Re: [ARSCLIST] Preservation media WAS: Cataloguing still :-)

[Don Cox <doncox@xxxxxxxxxxxxxx>]

On 03/09/06, steven c wrote:

> Well, there is an upper limit to the accuracy of any analog
> data-storage technology...if nothing else, the molecular qualities of
> storage media define a limit (albeit a VERY large one!). In fact, I
> would suppose this would also apply to digital media?!

It has to, because digital media are physical objects just like analog
media. The zeros and ones on a hard drive are recorded in exactly the
same way as an audio signal on a cassette, or a digital record on the
audio cassettes that were used for computers around 1980.

The signal on a manufactured CD is basically the shape of a piece of
plastic, just like a vinyl LP.

What digital adds is error correction. 

In the future, when optical readouts from grooved discs are more
advanced, it will be possible to do some error correction by examining
different parts of the groove to find the least worn or damaged zone.

> 
> However, in most cases there is a much smaller natural
> limit...varying, depending on the physical quality of each medium.
> Writing and drawing were limited by the size of available pen points,
> for example...and cunieform inscriptions by the size of the "scriber"
> used to create them. Since there is also an accuracy limit with
> digital representations of analog data, it would seem there would be a
> point for each storage medium where the two forms of inaccuracy would
> coincide. For example, the accuracy of a photographic image is limited
> by the "grain," or the size of the particles which react to
> light..

Actually, as with tape, it is limited by the irregularity of the
particles, which sets a noise floor below which the signal is hard to
detect. There is no exact equivalent of the "pixel" in analog film or
tape. Therefore, there is no hard cutoff at the upper frequency or lower
signal limits.

Dithering restores some of this advantage to digital records.


> .likewise, the accuracy of a scanned image is limited by the
> number of "pixels"... which in turn is dependent on the density of
> light-detecting entities. Finally, there is another limit, since the
> data is to be interpreted by humans (at least in most cases?)...and
> this is dependent on the degree of inaccuracy our humain brains will
> notice! Obviously, this last is the absolute limit (until working
> Artificial Intelligence becomes feasible)!
> 
> It will be interesting (unless we blow ourselves into an accumulation
> of radioactive particles?) to see what computers will be able to do
> in 20 years! Twenty years AGO, practical minicomputers had just
> become reality...and the (still in use) 640KB of internal memory
> was available on the fanciest and most expensive machines, which
> ran at (IIRC) 6MHz and had 10MB hard drives...

How did engineers make the first CDs, when hard drives were not big
enough to hold 600 Megs of data?

Regards
-- 
Don Cox
doncox@xxxxxxxxxxxxxx