have the highest video and audio quality I can muster, but still have them manageable in filesize and easy for YouTube to playback.

 

I was surprised today to see that ADOBE PREMIERE, when you export a movie, gives you some choices on how to interleave your audio track...

 

It offers choices like "one second", "half-second", "quarter-second", "eighth-second".

 

Gee, I dunno even what this is referring to. Can any of you explain? I've heard of interleaving video, but not audio.

 

 

Thanks, ras

[slight-return]

I believe they are talking about how the audio is laid into the serial stream of data that makes up the AV file.

OK, to read (and play back) a file you still are, at some point, reading serially (or probably more accurately, sequentially) -- so you are going to either have to come up with an encoding scheme that mashes everything together (which will also require more decode overhead which s going to slow stuff down) or write things in "chunks" (some video, then some audio, maybe a SAP, etc)

 

i think (grain of salt here, multimedia programming is not my area) what they are talking about is how long (and consequentillly how "often") the audio chunks are inserted between (interleaved) with chunks of video data...

so you get sort of a "buffer effect" because you read some audio, then some video -- hopefully by the time the audio is run out, the next chunk is there without having to wait

 

[additional clarificton : so dont thing about it like video interlacing - where you are interlacing vid dat with more detailed vid data, but interleaving different data streams...it's basically defining the MUXing scheme]