What is archival?
Royce Bair
<jno...@sirius.com> wrote:
>> ..It might surprise you to see Cibachromes given only a 29
>> year lifespan.
>>
>> Iris prints using archival inks and papers are known to
>> be more archival than any
>> color printing process used traditionally. I've heard Fuji's Crystal
>> Archive paper comes close, though.
Charles Eicher <cei...@inav.net> replied:
> Never heard of dye transfer or autochrome, I suppose? There are plenty
of
> archival color photographic mediums conservatively rated at lifetimes
of
> 300 to 500 years.. I even use my Epson inkjet as an intermediate step
in
> one of these processes, but no inkjet inks ever appear in the final
> product. I'd tell you all about it, but its my trade secret and I
haven't
> exhibited or published it yet. Its not that hard to do, though.
> There is NO SUCH THING as an archival inkjet print.
We've had a lot of discussion in the "rec.photo.technique.art" newsgroup
under the
topic of "New Photo Giclee Show" about whether inkjet art is now
starting to become
"archival", compared to other mediums. I'm reposting some of my
thoughts under
this new subject (What is archival?), so that others might take notice
of what I think is an important topic.
Charles Eicher makes some good arguments about archival. But I think
we're talking semantics here. What is "archival" to one person is not
archival to another. Where do you stop in your definitions? Is 300 to
500 years long enough?
My friend, Charles Berger, who developed the UltraStale Color System (a
pigment transfer system) <http://www.ultrastable.com/>, which WILL allow
one to produce color prints that will last 500 years, also produces our
UltraStable Canvas for inkjet printing
<http://www.tssphoto.com/sp/dg/canvas.html>. With moderately archival
inks like Equipoise, the canvas will hold an inkjet image for about 24
years before noticeable fading occurs (remember, this is before any
"noticeable" fading occurs, using a dark storage print for side-by-side
comparison -- not a total fade to blank canvas). The print could last
much longer with a more archival ink like Lysonic.
So far, the UltraStable canvas is the most archival and stable canvas
that has been tested by Henry Wilheim. As I've said, the Lysonic Ink is
much more archival than Equipoise ink, but most artists and printers
still perfer the Equipoise because it has more color saturation, and a
wider color gamut. Most artists would prefer that a larger audience
enjoyed a BEAUTIFUL version of their work right now (even for only 24
years), than a less than beautiful version that would last much longer.
The argument being, "what good is archival, if it's not reflecting my
best work?"
I'm not worried about my inkjet art fading away at some point -- I have
archived digital files of all these photographs onto Kodak gold CD-Rs
<http://www.tssphoto.com/sp/dg/kodak_cd.html>. Independent tests show
that these should last over 300 years under normal conditions, and may
last well over 1,000 years in dark storage. As inkjet longevity
improves I or someone else can reprint these images.
I want my photographic art to be seen right NOW in its best form and on
the brightest, the most colorful medium, and with the most flexability
of control -- and this for me is the inkjet or giclee process. A giclee
allows more people to see my art (through editions) and a good giclee is
much more beautiful and stable (archival) than the standard 4-color
reproduction process that artists have been using to sell expensive
limited editions of their art in the past. (Charles Eicher has been
discussing how artists have been ripping people off with their
high-priced inkjet prints -- what about the artists who have been
selling these limited edition 4-color process prints for $200 to $500
each -- these will fade much faster than a good giclee because virually
all of them are printed using standard 4-color process inks!)
Royce Bair, Director <ro...@tssphoto.com>
The Stock Solution
(Archival, fine art & photo-realistic solutions for inkjet printing)
http://www.tssphoto.com/sp/dg/
Aperture Photo Gallery <http://www.tssphoto.com/art.html>
Marilyn
see his post,
There was a photography/print exhibition in Vancouver, where
the artist supplied computer disks with each work framed work on the wall.
A digital copy of the work was loaded onto disk. This seems to be one solution,
if as you say the disks last for 500 years. As soon as the photograph
or the print begins to fade the owner can reprint it. (Some kind of
copywrite agreement would have to exist).
A mute point, with our civilization poised on the brink of global
climate change.
Jurgen Eidt
>archived digital files of all these photographs onto Kodak gold CD-Rs
><http://www.tssphoto.com/sp/dg/kodak_cd.html>. Independent tests show
>that these should last over 300 years under normal conditions, and may
>last well over 1,000 years in dark storage. As inkjet longevity
>improves I or someone else can reprint these images.
But you hardly find a CDROM in future centuries.
Try to read a 8 inch disk today...
Don Stauffer
are well aware that they blew it with some of the past digital storage
media. My suspicion is that they are well enough aware of it that there
will be either the keeping of the appropriate machines at the archives, or
else they will recopy collections onto newer media before they trash old
drives.
Jurgen Eidt wrote:
--
Don Stauffer in Minneapolis
home web site- http://home1.gte.net/stauffer/
home email- stau...@gte.net
work email- stau...@htc.honeywell.com
Joshua_Putnam
>>I'm not worried about my inkjet art fading away at some point -- I have
>>archived digital files of all these photographs onto Kodak gold CD-Rs
>><http://www.tssphoto.com/sp/dg/kodak_cd.html>. Independent tests show
>>that these should last over 300 years under normal conditions, and may
>>last well over 1,000 years in dark storage. As inkjet longevity
>>improves I or someone else can reprint these images.
>But you hardly find a CDROM in future centuries.
>Try to read a 8 inch disk today...
It's easy enough to get data transferred off of 8-inch disks in
any large city, since there are businesses that specialize in
obsolete data recovery. But CD-ROM has an added advantage over
most other storage devices that are or will soon be obsolete for
computer use -- consumer audio. Just as you can still buy brand
new turntables that can play 45 and 78 rpm records, it will be a
long time before drives to read standard CDs become unavailable.
(Note this applies only to audio formats that are truly
well-accepted world standards -- it's already difficult to get
good players for 8-track tapes from the '70s.)
--
Jo...@WolfeNet.com is Joshua Putnam / P.O. Box 13220 / Burton, WA 98013
"My other bike is a car."
http://www.wolfenet.com/~josh/
Edward Bigelow
does...or the Metropolitan Museum...or the NYT??
Ed
Don Stauffer wrote:
> Archival technology is worthless unless you have an archive. The archivists
> are well aware that they blew it with some of the past digital storage
> media. My suspicion is that they are well enough aware of it that there
> will be either the keeping of the appropriate machines at the archives, or
> else they will recopy collections onto newer media before they trash old
> drives.
>
> Jurgen Eidt wrote:
>
> > >I'm not worried about my inkjet art fading away at some point -- I have
> > >archived digital files of all these photographs onto Kodak gold CD-Rs
> > ><http://www.tssphoto.com/sp/dg/kodak_cd.html>. Independent tests show
> > >that these should last over 300 years under normal conditions, and may
> > >last well over 1,000 years in dark storage. As inkjet longevity
> > >improves I or someone else can reprint these images.
> >
> > But you hardly find a CDROM in future centuries.
> > Try to read a 8 inch disk today...
>
Keith Clark
>
> >
> > Jurgen Eidt wrote:
> >
> > > >I'm not worried about my inkjet art fading away at some point -- I have
> > > >archived digital files of all these photographs onto Kodak gold CD-Rs
> > > ><http://www.tssphoto.com/sp/dg/kodak_cd.html>. Independent tests show
> > > >that these should last over 300 years under normal conditions, and may
> > > >last well over 1,000 years in dark storage. As inkjet longevity
> > > >improves I or someone else can reprint these images.
> > >
> > >
This will be unpopular with some folks, but here goes nothing...
"Archiving" digital images on a CD is false security. OK, so the CD will last for
a hundred years or more without damage. Granted. I accept that claim (for the sake
of argument).
But how will you read the data in 100 years? CD players will long have been
obsolete, and if you use some non standard format, or use jpeg to "save space"
then how will the files be read?
I don't work for a photo lab, so I can say this with no commercial overtones...
Your best archival method for your digital images is to take your file to a
digital photo house and have them output your file to a NEGATIVE, and store the
FILM in an archival manner.
Given the extreme pace of change in the computer industry, which will be
-accessible- in 150 years : proprietary encoded data on a decades-obsolete gold
disk, or a negative that any non-techie can look at with a loupe?
Hey, I put my scans on CD too (I scanned about 4 gigabytes last week for a new
project). Except I don't expect them to be "archival".
Just my two cents worth.
Jason O'Rourke
>But you hardly find a CDROM in future centuries.
>Try to read a 8 inch disk today...
i saw this the last time archiving was discussed. It's still a lousy
example. The 8" disappeared before PCs became common and in a time when
mass storage was 360k, not 650M (a 2000X change).
You should note that the upcoming replacement to the CD, the DVD disk,
comes in the same form factor and its drives read cds perfectly well.
Perhaps in 2010 something will take over in its place, but if that is the
case you can xfer all of your cdroms to a single one of whatever that is.
Properly stored film may well be a longer lived format, but most people
aren't going to 'properly store film.' CDs are a little more
accomodating.
--
Jason O'Rourke j...@best.com www.jor.com
'96 BMW r850R
last dive: February 21st, McAbee Beach (Monterey) - 36ft max for 40 mins.
Blue Moon
>But CD-ROM has an added advantage over
>most other storage devices that are or will soon be obsolete for
>computer use.
Today's CD's are already becoming obsolete as DVD takes over
with multiple layered and two-sided discs of greater capacity,
and soon there will be something replacing DVD. The thing about
CD's that is often overlooked is that they are PLASTIC. There is
no known plastic that will survive for longer than a century if
it lasts that long. Plastic items in museums are a major source of
headaches for conservators since they eventually disintegrate.
Michael O.
Joshua_Putnam wrote in message <7av7p9$s1p$1...@sparky.wolfe.net>...
>Just as you can still buy brand
>new turntables that can play 45 and 78 rpm records, it will be a
>long time before drives to read standard CDs become unavailable.
>(Note this applies only to audio formats that are truly
>well-accepted world standards -- it's already difficult to get
>good players for 8-track tapes from the '70s.)
>
> Jo...@WolfeNet.com is Joshua Putnam / P.O. Box 13220 / Burton, WA 98013
> "My other bike is a car."
> http://www.wolfenet.com/~josh/
By the way, I am a HUGE fan of 8-track and I'll be damned if I am going to
stand here and let you bash that medium as not a "truly well-accepted world
standard audio format."
Working in the commercial photography field, there are numerous reasons to
go digital. Most times, archival is not an issue. (Whose been selling the
same products for 50 yrs?). But I have a real problem with people calling
'digital photography' art. Yes, I'm an old, stale, son-of-bitch who loves
his hassy, enlarger, and the smell of good, fresh stop bath. Digital
anything to me is a cheap, fast, no frills subsititute to the tried and
tested old habits that work. This happens to be great if you are an
accountant (I've done my share of 13 coloumn work) and word processing (I
have used my share of white-out). But I have a hard time watching an Iris
or Epson printer zip back and forth printing MY work FOR me and calling it
'art.' At work we have Shop 5.0 in every bay, with megavision backs on our
hassy's and leaf backs on our Sinars. I have no problem watching a print
come out of our Iris of a piece of jewerly we just shot. In fact, I think
that is just bitchen. In keeps us moving so we dont have to wait for the
chromes to come back from the lab. It is good for *production.* It is not
art.
Now on to cost. These Iris printers cost about 30-50,000 cajones depending
how they're set-up. I am not sure how a frugal, meager photographer/artist
(like me and most like me) can afford a $50,000 printer - thats mucho
cajones, at least for me. (I'm sure there are spoiled brat artists out
there that mommy and daddy buy them anything though) And anyway, if I had
50,000 cajones, I could build one damn nice darkroom - even could have my
Durst L1200 multigraph I want. I have also seen the fairly new Epson 5000.
I read articles that claimed this was supposed to be the "Iris Killer"
costing the meager *artist* only 10,000 cajones. Geez what a savings. Now
I can buy a car....yeeeeahh. And, yes, they reproduce a digital image
*perfectly* (as interpreted by the photographer/artist). I have also seen
what can be done even with a consumer based Epson Stylus and the prints come
out just like a photo - not unlike a good roll of 35mm Tri-X left in the sun
for two years, loaded in the camera and pushed to 3200 and then blown up to
20x24.
Seriously, my two cents.....If you are to produce a photographic image,
IMHO, and call it 'art', it should be processed as they have the past 100
years. Instead of photographers arguing with the computer dweebs about what
is 'photographic art', I would like to suggest we just take action and put
a circle with a line thru it and the word 'Digital' in the middle and stamp
it on the back of our archival prints (with acid-free archival ink, of
course). This would mean the photo has not been digitally
made/altered/retouched in any way, shape or form. Let people who buy our
prints interpret what they consider 'photographic art' when they go to the
auction. Which one would you buy? Here is a weird, hypothetical
situation- If Ansel would have made some of his prints on an Iris (maybe he
did, I'm not sure), I know, being a photographer and printer, that if I were
at an auction and had a choice of either medium of Ansel Adams' work, I
would want the fiberbase print that came out of his fix bath, not the
'print' that came out of an Iris Printer - no matter how damn three
dimensional it looks.
Mike
Royce Bair
I've heard this argument before, and I think you make several very valid points. As
one person also said:
> But you hardly find a CDROM in future centuries.
> Try to read a 8 inch disk today...
Also, a CD-ROM or a CD-R only have to have an error in ONE bit in the bit stream and
the whole CD is toast, while one small scratch in an analog negative does not wipe
out the rest of the image!
Still, I have hope that beween my negative and 2 CD-Rs (the master and a backup) that
one will survive, and a digital reader will be available in the future.
Garry W
>Your best archival method for your digital images is to take your file to a
>digital photo house and have them output your file to a NEGATIVE, and store the
>FILM in an archival manner.
>or a negative that any non-techie can look at with a loupe?
In a hundred years film -- and the film scanners that converted the film to
digital during the "digital transition" -- will be a thing of the distant
past. This will cause problems for your descendants who find your
safe-deposit box full of photo film. They will indeed still be able to look
at that negative in a hundred years, assuming it hasn't gotten trashed. But
getting an analog print from it will be very difficult: 1990s-era photo print
shops will no longer exist. They will need a very rare, very specialized
service shop to print directly from that quaint medium of "film".
I think the only way one can have first-rate archiving is to
1) store the pictures digitally, so that new "master" copies may be made
without loss of quality, and then
2) re-master your entire archives to the current format-du-jour as necessary,
for example every 10 years.
I use the most basic computer physical medium (which at present is hard
drives) -- as my archival storage medium. I do not mess with CD or tape at
all (except for in-case-of-disaster backups). If one always uses the most
basic format, making the new "master" physical copy will always be trivial,
at each step of the way as the hundred years passes by. Because there too
many millions of people out there now with too many billions of bytes of data
for the market to allow the copying to be hard.
(And hopefully as the computer industry matures the *software* will start
also providing trivial upgrade routes, driven by the force of the same
millions of people, with their billions of Web page images.)
Right now the fine-art film medium and market does have billions of drugstore
prints, newspaper snapshots, etc, backing it up. But in 15 years those
drugstores and journalists will, of course, be completely gone. They'll be
digital. At that point will you still want to be archiving onto film?
>Hey, I put my scans on CD too (I scanned about 4 gigabytes last week for a new
>project).
And I'm starting to get good use out of that Photosmart film scanner you sold
me last week. It's a whole lot better than scanning from prints. I had it
plugged into my two-pound laptop for a while - it was about 8 times bigger
than the computer that was driving it. Cute.
>Except I don't expect them to be "archival".
It's the CDs that may the problem. Use hard drives, if you can afford to.
Do you like your new Nikon scanner? I'm *already* thinking about upgrading -
the Photosmart isn't perfectly perfect. Of course.:)
Garry
Don Stauffer
Keith Clark wrote:
>
>
> This will be unpopular with some folks, but here goes nothing...
>
> "Archiving" digital images on a CD is false security. OK, so the CD will last for
> a hundred years or more without damage. Granted. I accept that claim (for the sake
> of argument).
>
> But how will you read the data in 100 years? CD players will long have been
> obsolete, and if you use some non standard format, or use jpeg to "save space"
> then how will the files be read?
>
> snip
You are one of many folks making this claim. Of course, it is very hard to prove or
disprove prognostications about the future. But what is the old saying about, "once
burned, twice wary?" My belief is that archivers have learned a lesson by not keeping
8 track (computer tape not music tape) readers and whatnot. These people are not
dummies. They are professionals, and they know they made a mistake. To think that
they will continue to make the same mistake over and over is somewhat unwarranted, in
my judgement.
BTW, there are now firms who offer to convert old file formats like WordStar to Word
or whatever, for a price.
SPECTRUM
<stau...@htc.honeywell.com> wrote:
>These people are not
>dummies.
Oh I wouldn't be so sure. I recently had a friend
bring me several pieces to a computer and ask me to install
them. None worked and all were less than 10 years old.
CDROMS, video card, sound card. Nada. I was quite surprised
at this. 'sOK though. I sold him my oldies for what a new
36X cost me. After all he doesn't need a 36X in a 486 !
Regards,
John S. Douglas http://www.spectrumphoto.com
Darkroom formulas & facts. F.A.Q.'s Photo & computer links.
Prof. portrait, bridal, stock & commercial photography.
WORLD FIELD PHOTOGRAPHERS ASSOCIATION
Robert J. Niland
> Properly stored film may well be a longer lived format, but most people
> aren't going to 'properly store film.' CDs are a little more
> accomodating.
Don't count on it. Some of my 1994 Kodak PhotoCD CDRs are
already unreadable.
Regards, 1001-A East Harmony Road
Bob Niland Suite 503
mailto:r...@frii.com Fort Collins
http://www.frii.com/~rjn/ Colorado 80525-3354 USA
Unless otherwise specifically stated expressing personal
opinions and NOT speaking for any employer, client or Internet
Service Provider.
Mike Andrews
: Jason O'Rourke (j...@best.com) wrote:
: > Properly stored film may well be a longer lived format, but most people
: > aren't going to 'properly store film.' CDs are a little more
: > accomodating.
: Don't count on it. Some of my 1994 Kodak PhotoCD CDRs are
: already unreadable.
Any idea what the cause(s) of the degradation might be?
E.g., surface abrasion, scratches, deformation of the pits
or bumps with age, etc.
Are the PhotoCDs unreadable on several different readers?
This may have some bearing on our archiving choices here
at ODOT. Anyone else seen anything of this sort?
--
Mike Andrews | speaking for himself
MAnd...@odot.org (when it works) | posting work-related
else mand...@notes9a.okladot.state.ok.us | during work hours
Currently doing time as datacenter director | IAW ODOT policy
Joe McCary
<eb...@ix.netcom.com> wrote:
|So where are the archives? Does anyone know what the Library of Congress
|does...or the Metropolitan Museum...or the NYT??
While I have no idea what the Library of Congress does in the matter,
I do have "some" knowledge of what the US National Archives does, my
business is a vendor for photos at the National Archives. As they
have somewhere between 9 million and 18 million images on file, in
cold storage, at their College Park Offices (Archives II) NARA
(National Archives & Records Administration) is moving toward an all
digital storage. But it sees some problems. The volume of images on
file is so large that it will take years to make digital copies and
file these. Next comes the storage media problem. Everyone seems to
think that CDs are "archival." And while they seem to be fine we just
do not know. How can we intrust the records of our nation to a media
we are not positive will last 100+ years? I have copied photos of
Mahew Bady and others that are well over 100 years and they are fine.
I have also copied images from the Nixon era that are in such poor
condition that they are unusable. Modern "progress" isn't always
progress. The question is how long will the media of CDs last. Lets
look at another storage media that has evolved over the years, the
music recording business started out on wire, then quickly moved to
some very fragile (I have no idea what it was made of) records. They
moved to Vinyl. How many people have been using vinyl records since
the CDs were introduced in 1981? "Record" stores don't exist anymore
they are all now music stores. What is next, we have already seen the
advent of DVDs. Is this media to replace CD ROMs? What is next? and
better yet, how soon? How many PC users out there have a 5ź floppy
disk on the machine they are using now? It was once the standard not
so long ago. So if the Archives it to begin placing their records on
CD ROM media they want to be sure it will last at least some time. I
am sure they do not want to be recopying the material several times
every hundred years while the computer industry jumps on a new
bandwagon of storage media.
That is the first question. The next one is how large should the
files be? There are lots of 4x5 negatives in the Archives, should
these all be scanned at what dpi? 300 dpi? maybe they should all be
scanned as 8x10s @ 300 dpi? If that were to happen, how would someone
use that image in a big print? So many questions, so few answers that
anyone agrees on. I think that scanning all the images is a great
idea, so many silver images are fading that they should start with
these first. If they were to identify the perhaps źth of the
collection that is in "danger" and make digital copies of these it
would be a start (they already do silver copies for just such
reasons).
Joe McCary
Photo Response
Joe McCary
Robert J. Niland
> : already unreadable.
> Any idea what the cause(s) of the degradation might be?
> E.g., surface abrasion, scratches, deformation of the pits
> or bumps with age, etc.
No idea. These CDRs have always been stored in their jewel boxes,
out of direct sun, at room temp, and handled carefully (not even
a fingerprint on the data side of the media). If no one else
is having similar problems, I can only suspect the writer at
the local PCD processing lab.
> Are the PhotoCDs unreadable on several different readers?
Yup. IDE, SCSI, CDR, CD-ROM, makes no difference. I've tried
about 8 different drives.
> This may have some bearing on our archiving choices here
> at ODOT. Anyone else seen anything of this sort?
Anyone archiving on unproved media (which is any computer
media newer than punched cards :-), needs to have a policy
of read-testing their media every year or so, using a
reader that can report corrected errors, so that you
can plan to migrate the data when corrections grow, but
before the errors become uncorrectable.
Andrew J Robbie
>
> > But you hardly find a CDROM in future centuries.
> > Try to read a 8 inch disk today...
>
> Also, a CD-ROM or a CD-R only have to have an error in ONE bit in the
bit stream and
> the whole CD is toast, while one small scratch in an analog negative
does not wipe
> out the rest of the image!
This is simply not true! CDs have very sophisticated error detection and
correction schemes built-in -- you can drill holes in music CDs and they
will still play without any loss of quality. Also, a well designed format
allows for gradual degradation. Thus very severe damage will increase the
signal to noise level but hopefully the image could be useable.
I think we have to seperate archival into (1) maintained archival, and
(2) unmaintained archival. For (1) digital storage on modern hardware is
good. Film, if kept cool & dry, may last longer if unsupervised (but kept
in vaults, etc). However, for (2) you really need to go for "proven
technology", eg pyramids :-) Or do what NASA did for the Voyager probe:
put engraved instructions on gold sheets. On that basis, etchings would
seem to have the most potential.
Personally, I suspect CD-Rs will have poor archival performance: the dye
layer will fade, the CD will delaminate etc; technology will dramatically
pass it by. Just the other day I pulled out some Kodachromes shot by my
grandfather ~45 years ago and they were exquisite. I also found some old
medium format B&W negs which must have been 50-60 years old and they were
in great shape. OTOH it would be very expensive to recover data from any
1960s mag tape.
Regards,
Andrew
--
a.robbie (at) student.unimelb.edu.au
Garry W
>In article <36D387B8...@tssphoto.com>, ro...@tssphoto.com wrote:
>> Also, a CD-ROM or a CD-R only have to have an error in ONE bit in the
>bit stream and
>> the whole CD is toast, while one small scratch in an analog negative
>does not wipe
>> out the rest of the image!
>
>This is simply not true! CDs have very sophisticated error detection and
>correction schemes built-in -- you can drill holes in music CDs and they
If you do that to an audio CD, you'll get a burp in the music but it'll keep
playing.
If you do that to a data CD, you'll get a burp in the data, and the CD will
probably thereafter be dead for its intended purpose.
I have a number of commercial software CDs here at home which have received
minor scratches or abrasions and are now unuseable for their intended
purpose. I.e., the software installs start, then crash. I now treat software
CDs as fragile and delicate objects. It's been a rather expensive lesson to
me.
Garry
PS - The CD "scratch remover" kits sold at the computer stores do no
detectable good, in my experience.
Don Stauffer
a data file. Yes, a single bad bit, if not caught by error checking in a file
transfer, CAN ruin a whole executable file.
But, the same is not necessarily true in a data file. An error in the header
MAY be catastrophic, but an error in the true data part only affects a small
portion of the data. For example, a single bit error in a word processing file
may screw up one character. Similarly, a bit error in an image file will only
screw up one pixel.
Garry W wrote:
--
Peter H.M. Brooks
stau...@htc.honeywell.com "Don Stauffer" writes:
>There is a BIG difference in the effects of a bit error in executable code vs in> a data file. Yes, a single bad bit, if not caught by error checking in a file
>transfer, CAN ruin a whole executable file.
>
>But, the same is not necessarily true in a data file. An error in the header
>MAY be catastrophic, but an error in the true data part only affects a small
>portion of the data. For example, a single bit error in a word processing file
>may screw up one character. Similarly, a bit error in an image file will only
>screw up one pixel.
>
This is true, as far as it goes. However, most data on CD-ROMS is kept
in files, many of these files are encoded with a checksum so will fail
to work if a single bit is wrong - this is true of executable or
binary files.
--
Peter H.M. Brooks
Garry W
>But, the same is not necessarily true in a data file. An error in the header
>MAY be catastrophic, but an error in the true data part only affects a small
>portion of the data. For example, a single bit error in a word processing file
>may screw up one character. Similarly, a bit error in an image file will only
>screw up one pixel.
Most digital images are stored in compressed/encoded formats nowadays. In
these formats I think a single bit error will tend to wreak complete havoc.
You *really* want to be gentle with your data CDs.
But accidents will happen. The last time I lost one, the drive door closed
with the disk off-kilter. After rescuing the disk (btw, don't force the door,
instead power off and gently disassemble your drive), it was no longer
readable. "Microsoft...? Can you send me a new Office 97...?" Yeah, right.
Garry
Material Fellow
> I have a number of commercial software CDs here at home which have received
> minor scratches or abrasions and are now unuseable for their intended
> purpose. I.e., the software installs start, then crash. I now treat software
> CDs as fragile and delicate objects. It's been a rather expensive lesson to
> me.
>
> Garry
>
> PS - The CD "scratch remover" kits sold at the computer stores do no
> detectable good, in my experience.
In contrast I have had good luck. However you must work at it until the
scratch is almost optically invisible.
Sometimes, nose oil can be applied to level out the scratch enough to
allow the mechanism to optically read through it.
A pressed CD is much easier to restore than in a CD-R, because the
reflective signal is stronger in the first place, I guess.
jim buch
Shez
> An error in the header
>MAY be catastrophic, but an error in the true data part only affects a small
>portion of the data. For example, a single bit error in a word processing file
>may screw up one character. Similarly, a bit error in an image file will only
>screw up one pixel.
Not in a compressed TIFF or Jpeg - in my experience a single bit error
will screw up the rest of the picture from that point, and many programs
won't load such an image at all.
-Shez.
--
____________________________________________________________
Boling's postulate:If you're feeling good, don't worry. You'll get over it.
____________________________________________________________
Address any email replies to Shez (email to "news" is rejected)
Take a break at the Last Stop Cafe at <URL: http://www.xerez.demon.co.uk/>
Don Stauffer
between the two. First is personal archives. You have some pictures you want to
save so your grandchildren can see them fifty years from now. Second is a
professional, institutional archive.
Now if you put your own pictures on a CD ROM and it gets some errors, then yes it
may be completely unusable to your grandchildren, while a silver print, though it
may fade, would still be useful.
If we have material in an institutional archive, I would suspect that it would not
take much software development to come up with reading and transcribing schemes that
can ignore checksum errors. Compressed images are something else. I think we can
discount jpeg. Who would want to archive jpeg files? With something like tiff, it
may well depend on where (header or image) the error is. If I were in charge of an
archive, I would want to look into something like that.
Now, back to the consumer. If this turns out to be a problem for the consumer, my
belief is that that is an opportunity for some entrepreneur to step in and offer
recovery services. There were and ARE firms that do conversions of business records
in old formats. There has not been the demand for this kind of thing for the
consumer. But if in the future there is, I think there will be folks who will find
a way to fix it.
Jan Steinman -- jan AT bytesmiths DOT com
People are seriously confusing "single bit error" in files with bit errors
on media. CD-ROMs have sophisticated multi-bit detection and correction
capabilities built into each and every drive -- you CAN do considerable
damage to a disk before such damage will even show up at the drive's SCSI
(or IDE) port.
The correction mechanism is such that CDs are much more forgiving of
radial scratches than circular scratches. For this reason, if you do any
surface treatment, you should always apply it radially, NOT circularly --
this may explain the different success reported with scratch-removing
treatments.
While arguing against the fragility of CD-ROMs, I strongly agree that
technology is a bigger enemy of un-maintained archiving. It may be true
that today's DVD drives can deal with CD-ROMs, but once DVDs ramp up and
CDs begin to ramp down, expect that to change. I once had a floppy disk
drive that was capable of reading and writing both 80kB and 340kB formats
-- big deal! :-)
Scientific American had a very good article about the "Coming Dark Age" a
couple years ago -- worth reading by those who are unconcerned about
technology obsolescence.
As others have pointed out, "maintained" archive is MUCH different than
"un-maintained." As long as you're only migrating one technology step,
it's no big deal, but people who are putting their family photo albums on
CD-ROM and throwing away the negatives, and planning on their
great-grandchildren being able to read them far in the future are just
kidding themselves. (Their great-grandchildren may tear themselves away
from the holodeck games long enough to play with the "frisbees" for a few
minutes before boredom sets in... :-)
--
: Jan Steinman -- Jan AT Bytesmiths DOT com
: Spammers: please add:
: <mailto:commerc...@murkowski.senate.gov>
: to your list -- he threatened me for forwarding him your
: spam with a protest against his pro-spam bill, so I'm sure
: he'd love to hear from you all! :-)
William Bell
What may work for a wealthy company's modest needs for storage of a
limited number of documents or images may not be a good choice for a
large public archive or library.
I've been banging around digital imaging since the beginning of the PC
age and trad photography for much longer. If someone asked me to make
archival copies of a limited number of things, say a number of photos of
historical significance, especially if they were in color, this is what
I'd do:
I'd use either a stat camera, a high quality scanner and film recorder,
or a good old 4x5 camera and make separation negatives with clear
registration marks. Archivally process and store in clearly marked
archival packages. Even a simple one or two paragraph explanation of
what was done. Oh, I might save the scans on whatever media was the
flavor of the day, but that would be more for convenience than anything.
Media types for electronic storage may or may not hold up. We know that
much of the recorded mag audio tape from the era of tape has not held up
so well. A good condition lp will sound much better than the master
tapes they were pressed from in many cases. We do know and have lost of
experience with b+w negatives and know that they will hold up very well
if even reasonable care is taken in their storage. We have also had
100+ years to figure out how process it properly.
No matter what technology comes and goes, the laws of the physics of
light will be a constant. Even if, as has been suggested, there will be
no optical enlargers in 50, 100, 150 years, any reasonable person
involved with imaging should be able to assemble the image from the
separation negatives. Will it be the most easy way to reproduce the
image in say 10 or so years, if needed? Of course not, but by storing
the information in its most simple format, finding an expert on error
correcting image pack or TIFF files when they haven't been used in 50
years will be avoided. There will be many ways to reassemble the image
instead of just one tied to a single technology.
Sometimes I think you can be blinded to the goal and forget the advice
most of us have been given at one time or another - K.I.S.S - keep it
simple, stupid.
Bill