Question about Preserving Historic Digital Photographs

<p>I would like to know if there is anyway I can save and pre-serve historic digital photos I have taken for a religious organization. The purpose of saving is two fold: (a) I will be handling this responsibility to another generation who I think will continue in my foot steps and (b) Historically saved photos be available for publication and future references.</p>

<p>Thanks</p>

 

<p>Only by constantly backing up the files onto the latest storage media. In view of the large number of storage systems in widespread use in the past which are now obsolete and for which readers in good working order are now hard or impossible to find, anything else is just too risky!</p>

<p>At the moment, the best electronic storage system is a dedicated external hard drive. About 20 years before fading can occur, but assumes you use it only for archiving and not normal daily use.</p>

<p>Another is to print the photos on archival paper and store them properly for longevity.</p>

<p>The second best option is to print them as photobooks. Maybe two copies. One set for yourself for your own records, and a second set that goes to the organization. Choose a good book making company like Bay Photo, not the cheaper companies. and make sure all of the books are the same size which looks nicer on the shelf.</p>

<p>CHEERS...Mathew</p>

<p>M-Disk claim a lifespan of 1000 years for their new digital DVD and BluRay media - http://www.mdisc.com/ - whether or not equipment to read 1000 year old DVD will be available in 3015 is less certain. It will fairly certainly be around for the next few decades though.<br>

<br /> Conventional DVDs have a lifespan of maybe 25 years when stored properly, possibly longer. However the lifespan could be a short as a few years if if not stored under proper conditions or if there are manufacturing defects.</p>

<p>BTW this isn't a Canon EOS question, so the thread will be zapped to to the Digital Darkroom forum.</p>

<p>This is huge subject for which much has been written, including a lot by the US Library of Congress. Far too much for a Photo.net post. In addition to media concerns, you have to think about preservation of the media in the event of all manner of destruction, and also ensuring that appropriate metadata is stored in the files. Also, you need to define what you mean by preservation. 1000 years and able to withstand nuclear war is different from 25 years and able to withstand a fire.</p>

<p><em>1000 years and able to withstand nuclear war</em><br>

To be honest, Marc, I would guess the OP would be happy to store his files in a way which ensures they can be printed in a few decades' time. If nuclear war occurs, I imagine he will have other concerns. I would say storing the files on three identical external hard disks, transferring the contents onto new HDDs every 10 years and storing each of the 3 disks in a different place would provide adequate security against calamities including fire and meet the OP's needs without the expense of underground lead-shielded bunkers :-)</p>

<p>M-Disc sounds rather exciting! I hadn't heard about it and will definitely be using it.</p>

<p>However....</p>

<p>I would highly recommend something rather old-fashioned. In addition to state-of-the-art digital archiving methods, I would also create state-of-the-art archival digital prints (pigment inkjet prints on archival paper). I'd frame them well, with archival materials, under glass, and display them in areas of the religious institution artificially lit with either incandescent/halogen or LED lighting. Avoid sunlight and fluorescent lighting. Or you can store the prints in good conditions -- low humidity, moderate/low temperature, and in the dark. There are no guarantees with any storage form, but a good print may outlast our digital technologies, provided it is well treated.</p>

<p>The best survival concept for anything seems, multiple copies to start with, which get reproduced on their own.<br>

Instead of buying the wonder BluRays, I'd try to pass the responsibility for the archive to at least 2 dozen semi sane PC users. (I know enough who don't really back up, but assume if you have enough of them interlinked they can get a fresh copy from a co-owner and safe it on their replacement HDD)<br>

How muich data are we talking about in total? - I'd guess every average member of the organisation might be able to provide shelter for some 10 GB?</p>

<p>You missed my point David, which was "you need to define what you mean by preservation." I enumerated two extremes. Your comment that one of my extremes is extreme doesn't make any sense. (To me, anyway.)</p>

Making prints will give you prints you can look at and which will have a 'shelf life' of their own, transforming the data is not a way to conserve the original digital information.<br>I'm with Jochen (and others): make multiple, independent copies on convenient digital media (no BlueRays or DVDs or tapes. Hard drives will do fine, are cheap and fast). Make new, fresh copies now and again. Keep an eye on the life span of the fileformat used and change that should it look like it is on its way out. And indeed share when possible: it shares the burden of maintaining maultiple copies and distributes the copies over several separate places.

<p>Any, and I do mean ANY, archival system, digital or film, requires a certain minimal maintenance.</p>

<p>As said, HDs are the best right now, but both the data <strong><em>and the formats</em></strong> of the images need to be regularly re-archived and updated. Imagine your images were all in MacPaint format, just for an example - I'm not sure anything can read those files nowadays. I would not bet that PCT, TIFF, or JPG will be easily read by whatever kinds of computers/cyborgs exist 100, much less 200, years from now. Also, try to find a 12 inch floppy drive today.<br /> <br /> In 200 years, all the world's images will be on a simple cube 1cm square; and then somebody will misplace it.</p>

<p>Platinum or more often these days, paladium, prints -especially of color separations for color on really archival paper- are more permanent than silver, even toned, but they are expensive and, as QG points out, prints require re-scanning into some future digital format at the end when they are to be used.</p>

<p><em>So far</em> I've never had DVDs 'age out' or fail on me after writing; but although I keep copies on optical media as an additional backup, I depend nearly completely on HDs now.</p>

<p>As for claims of long-term archival status --<br>

Well, I remember from my local histories that the traveling medicine man always moved on to another town before anyone had a chance to see if his secret Indian chief potion actually worked or not.</p>

Not just something remembered in histories, JDM. Most project and change managers of our days are very much the same.<br><br>Claims are no guarantees. An expected life span of 10 years means that the mean life span is 10 years. So do check and recopy long before you think, based on figures from wherever, it could be about time.<br>Another advantage of magnetic media over optical discs: every time you want to refresh your DVD or BlueRay copies it will cost a considerable amount of money in fresh discs, making you a bit hesitant to do so.

<p>If you're treating this as an archive, and if you're going to commit these to multiple digital media storage types, and either you - or someone else - is going to move them from device to device over the years - investigate using a utility to generate a checksum of each file, and store a list of those separately and with the file.<br>

It allows you to validate that every copy you made is identical to the original, and it would allow someone in the future to insure that a particular backup copy hasn't become corrupt, and that any copy they make matches the original. I've seen defective computer memory - on a computer that seemed to be working fine until subjected to an intensive memory test - cause damage to large digital files when copied.<br>

A number or archives, libraries, universities and professional organizations have been having ongoing discussions about this for a few years, and if you do a couple of google searches you may find some best practices and suggestions of utilities to use that could work for you. It may be overkill, but it's something to be aware of.</p>

<p>Here's a 10,000 year old digital* pigment inkjet print. It looks pretty good to me:</p>

<p><a href="https://en.wikipedia.org/wiki/Cave_painting#/media/File:SantaCruz-CuevaManos-P2210651b.jpg">https://en.wikipedia.org/wiki/Cave_painting#/media/File:SantaCruz-CuevaManos-P2210651b.jpg</a></p>

<p>And I examined a 17th Century book today, printed on linen paper. It was difficult for me to read, but only because my French isn't so good. Otherwise the book looked perfectly fine.</p>

<p>I do question whether the plastic in an M-disc will remain clear and distortion-free for a millennium, and I also wonder what will be required to read a DVD 1000 years from now -- or even 100 years. I would consider a claim of 100 years longevity more credible.</p>

<p>I've been working with digital data since the 1970's and still have almost all of my earliest files intact. However, I also recognize there has been data loss throughout the years, even when I've been the one tasked with transferring it from medium to medium, format to format, technology to technology. I have no trust that future generations will preserve my digital data properly. In fact it's been my experience that data suffer a half-life of between 5 and 10 years in the hands of others. So if I want something to last for future generations (e.g. a family photo), an archival pigment print is just about the only thing I trust to make the journey. No other medium can withstand nearly as much neglect and still be usable.</p>

<p>*resembling or relating to a digit, especially a finger.</p>

A print of the digital image, however, is a transformation which is neither 'one-to-one' nor reversible. It, the process, imprints it's own look on the image, and the print is never the same as the original. There is always some loss, some additions, and neither allow reconstructing the original.<br>So though a print may last long, it is not a preservation method of something that is not a print.

<p>... which is acceptable to me.</p>

... but not the aim of the exercise.<br>You can't, and thus don't, preserve an Ansel Adams photo by having me or you paint our impressions of it. If not the original form itself, the digital capture of the historic photos will already not be the same as the photos themselves. No reason to deviate (even further) by introducing the peculiarities of yet another process. That's not conservation or preservation.<br><br>If a print as a method has to be chosen, a printout of the hexdump of the files will always allow an exact reconstruction of the original data, i.e. allow to replicate the exact digital image, pixel for pixel. As long as there is knowledge of how these were arranged inside the particular file format (which is something that can be printed out too), it does not matter whether there are still machines and programs that can read the file format or not, because it will always be possible to enter the data in whatever container format happens to be current.

<p>This begs the question of whether you'd like to have something that's 95% accurate or have nothing at all. I choose the former. But of course my response, if you had read it carefully, is to archive both ways. The digital data might just survive, in which case the OP would still have the original image as faithfully represented as digital photography will currently allow. But if that fails (and it probably will, given enough time), the print is the fall-back. It might be spotted, and slightly yellowed. It might even fade a bit (even if it's a pigment print), but heck if it isn't still usable and restorable! Will it be as good as the digital file that was lost 50 years previously? Of course not. But at least it will be something, which is infinitely better than nothing.</p>

Well, i'm not one to argue... ;-) But:<br>The question is how to make sure you have something that is 100% accurate. And the answer to that is not to rely on something that is 95% accurate. There are better answers than that.<br>Of course 'at least something will be better than nothing'. But that begs the question why we would think the only other option would be to have 'nothing'.

<p>Are you certain the original digital image (even the RAW file) is 100% accurate? Is it even possible to have a 100% accurate image?</p>

Of course not, Sarah.<br>But the question was how to preserve what they have. Not how accurate the thing they have is.<br>Accuracy here means no more than how closely the thing they have in, say, 100 years from now still resembles what they have now.

<p>Then my answer to you is that archiving indefinitely into the future with no loss of data/fidelity is not possible. Any statistician will tell you that whatever is possible will eventually happen, given enough trials/time. The vernacular term is "Murphy's Law." If data loss is possible, then it will occur. Period. You can either accept that and do the best you can to preserve data as long and as well as possible, or you can simply not try. But to expect 100% success in your preservation methods is foolish.</p>

<p>Now if you'll pardon me, I've got to make bread.</p>

And that is why we should not aim for whatever is possible, and instead help data decay set in right at the start? Talk about foolish...<br>It of course makes no sense to attempt something and give up immediately because of fuzzy things like the idea that anything that can happen will happen.<br>If the goal is to preserve something as it is, the only sensible thing to do is try to preserve it as it is. Period.<br>That i could be so that there may be no 100% secure way to do so is why questions how to best attempt it make sense.<br>And if statistically anything that can happen will happen, it will indeed happen that these data will be preserved 100% as they are for ever (don't insult statisticians like that, by the way. They generally do know better). Except, of course, when the attempt begins with corrupting them straight away.

<blockquote>

<p>Are you certain the original digital image (even the RAW file) is 100% accurate? Is it even possible to have a 100% accurate image?</p>

</blockquote>

<p>It (the raw) is not. Certainly not '<em>accurate</em>' to the scene colorimetry or what we saw. Not even close.</p>