[Tig] Digital data storage (was something else)
Wed Mar 9 14:14:14 GMT 2005
You're right- I didn't read your post (I only get the digests). But now that
I have I still stand by what I've said. In fact, forgive me if I may have
missed your point somewhat.
I'm sure you're very excited about future trends, MEMS-based storage sounds
wonderful (I myself can't wait for the latest line of bionic eye implants
slated for 2020), but I'm talking from a purely practical, measurable,
working practice that can be used today. Sure let's re-evaluate come 2007-
let's see if it lives up to all the useful properties you've mentioned-or if
it goes the way of betamax and several "next-generation DVD" designs.
Right now in 2005, I don't think anyone responsible for image archiving will
argue with the comments I've made. My comments only apply to the film
industry, because those are the needs of the film industry. Financial
information can quite happily be stored in format-independent data, as can
text (ascii is not likely to go anywhere for quite a while) but pictures is
not numbers. So if you can't agree on the best way to make a picture into
numbers, how can you successfully store the numbers, using MEMS, optical
storage hard disc or otherwise?
Similarly, it is not practical to constantly copy/transcode film data. One
of the big issues in all of this: who will pay/be responsible for this?
Goodness, if the studios were so inclined to protect the films, they would
have properly stored the majority of films made. And there's a very good
reason why they don't: copyright. See, in the short-term, studios own the
film. But in the long-term, copyright expires, and the film becomes public
domain. Which means that the studios would end up having invested a lot of
money in a product they no longer own. I appreciate that studios are trying
to extend copyright laws to circumvent this (micky mouse law, etc.), but let
me re-iterate the point- ultimately the only people who benefit from
long-term storage are the filmmakers (in terms of posterity) and the public,
neither of whom can afford to (nor probably have the will) to re-archive
material every so often.
So, back to 35mm film: a simple, compact, fairly inexpensive media that
usually exists as a by-product of most film productions (hell, you can even
just keep one of the release prints if you're desperate), that is good for
many years, even if you treat it badly and lock it in a cupboard under the
By the way, anyone who's interested in this topic should definitely read
"The Death of Cinema" by Paolo Cherchi Usai.
Maybe you didn't read my post. MEMS-based storage is coming. Beyond-belief
ruggedness, reliability, speed, density, etc. How does this make digital
storage "almost exclusively the worst form of long-term archiving". To the
contrary, it will make it almost exclusively the best form of long-term
Don't confuse consumer junk with serious data archival. The financial
industries, for example, dwarf, by far, the data archival needs of the
motion-picture industy. You can bet that there's a serious approach to
safeguarding this data. It may very well be that today this means
microfilm. I don't know. Nobody is saying that this is a bad choice today.
However, in the future? Nah, it'll have to be digital, for all the reasons
I stated earlier. There will be media that'll be able to store terabits per
square inch, rather than gigabits. Benefit will be found in physical
volume, power requirements, access and duplication speed as well as raw
robustness and longevity.
If you care about the data you'll make sure that a system is in place to
make it last. As an example, I have AutoCAD files from 1984. Back then
they existed on 8-inch floppies. As time went by I moved my digital library
from medium to medium. Now it's on CD's and on dormant USB hard drives
(powered-down after backup operation). Multiple copies are kept in
In other words, the media is of no consequence if a proper approach to data
archival is taken. With film it is about the medium. With digital it
isn't, it's about the data. The medium is just a state.
Ultimately it is the content owner that will have to decide how seriously
data archival needs to be approached. Hopefully facilities are able to
guide their hand by highlighting available choices.
But, yes, if today you pick any contemporary data storage medium, there is
almost no doubt that it will not survive as long as properly handled and
preserved film. If you think of data storage as being tied to a particular
medium forever, then, yes, it's the worst possible choice. However, if data
will simply "lease time" on a medium until it can move to something
different, then, digital can be far superior, in all regards, to the
More information about the Tig