Miata Mailing List: February 1996, Message #63
sponsored by



     


   From: Mike Lastort 
Subject: Re: 10 megabytes????
   Date: Fri, 2 Feb 1996 08:33:20 -0500

I tried to send this once (make that twice now!) already, but the server 
bounced it. Here's another try.

Kerim@aol.com wrote:
>
> In a message dated 96-01-30 23:42:56 EST, keith writes:
>
> >Just to throw in a data point, one minute of CD audio takes about 10 Mb of
> >storage. Thus your 5 second buffer needs to be about 800 Kb.
>
> Hmm, I'm a little skeptical on those numbers. Maybe I'm not remembering my
> sample rates to data space conversion, but I am pretty sure that (with modern
> compression routines) 1 minute <> 10Mb, so 5 seconds <> 800Kb.

I'm not sure who posted the original message, but 10MB/minute for CD quality
audio is about right. CD quality is defined as 44,100 samples of 16 bit stereo
audio per second. That's this many bits per second:

44,100 * 16 * 2 =  1,411,200 bits per second (bps)
(It's multiplied by 2 for stereo. It would be 44,100 * 16 for mono.)

Now figure out how many bits were talking in a minute:

1,411,200 * 60 =  84,672,000 bits per minute

There are 8 bits in a byte, so that gives us:

84,672,000 / 8 = 10,584,000 bytes per minute.

Now divide that value by 1024 to get the number of Kilobytes:

10,584,000 / 1024 = 10335.9375 Kbyte per minute

Divide again by 1024 for the number of Megabytes:

10335.9375 / 1024 =  10.09368896484 Mbyte per minute.

This also implies that a 5 second CD quality sample is:

1,411,200 * 5 / 8 / 1024 =  861.328125 KBytes

There is no compression with CD quality audio. All audio and video compression
algorithms are lossy (which means they lose data when compressed). True CD
quality cannot be compressed, because as soon as you lose data, you're not at
CD quality anymore.

Also, that rate of 1,411,200 bps is nearly DS-1 (often erroneously referred to
as a T-1 connection) speed. DS-1 is 1,544,000 bps. Therefore, for any real time
transmission of CD quality audio across the Internet, the connection from point
to point has to be DS-1 (T-1) all the way.

> After all, if your numbers are right, how can a DiscMan provide a *10*
> second buffer on a portable unit?

I'm not exactly sure I understand your question. The portability of the unit
has no bearing. A CD is nothing but a storage medium. The DiskMan is nothing
but a device to access that medium. After all, laptop PC's with CD-ROM readers
can access a CD-ROM just as effectively as a desktop PC with a CD-ROM reader
can.

Miata content: More snow on the way. Damn! I like my Rodeo, but I need some top
down Miata weather soon!

----------------------------------------------------------------
|                          | "Smell the burning flesh. Taste   |
| Mike Lastort             | the tangy sulfur air. Volcano     |
| lastort@access.digex.net | season."                          |
|                          | - A Haiku by Moltar, engineer in  |
|                          |   SpaceGhost Coast to Coast       |
|--------------------------------------------------------------|
| GCS d- s+:++ a C++++(--)$ US+ P L+ E-- W+(-) N++(-) K  w+++  |
| O-- M-- !V PS+++ !PE Y+ PGP+ t++(--)@ 5+++ X+++ R(-) tv+ b++ |
| DI++ D+ G-- e+ h-(--) r+++ y+++                              |
----------------------------------------------------------------