Creative Production Workflow Tips
3 benefits of storing metadata in XMP files
November 18 | 2015
Written by Brian Stuckey, Senior Product Manager at globaledit.
Most people don’t find metadata or file formats sexy. It’s all of the data around the image without any of the nice visual appeal of a photograph. So, launching into the realm of the esoteric, I wanted to take a moment to talk about XMP files.
XMP files are awesome. XMP files allow you to transfer your image metadata, ratings, tags, keywords, geolocation, and other attributes about the image outside of the actual JPG, CR2, or other file. The XMP format is a standard way of recording these data so that other programs can carry all of the additional data you added to your images.
This is super useful in practice. Most creative work doesn’t live in a single piece of software. You might start by capturing images in Capture One. There you might do your quick selects, add job information, and other relevant data. Next, you might organize the files using Adobe Bridge. From there, you might pass the RAWs off to a retoucher and get back a finished TIFF file. When you to import this into your DAM – the TIFF file may no longer contain the job information, photographer, talent, or the other metadata you meticulously added earlier.
This is where XMP files save the day — when you import them into your DAM, your software can read the metadata and append all of the missing information to the asset.
(One technical note: XMP content is frequently embedded in images such as TIFFs, JPEGs, and a few other formats. For the sake of this document, I’m focusing on sidecar file format. This is the kind where you have a file named SOMETHING.XMP in the same folder as SOMETHING.CR2.)
How XMP files work in practice
A single photograph might have several different associated files. For example:
image_2345.CR2 (The camera raw)
image_2345.JPG (JPEG preview)
image_2345.TIFF (High resolution retouched photo)
image_2345.XMP (Sidecar file that contains all of the shoot information)
XMP files typically have the same name as the photograph and live in the same folder as all of the other assets. Most software (Bridge, Lightroom, Capture One, Media Pro, etc.) are smart enough to understand the relationship and automatically recognize the data contained in them.
Example One: My vacation
Earlier this year, I went to several countries in Europe. I took nearly 7,000 photographs across three devices and was faced with a challenge: How do I tag my photographs, select the good ones, and come up with the Facebook album version?
The big problem is that Lightroom is meant for a single user. Two people can’t work on a library simultaneously.
I started by uploading all of my JPEG preview images to globaledit. JPEGs are smaller and the transfer time was short. This allowed my wife and I to see all of the images and to collaborate on them. I culled the 7,000 photos down to 500 and she picked out the final ~50. At the same time, we bulk-edited photos to add the city information and for a few key photographs, we added additional information (the name of a landmark, who was in the photo, etc.)
This distributed approach allowed us to quickly comb through thousands of images at the same time and be always in sync. I worked on my desktop and she worked on her iPad. When we were done, I downloaded the XMP files which only took a few seconds and added them to my Lightroom collection.
Now, Adobe Lightroom had all of my keywords, subject information, location, ratings and selects. I appended this to the RAW files and I processed the final 50 images in Lightroom.
Because I was able to quickly transport metadata in the XMP files, I could distribute my workflow without fear of losing valuable information about the images. I didn’t have to transfer tens of gigabytes of information to just make a few decisions.
Example Two: Backing Up
Adobe Lightroom is my go-to package for my personal photographs. I’ve spent thousands of hours adding information about people, places, geotagging, and categorizing my images.
What if I want to switch to, say, Capture One? What if my library becomes completely corrupted?
Adobe Lightroom allows you to actually write all metadata to XMP files in addition to their library. It slows the software down a little bit but this gives you a new level of security. As long as you back up the XMP files with your images, you’ll never lose any of your metadata.
Example Three: Migrating content between applications
Most of my RAW processing for casual images is done in Lightroom. However, I prefer the look and options that Capture One affords me and so my workflow is split primarily between these two applications.
I really like the organization tools that Lightroom provides. The face identification, geo-tagging, and keyword management tools are second to none for a home user. I use Lightroom to tag ALL of my images, regardless of whether I process them in another application.
To enable this mixed-environment approach to handling images, I use XMP sidecars to transfer the metadata I’ve curated in Adobe in to Capture One. Capture One reads the XMP sidecar files and all of my people tags are instantly recognized. I can process my images like I prefer and rest assured that I can always find my images by metadata regardless of the software I used to render the final image.
The XMP format was created by Adobe to standardize how metadata is stored and transferred. It is now managed as an ISO standard and free for anyone to use.
For more information about XMP: