Heresy In Library Organization

Scanning the horizon, there are so-called workflow tutorials everywhere you look. Each and every one of them is promoted under the guise that it will help you become "fast and efficient." But what fast and efficient really means for most of the gurus out there is automating everything, and trying to do all 10 steps (or whatever) of the workflow at the same time, during import.

But automation may not be all it’s cracked up to be. What you’re doing when you adopt one of these workflows, unfortunately, is buying into a routine that’s narrowly dictated by the capabilities of the chosen software and by the design of someone else’s digital photo library. The reality is that the pro workflows are highly customized and rarely fit into a neatly prepackaged routine of doing everything during import. If your library design and workflow are well thought out, they will be based on your individual ideas about how you need to store, organize and find the photos that you work with on a daily basis. Your workflow and your library design are married in this adventure, and the library design is the driving force in the marriage. The workflow should follow the lead of your library design and should never dictate the basic structure or capabilities of the library.

Figure 1: A Short List of Collections by Subject

So let’s start with library design. Most photographers organize their photo libraries by subject because that’s the way our brains work and because that’s the way we organize objects in the real world. An example of organizing your library by subject would have portrait photos going into one folder and photos taken in Thailand going into another folder. I can smell trouble with this thinking right from the start, where two seemingly unrelated subjects, Portrait and Thailand, become top-level folder structures in your library. But once you have tens of thousands of digital photos, the system starts to break down. What happens when you take portraits while traveling in Thailand? Do they go into the Portrait folder or into the Thailand folder?

Besides that dilemma, once you have a large enough library, you don’t find some specific photo by digging through folders. So why organize them into folders by subject in the first place? Are you going to physically separate those photos that include portraits taken during one day in Thailand? The idea of exactly where a digital photo is should have been your first clue that you’ve maybe taken an outdated idea of organization a bit too far. Trying to think literally about where your photos are inside the computer or on a hard drive is an illusion anyway. It’s all just bits. There is no where!

Anyway, this is what databases are for. Once you’ve started using a database-driven cataloging software like Lightroom, the way that you approach finding your photos changes. With a visual database, there are basically two modes of search. First, if I’ve ever published the photo before, it will be in a collection. A collection is a virtual grouping by subject that might contain photos from a dozen different "places" (folders) on my hard drive. It’s a selection of photos that have been arranged to tell some story, but because it’s a virtual grouping, the photos need not be duplicated or physically together in the same folder. I don’t care "where" a given photo is; I can always find it visually by looking into one of a short list of collections. These collections have become the "organize-by-subject" structure that I rely on, rather than the physical organization of folders on disk (Fig. 1).

If the photo I’m looking for isn’t a member of a collection, it becomes a more general search, by text or keyword. Searching by text or keyword doesn’t return photos neatly edited into groups by subject the way collections do; rather it returns a lot more photos that have been tagged by theme. Common keyword or text searches might be for places like "Mendocino," or for broader subjects like "salmon" or "bridge." In my library, the results returned from searches like this will be much larger, so this is where chronology plays a role.

Once you’ve done a text or keyword search, narrowing the results becomes a visual search. And when you’re in visual search mode, chronologically is the only context you have to help you narrow the search down to the one.

Figure 2: When, Who, What, Where Folder Names

Then, there’s the question of legacy. While I don’t have any illusions that my own digital library will have great value after I’m pushing up daisies, I do at least want it to be understandable and navigable to those who inherit it. This requirement also dictates that having easily navigated chronological groupings in the file system will make more sense than subjective groupings like "Portrait" or "Thailand." So my library is indeed organized by subject, to the extent that the smallest unit of organization—what I call shoot folders—are named with the date first, in the format YYYYMMDD. Now they sort chronologically—both in Lightroom and in the file system.

I name each shoot folder with the four objective pieces of metadata (call them keywords, if you like) that I want associated with every photo in the folder: when, who, what and where (Fig. 2). So if I’m teaching in Italy for the ICP, my shoot folder names will look like this:

The When keyword is the date. Who can be a person’s name, a client’s name or even a job number. In this case, it’s the workshop organizer’s name, ICP. The What is a subject, in this case, Workshop, and the Where is generally the city and state. These keywords almost always will end up as IPTC keywords for individual photos, as well, but I want these four that I think of as objective metadata for the entire shoot in the folder name. This groups actual photo files together logically in the file system, even though I’ll rarely ever need to physically go into that folder for any reason. In the long run (we already talked about daises, remember?), having shoots organized into well-named folders that sort chronologically gives the library an easy-to-understand navigation should you ever need to dig into the folder view. It’s also helpful for legacy purposes.

Having this folder structure also gives me the peace of mind that if I ever decided to switch to Bridge or another database besides Lightroom, I would be free to do that. My library structure is completely independent of the catalog that I use to access it.

When it comes to daily workflow, I start by copying new photos into my photo library just like everyone else. You can do this during the import process using Lightroom, but when leading workshops I like to separate these two things. There’s so much confusion about the relationship between the Lightroom catalog and the actual photo library that I find splitting up these workflow items helps clarify things. It’s only after I’ve set up my shoot folders and copied the new photos to their permanent home in the library that I even start Lightroom.

If you copy your photos to your photo library first, once you’ve started Lightroom, the Import process is completely straightforward using the Add method. Photos are added to the catalog without copying them or moving them.

Next comes filenaming. I don’t renam
e photos during import for two major reasons. First, I want to preserve the original camera-generated filename in metadata, and if you rename during import, Lightroom simply throws that away. (Renaming after import, in the library, preserves the original filename in metadata.) The second reason is that I include the timestamp metadata in my filenames, so I want to be sure that’s accurate before I rename anything.

Figure 3: The Lightroom Filename Template Editor can be used to create complex filenames using nearly any piece of available metadata, including hour, minute and second info derived from timestamps.

As long as I’m structuring my entire library based on chronology, the exact moment a photo was shot is already the perfect sequence number. Why would I want to add anything to that? Relying on camera-generated sequences doesn’t work if you’re shooting with more than one camera, and camera sequences roll over occasionally, too. So for my library, the ultimate sequence number is made up of the date (YYYYMMDD) plus the timestamp (in the form of HHMMSS). Then, to preserve the original camera-generated filename, I include that at the end to complete the sequence. Including the date in the filenames also gives you a one-to-one correspondence for individual photo files and their parent shoot folder; you’ll never wonder in what folder any given photo belongs (Fig. 3).

Figure 4: Two Cameras, Chronological Sequencing

Getting two cameras synchronized for chronological sorting can be tricky, but I either can try to synchronize them to a world clock on my smartphone or GPS unit, or I can simply photograph the timestamp settings on one camera LCD using the other camera. Then when I get home, it’s simply a matter of comparing the timestamp shown in the photo with the metadata timestamp on that photo that was created by the second camera and adjust the difference until they match perfectly.

Once the timestamps from two or more cameras are in sync, I’ll re-name all the photos in this format: YYYYMMDD-HHMMSS-original filename.cr2. Once sorted by filename, photos from any number of cameras will display in the exact order they were shot (Fig. 4).

Remember, in the end, all you’re really doing is creating a sort based on a number that already exists in nature. An accurate timestamp! If you use an external GPS device, accuracy will be especially important when you’re auto-tagging photos to the GPS track.

The next step in the workflow is to finish up with other important metadata. If I didn’t apply a basic copyright during import, now is the time. The easiest way to do that, either during import or after, is to use a metadata preset. I generally apply a metadata preset to every photo I shoot, which includes all my relevant contact information, copyright, rights usage terms and URL info. As long as I still have everything selected in the Lightroom grid, I’ll take this opportunity to add keywords that are relevant to every photo in the shoot. When traveling on location, this generally will include location info, and generic search terms such as "tourism" and "travel" that I want applied to every photo.

Once I’ve applied global keywords to the entire shoot, I begin a very rough edit process that serves several purposes. Going through the photos one by one or in groups, I’ll add more specific keywords that didn’t apply to every photo in the shoot. It’s also during this first pass that I’ll start to generate a quick collection of my better shots. When I’m finished with this first pass through the photos, I’ll move to the quick collection to refine the edit. This is where I’ll check more critically for accurate focus and start to make my first tweaks to cropping and color correction on shots that deserve it. I’ll also narrow down sequences to find the best frames, eliminating the others from the quick collection. Once I have a rough edit completed, if the photos are worthy, I’ll make the quick collection a permanent collection.

Finally, it’s time for a quick export. When I’m traveling, I’ll frequently select out my best one or two shots from the day to email to friends and family. A couple of quick clicks in the Export dialog, and I’m done.

Of course, somewhere along the line, backing up has to be considered, too, especially when shooting on location. But the point is that the most efficient workflow routine will be the one that’s customized by you, based on your individual needs and on the design of your library.

Find George Jardine’s extensive collection of tutorials at his website,

Leave a Reply