Re: Data Compression

Marc Lindahl (shark@dti.net)
Sat, 6 Apr 1996 00:47:56 -0400


>V42 is NOT compression, it is error correction, V42Bis is compression
>and error correction. (Fine point I understand)

I stand corrected. I'm so used to saying V42 as shorthand -- does anyone
actually use the error correction without the compression?

>actual expends time. This TIME delay acts as if it is slowing the data
>throughput. This is also true if you are sending data that has been compressed

Not neccessarily. I know the AT&T and Rockwell chipsets have a different
processor to do the V42bis stuff, so it actually occurs in parallel. In
fact, I believe that it's the time it has while the modem part send the
data that it uses to decide if it can compress (i.e. if it can't decide and
do something by the time the buffer's almost done, it gives up).

>> With video/audio data like CU-seeme, it's useless, but for other
>> peripheral data, who knows?
>
>Not true. Video data is not random. It has very obvious patterns.
>Compression works by looking at repeatable patterns and reducing the

Right, but CU-seeme does compress the audio and video, no?

_/_/_/_/ _/ _/ _/_/ _/_/_/ _/ _/
_/ _/ _/ _/ _/ _/ _/ _/ _/
_/_/_/_/ _/_/_/_/ _/_/_/ _/_/_/ _/_/
_/ _/ _/ _/ _/ _/ _/ _/ _/
_/_/_/_/ _/ _/ _/ _/ _/ _/ _/ _/
shark@dti.net (Marc Lindahl)
<A HREF="http://home.dti.net/shark">My Home Page</A>
<A HREF="http://www.ariel.com/~bowery">Visit my Bowery Records site</A>