h a l f b a k e r y
You gonna finish that?
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
or get an account
One of the great unsolved problems of modern
photography, which by definition is digital, is the lack
a robust means of photo archival that can survive long
spans of time and neglect -- and by that I'm not talking
about 5 to 10 years in someone's climate-controlled
a cloud provider's fully-staffed data center,
more along the
lines of 50 to 1,000 years
in a forgotten corner of a hot and humid warehouse
subject to occasional freezing.
Think ahead 1,000 years. We have pre-historic cave
paintings today that will still be around after the
Solar Electro-Magnetic Flare Of 2,XXX (insert year) will
have wiped out all of the planet's hard-drives & flash-
drives containing the majority of modern-era visual
media. Only photos on CDs & DVDs which were
a brief 20-year period (1995 to 2015) will have
the Great Solar EMF, and even fewer would still be
I propose a product to be sold to museums,
governments, well-funded artists, and others
in the long-term preservation of digital photographs in
their full glory: the
Perma-Print. Each Perma-Print is a pixel-accurate
digital image printed onto
the core layer of a 200-lb, 50-inch wide monolith of
synthetic diamond. The size and heft
are required due to the physical structure of each
pixel, as described below. Hardened glass could be
used in place of synthetic diamond, but the latter is
preferred for its improved
resistance to fire and crushing forces and better
ability to retain shape.
Each pixel in the print is represented by a "print-site"
- a set of micro-
crystalline blobs of pigment
corresponding to one Bayer Color Filter Array. Each
Bayer CFA sub-pixel's surface color provides accurate
representation to a human viewer, while its
structure offers precise machine-readable 16-bit RGB
(Red-Green-Blue) composition for future restoration
digital media. The crystalline structure provides a
quantized, countable number of possible sizes for
blob (e.g. a small red hexagonal blob is a red pixel of
value 6; a mid-tone red is a much larger red polygon
32,768 edges). The use of Bayer Arrays of single-color
sub-pixels avoids the
chemical and physical complications that would arise
from mixing pigments.
At 16 bits per color, dynamic range
(shadow and highlight details) actually exceeds that of
color pigments, and can accommodate the RAW
format of 2012-era digital cameras without loss of
detail. This means a future archaeologist will be able
to visually inspect the image to see what they got,
and then perform a full digital restoration which is not
subject to the limited dynamic range that the color
pigments are able to transmit.
Consumer-grade, cheaper versions of the product can
also be marketed which would be smaller and use
in place of synthetic diamond or hardened glass. The
smaller pixels would only support 8-bit RGB, which is
adequate for consumer use. These would be water-
and EMP-proof, but not fire-proof.
Teflon data storage
Glass has the Big Problem of being brittle. Diamond, too! [Vernon, Dec 31 2011, last modified Jan 03 2012]
another long term data storage idea...
.. includes GoatTV [not_morrison_rm, Jan 01 2012]
[theircompetitor, Jan 03 2012]
Punched tape reader
[MaxwellBuchanan, Jan 04 2012]
Punch card reader
Three seconds of google search reveals... [ye_river_xiv, Jan 05 2012]
||@Vernon - I revised the design to use synthetic diamond in
place of hardened glass. Glass also has the problem of
melting rather easily.
||Given the exponential rate of growth of storage
media, I don't think this is likely to be a problem.
I think that, over the centuries, data will be
duplicated, downloaded, backed up, replicated,
format-converted, reposted umpteen times. In
2000 years from now, my guess is that all of
today's data will exist in a billion locations, taking
up too little space to make it worth deleting.
||Media integrity aside, it's not necessary to worry
about formats either. To imagine that today's
information will be lost simply because nobody in
the year 2984 can figure out how to decode a jpeg
file is foolish.
||@Maxwell, your comment fails to address the coming
Great Solar Electro-Magnetic Flare. We are utterly
unprepared for that!
||I am pretty sure the GSEMF will have a devastating
effect, much like the Y2K bug did.
||If anyone's really worried, perhaps they should start
building sarcophagi out of heat-moulded CDs. In
millennia to come, archeologists will lovingly peel
them apart and reconstruct the lost data.
||// over the centuries, data will be duplicated,
downloaded, backed up, replicated, format-converted,
reposted umpteen times //
||This method of data storage is already baked. It's called
||Unfortunately, the data therein is irreversibly corrupted.
||So are most of the people who read it and believe it.
||The merits of diamond as a data storage substrate will
undoubtedly be as you say [ignobel] but I wouldn't use it if
there are non-valuable alternatives. It would be like
inscribing important information on gold due to its
resistance to corrosion. The substrate wouldn't fail in itself
but would soon be appropriated by others for their more
base and selfish purposes.
||Teflon, as suggested by [Vernon], or 15 copies of the glass
version scattered across the land would be much more
likely to survive the time periods you are positing.
||MB, With such a mound of bits, seems like the
problem will be how to find the relevant data.
||I don't know that it matters, [Ray]. Regardless of the
subject matter they will be "official photographs" of a
"former world leader" or "high court official", if
experience of the last few hundred years of
archaeological finds is anything to go by.
||We could just put everything into one huge ZIP file
and then email it to ourselves...
||Re: melting, brittleness issues:
||1. Store archive in cool, dry place, such as Yucca Mountain.
||2. Inscribe entrance to storage facility with Rosetta-style
message in Binary, English, Japanese, Russian, Arabic, and,
oh I don't know, Esperanto or something. Message reads as
|| "World data archive within. Do not melt. Do not strike
with blunt objects. No horseplay."
||Wonder why the ancients don't want us to play with
||I'm not sure about diamond as a storage medium - it's pretty brittle and it burns quite easily (producing CO2).
Then the problem with a message (as [Alterother] suggests) is that no one will understand it. We have pretty much no idea what the ancient Egyptians wrote, and that was only 5,000 years ago.
||I wonder if that's true, hippo. Presuming continuous
use of digital media (a big assumption, granted),
language drift might truly slow down, since recorded
media provides a self correcting mechanism -- e.g.,
no idea how Shakespearean actors pronounced
Shakespeare's plays, but no doubt whatsoever as to
how Olivier did.
||...unless use of digital media actually accelerates language drift by making archive material unavailable. E.g. If I were given a 5.25" disk from the 1980s now, I'm not sure how I'd go about finding out what it contained.
||Some of your old home videos maybe. Maybe if
Facebook was to somehow dissapear, even much of
the stuff you've uploaded (though that's a lot less
likely as cloud storage permeates). But Hamlet is
quite certain to be available.
||//given a 5.25" disk from the 1980s now, I'm not sure
how I'd go about finding out what it contained.//
||On the other hand, numerous librarians, digital
archivists and other people would be able to read in
five minutes. We can still (where need arises) play
Edison cylinders, decode Morse, and view 16mm
||and can read several different layers of the Dead Sea
Scrolls, to boot. I find the whole topic a combination
of the typical good old days nostalgia /end of the
world is coming hysteria. I think digital media is
going to be just fine.
||Really? Try reading something encoded on punched
||// that no one will understand it //
||Which is why I suggest that the first column of the message
be written in binary. Anyone who can't translate that has
no business inspecting digital media archives anyway.
||[infidel] you think punch cards are more difficult
read than Dead Sea Scrolls or the Rosetta Stone?
||The point is not that it may be difficult to read --
point is that if needed, it will be readable unless
powered devices themselves disappear -- and
that's conceivable, it's not very likely.
||As the network becomes ubiquitous, cloud storage
will in any case make this much less of an issue
than it even is today.
||//Try reading something encoded on punched
||I hope there's plenty of these glass pictures after the apocalypse. I do so like knowing there will be ample sources of sheet glass. It's the best for making arrow heads.