(A draft from all the way back in January 2007...)
Megacorporations, professional photographers and other people who work with photographs for a living will all have their own methods of archiving and storing photographs. Amateurs don't - and amateurs have now all switched to digital, without necessarily upgrading from their old film-based workflow.
Here is some common wisdom about film photography:
- Don't expose film to light, or it will be ruined.
- Make sure you wind the film on before taking the next photo.
- Don't get fingerprints on the negatives.
- Don't throw the negatives away.
These are things I reckon the vast majority of the population understands - film photography has been around for long enough for its techniques to be ingrained in popular culture. The pitfalls of film photography form the plots of major movies; who doesn't understand that a destroyed print can be reproduced if you still have the negatives?
Not so for digital. The film-problems listed above are all magically solved with digital cameras: You can switch flash cards in daylight, the camera doesn't need to be "wound on", flash cards can be handled and don't mind fingerprints, and the file is the negative, so you don't need to store anything extra.
Since digital technology is so poorly understood in general, people see these advantages and assume that digital cameras are somehow magic. This is very, very wrong. Digital photography has its own entirely unique problems which the general public has never heard of because they've never seen them in a movie.
Here are some of them:
Proprietary file formats
JPEG: JPEG escaped from patent hell by the skin of its teeth last year. Forgent Networks attacked it with a submarine patent. The US Patent Office found the patent invalid due to prior art (and further found that Forgent knew about the prior art, making them extra-slippery fellows), and it has since expired anyway. Nevertheless, it's a good reminder of the dangers of software patents, and a reminder of why technically superior file formats like JPEG2000 aren't very popular and don't deserve to be.
Nevertheless, when we talk about photographic file formats, we may as well just talk about JPEG. It's universally supported in everything from mobile phones to DVD players, embedded into operating systems, and fully implemented in Free software. It's been around since 1992 and has even entered the Oxford English Dictionary as a noun. There are billions upon billions of JPEGs in the wild. It's not going anywhere soon.
Raw: Raw formats (sometimes called RAW files, although "RAW" isn't a file format) are all proprietary; they are the almost completely unmodified output of the camera sensor, and they vary from camera to camera. There exists Free software to decode them (dcraw), which you may have to rely on if proprietary software drops support for less common RAW formats in the future. Nikon earned themselves a stab in the face recently when they encrypted part of their RAW format in order to lock out competing software companies who make raw decoders. The issue is now resolved, and the white balance data in question is now freely readable by most software, but it's yet another excellent illustration of why you cannot entrust your photographs to proprietary formats. Adobe's Digital Negative (.dng) is open, supported natively in some cameras, and a good candidate for long term storage. Although Free software can read DNG files, I'm not aware of a Free raw-to-DNG converter like Adobe's DNG Converter. I would be inclined to keep JPEG copies of all raw files until you can convert them all to DNG.
In 1941, Jorge Luis Borges showed that having a perfectly preserved collection of JPEGs and Adobe Digital Negative files is no good if the collection is so large that you can't find anything in it. I have about 17,000 photographs from the last 3 or 4 years, and finding one in particular is non-trivial. An index of some sort is required.
Modern cameras generate lots of metadata such as date and time by which you can sort the photos. This is usually stored inside JPEG files as EXIF data, so you don't have to worry about losing it. Unfortunately, it's not really good enough to find a photo in a large collection. Ideally I would tag and comment every image, but back-tagging 17,000 photos is something I'll do when I've retired. I find that date-based folders are perfectly adequate for breaking the photo stream up into manageable, meaningful chunks. Backup is easier, and you don't get into the quagmire of category hierarchies ("should Wedding Photos go under Events or Places?").
Picasa stores captions in the IPTC metadata, but not labels (the Picasa word for tags). Adobe's Photoshop Album is even worse - all metadata is stored in a proprietary database which gives you no exporting options. Once you start using it, you had better keep using it forever.
Oh wait, Adobe Photoshop Album is no longer available! Too bad! Adobe encourages you to buy Photoshop Elements instead, which has similar functionality, and costs $90 instead of $50. I wonder how long it will be before they discontinue the $90 Elements in favour of a New Improved package that costs $130.
Seriously, don't waste time entering metadata into proprietary databases. If you want to get into tagging and commenting your photos, Free software is the only sane choice. I use Digikam and it's great.
Your hard drive is not reliable. One day it will break. It could happen...
Did it? It might have. A hard drive is a big whirring box of delicate mechanical components that are constantly snapping back and forth and wearing on each other. Maybe you've had a disk that's lasted for 8 years without a single glitch, but if ever past performance was no indicator of future results, it's here that it ain't. There's a very good reason why the hard disk manufacturers won't give you anything longer than a 5 year warranty.
RAID-1 is good insurance, but it's not trivial. If automated mirroring of your hard disk sounds like a good idea, it's worth the time and effort to set it up.
The only other backup scheme accessible to most people is to burn DVDs. My 26GB photo collection could fit on 6 DVDs, but it's actually spread out over many more, as I'm very aware of the fact that DVD media isn't 100% reliable either. Duplicates aren't a problem as long as you don't create a proliferation of versions without tracking them. Taiyo-Yuden make the highest quality DVD media; they cost about 30p each from SVP. A bargain-basement pack of 50 for £4.99 is not something you want to entrust your precious data to.
Ultimately, any media can fail, and you can always be arbitrarily unlucky no matter how much time and money you spend doing backups. I'm not going to suggest you invest in tape drives and expensive backup services, but there's one form of remote backup that's within everyone's reach: friends and family! Just give a DVD full of photos to each of your friends and ask them to store it for you. You can offer to do the same for them. Make sure you write down who's got copies of what!
There will eventually be generations who grow up understanding the above naturally, but until then, the transition to digital photography may create a cultural memory-gap when a generation fail to look after their pictures.