How
I take a lot of digital photos. For the 'why', I'll refer you to this page. For how I keep them organized. keep reading.
I take at least a photo a day. When there is a big event, it isn't unusual for me to take more than 200 images in one day. This makes it difficult to find "that one image" or "all photos with a duck". The problem is only going to get worse. I'm not going to stop taking photos. As of this writing, I'm up to almost 60,000 images. Now that I have everything on one large external drive, I decided it is time to get serious about organization.
I started with some constraints. Whatever organization scheme I use must be able to deal with 60k (and counting) images. I wanted to be able to quickly and easily find photos from a particular date and photos with particular content. This started me down the field of tagging. I need an organization scheme that will continue to accomodate new images as I take them. This means I didn't want a classification scheme that would have to be redone if I ended up with too many of one type of image.
I didn't want to spend a huge amount of money on software. I was leary of any photo organizing system which put the meta-data into a database where it could lose the link with the original image if the database became corrupted.
I also wanted an easy way to make web photo galleries. If someone sees a photo on the web and wants one of the originals, I wanted the file name to remain consistent. I didn't want my web galleries to have different file names than my originals.
My first decision had to do with whether I went with some photo management system with a central database or if I came up with some other way of tagging the images. I used to have more photos than hard drive space which would have made a database-driven system difficult. Although there are photo management applications which work acrosss multiple volumes, I didn't want to keep track of that many CDs. I didn't seriously start my search for an organizational solution until I bought an external drive large enough to hold all my photos. Now about that centralized database...if I went with an image management system (something like iPhoto or a professional system like something from Canto) I would be tying myself to one vendor's solution. If that vendor changed directions or went out of business, I'd be stuck with all of my metadata in a proprietary database. I also had a fear of database corruption. While iPhoto database corruption is rare, it does happen. If I ran into that problem, I'd likely lose all my metadata. All work I'd put into categorizing photos would be lost.
My other problem had to do with how to categorize things. I'd first been introduced to the concept of 'tagging' a photo using keywords within Photoshop. However, Photoshop's interface wasn't going to work for 60,000 images. iPhoto also includes some methods for adding keywords to photos but the interface isn't all that easy to use. I also had scale issues. With 60k+ images, I needed more keywords than will comfortably work with the iPhoto interface.
The other problem has to do with taxonomy. If I take a picture of a duck, am I going to remember to also tag it as 'bird' and 'fowl?' I might remember that some of the time but certainly not every time. While folksonomies have their place, I needed more structure. I wanted my tagging to be consistent and to not rely on me to remember all the synonyms for a given term. I needed a formal taxonomy. I also needed something that could be modified. I didn't want to take someone's taxonomy and never modify it. I wanted to be able to add people's names along with synonyms for nicknames.
I also needed something that would allow me to rely heavily on drag & drop instead of typing everything. I didn't want to introduce typos to my tags. Selecting tags from a pre-populated list would help reduce typos.
Finally, once I get all these images tagged, I need to be able to find them. Whatever metadata is stored, must be stored in a way which I can find it again easily. Remember that picture of a duck? I want to be able to easily find that one duck picture or many duck pictures. Metadata is only as useful as what it will help me retrieve.
I'd already been renaming my files using the time stamp in the EXIF information. My file names are normalized. I can easily find photos for a particular date by searching on the file name.
My images are all in JPEG format. I'm not taking photos for publication. I also don't have a high degree of confidence in the longevity of any particular vendor's RAW format. I think JPEG is here to stay. While this is a lossy format, it's good enough for my purposes.
Another advantage of JPEG is the ability to append text data, like a tag, directly into the image file without affecting the image. The metadata is now stored as part of the image file. It isn't going to be lost with any database corruption or sidecars that somehow lose their associaton with the original image file.
My metatagging solution? IPTC, Controlled Vocabulary, Image Info Toolkit and Spotlight in Mac OS. IPTC is a standard for associating metadata to images. Controlled Vocabulary is a list of keywords put together by someone who knows how to do a taxonomy. I used that as my starting point. I've modified it to include things like people's names. Image Info Toolkit has some UI issues that bother me but it works well enough. The 'batch' and 'template' features allow me to easily tag large numbers of photos in a reasonable amount of time. The interface is awkward but not painful. I still think there is too much clicking involved to tag one photo but this is the best I've found to date. This isn't a perfect solution but it works.
Image Info Toolkit adds the meta data to the JPEG file. I choose tags from a Controlled Vocabulary list. Synonyms are automatically added. As long as I stick to clicking and dragging, I can't make a typo.
Once the meta data is a part of the JPEG file, Spotlight will find it. I can use Spotlight to find all those images of ducks.
When I create a web page, I use Photoshop's automated 'build web gallery' tool. This creates a simple web gallery which does not rename the files. This means I can publish a web page of images. When someone wants a copy of one of the original images, I can find that image using the original file name. This is one of the big reasons I don't use iPhoto for my web galleries. IPhoto renames the image files. If someone wants an image, I have to figure out which original image matches the image in the web gallery. I will never use a web gallery tool which doesn't preserve the original file names.