[ExtractStream] Explanation of records in ty chunks?

[Dale Reed]

willieb9000@y... wrote:

> ok, guys, I am looking at display stream. On the left in black are
> the records for the chunks. The first group of 4 bytes is the size
> and the type. WHat are the second group? I know it has something to
> do with the sequence, but what exactly do they tell us? The third
> group seems to be simply telling whether the record refers to real
> data or not and what does the 4th group do or tell us?
> 
> If it help, use these 2 examples:
> 
> 00 22 42 E0 30 22 69 54 00 00 00 B0 AD D2 5F 35 
> 00 01 06 E0 30 00 8D FC 00 00 00 B0 AD D2 5F 35 
> 
> I am trying to figure out how the programs decide what sequence they
> are expecting next. I can find the offending chunk in my streams, but
> I can't tell why SS doesn't like it.


If you grab ConvertStream from the tympeg zip, you can run it in
-v mode. It will give you a whole bunch of debug on each chunk,
including header, contents, etc. Its what I built/used when I was
trying to figure out why convertstream was failing on some
streams. Its very verbose, though. :(

http://www.ntworld.com/tivo/tympeg.zip

As an aside, I started working on a tivo application that is pretty
much a stripped down extractstream. My goal is to have a "debug"
tool that knows of the sequencing and errors and can give us physical
information in the FSIDs where these happen. If we can figure out
why bad data is getting into the streams and how to handle it, then
we can get on with the fun stuff. :)

Dale