Posted by Andrew Reilly on 11/14/12 11:28

On Wed, 05 Oct 2005 10:15:08 -0700, Radium wrote:
> Andrew Reilly wrote:
>
>> Why would you want the read speed to be different from the "bit-rate", on
>> average?
>
> Its not a question of wanting. Isn't it true that the bit-rate and the
> read-speed don't necessarily have to be the same?

It is only true (that they don't have to be the same), in the average,
when the medium is not being used for real-time performance. That is,
it's perfectly reasonable to use a higher read speed when copying the
medium, for example.

>>The alternatives are under-run or unbounded buffering.
>
> Whats are these?

Under-run is when the real-time performance engine requires some bits in
order to produce the next part of the real-time performance, but they
aren't available because the read-speed is too low. This happens all the
time with, for example, streaming audio and video over the internet, where
the instantaneous effective transfer rate can vary wildly (due to the
presence of other traffic on the shared medium, at least).

Unbounded buffering is essentially the copying exercise. If you read bits
off the medium faster than you use them, then you have to put them
somewhere: buffer them. If your read speed is *always* faster than your
consumption speed, then your buffer will need to accommodate data at the
rate (read-speed - consumption-speed) for as long as the performance
continues.

>
>> What
>> other significance do you hold for the term "bit rate"?
>
> How many bits transferred per in a specific period of time = bit-rate
>
> There is a difference between *sample* rate and *bit* rate.

Of course. That difference is determined by the data encoding. They are
not typically completely unrelated, though. For a given encoding, they
are usually proportional. The bit rate merely tells you the maximum
information capacity of the channel.

I had deliberately not mentioned sample rates, as I thought that there
were enough points that needed clarifying in the data storage and
transport parts of your fantasy... We can get on to sample rate issues if
you like. As Jerry said, waiting a little over three billion years per
frame is not something that I'd fancy as an entertainment.

--
Andrew

[Back to original message]