Image tagging has long been one of the least efficient areas of our workflow. In previous years, photos of the site during excavation were taken with digital cameras. At the end of the day the trench supervisors were expected to upload the digital images to a computer, rename the files and add captions to the photos’ metadata with Adobe Bridge, and store them on the server. In the best of circumstances this meant that an image taken early in the morning would be tagged 9-10 hours later, likely after numerous other photos had been taken. To make things worse, as the season progressed and the trench supervisors became busier they tended to defer these tagging processes for a day or two. By that time they often had difficulty remembering exactly what was intended by the photo—even for archaeologists, it is sometimes hard to tell the difference between two photos of dirt.
This year we decided to move the captioning process out into the field. One of the key pieces of technology that enabled this was the Eye-Fi Connect X2, a camera memory card with built-in Wi-Fi. This card allowed us to continue to use dedicated digital cameras, which currently produce images of a much higher quality than the cameras built into many tablets and mobile devices, and to operate away from existing Wi-Fi networks. Using Eye-Fi’s Direct Mode we paired each card/camera with one iPad. After a photo is taken the card automatically broadcasts a Wi-Fi network to which the iPad connects. The card then transfers images to the iPad, putting the photos directly into the Camera Roll. The Eye-Fi app on the iPad does the actual transfer, but it can run in the background. The entire process takes from 30 seconds to a couple of minutes, depending on the number of photos.
Once the images have been transferred, we can begin the process of adding captions and labels to the images’ metadata with Photosmith. One of the great things about Photosmith is its ability to directly access the iPad’s Camera Roll. This means that once the images have been transferred, all we need to do is open Photosmith and begin tagging the new images—no importing required. We only used two metadata fields, title and caption. Title was used for a descriptive subject code (so for a photo of context 1003 from trench T.1000 at the San Giovanni site, the code would be SG-T.1000-1003), while caption was for standard, plain-text descriptions. This whole process sounds pretty involved at first, but once you have it set up it is actually quite simple and seamless.
After the images have been tagged and the iPads have returned from the field, the next task is to upload them to the server. This is done by syncing Photosmith with Lightroom over Wi-Fi. Once the images are in Lightroom you can edit them, delete duplicates, and do whatever image processing you want. Now the photos are ready to be handled like any other image that had been imported from a camera the old-fashioned way.
At this point we put the new photos into an “Incoming Images” folder and ran a FileMaker Pro script. Similar to the image handling used by PARP:PS, our FileMaker and Automator scripts created a thumbnail and ran some other actions on each file, then imported them into FileMaker using the Troi File plugin. The plugin also has the ability to import metadata along with the image. This allows us to not only insert the metadata into fields within FileMaker so they can be read by users, but also to create and manipulate other data based on the information contained in that metadata. For example, our scripts took that SG-T.1000-1003 subject code that was entered earlier and parsed it in order to automatically link the photo with its subject (context 1003). With the file renaming and relocating abilities of the Troi plugin, that image file could also automatically be renamed according to its subject, then moved to a particular folder for that subject on the server (I have not actually tried this yet, but see no reason why it would not work). By pushing the tagging process out into the field and using structured tags, the image processing and importing workflow can potentially become much easier, with little or no input required by the user.
While the Eye-Fi cards worked well most of the time, we did run into one or two problems. Fortunately, many of the minor performance issues we had with the cards seem to have been fixed in recent updates to the Eye-Fi app and card firmware. There are, however, still some technical constraints.
- Range: While we didn’t run into any problems on site, the cards’ range in Direct Mode is only around 30-50 ft. This is much shorter than your standard Wi-Fi router, which is good to keep in mind.
- Compatibility: While nearly all point-and-shoot digital cameras use SD cards, the DSLR market is rather mixed between SD and Compact Flash (all of Canon and Nikon’s sub-$1,000 camera bodies use SD, but the higher-end models mostly use CF). Eye-Fi cards are reported to work with SD-to-CF adapters, but with some compromises in performance.
At the moment there are two good metadata editing apps for the iPad: Photosmith and Filterstorm Pro. While both apps offer the same basic ability to edit metadata, there are some important differences.
- Filterstorm Pro’s FTP support offers more flexibility for uploading images to a server, while Photosmith currently only syncs with Lightroom (there is limited support for Dropbox, but that only works if you have internet access). Filterstorm Pro also allows batch editing of metadata, which is extremely useful when processing multiple photos of the same object or context.
- Photosmith has the neat ability to directly access the iPad’s Camera Roll, which makes everything run much more smoothly when you take several rounds of photos in the same day.* It also lets the user edit metadata with the image nearly full-screen, while Filterstorm Pro only allows editing in thumbnail view (this was the main reason we decided to use Photosmith). The developers of Photosmith have said that FTP export and batch editing are on their radar, which would address the two biggest issues we encountered.
We discovered many more little configuration and workflow tricks that don’t really need to be included here, but if you are interested to hear more you can leave a comment below.
Update: With version 2.0, Filterstorm Pro can also directly access the iPad’s photo library. This makes it a much more viable option for use with Eye-Fi cards.