In the past I used Geotagger in conjunction with Google Earth, but that method writes the GPS data to JPG files. I wanted a way to geoencode the original RAW files in Lightroom. Jeffrey Friedl’s plugin makes this possible, since it can take GPS coordinates from a tracklog and store them in a Lightroom catalog.

Step 1: Record a Tracklog

The first step is to record a tracklog while you’re out shooting. If you don’t already have it, install Trails from the App Store.

Make sure your camera’s clock is set to the same time as your phone, then start a new track. Lock your phone and put it in your pocket. As you move around, Trails will track your movement.

If you have an iPhone 3Gs or iPhone 4 make sure you have the latest version of iOS. The multitasking allows Trails to run in the background and continually record your position.

On iPhones without multitasking, Trails will do its best to make a good tracklog by connecting good data points with a straight line. The following screenshots illustrate the difference between Trails tracklogs produced with and without multitasking.

The map on the left shows a trail recorded on an iPhone 3G without multitasking. The app was able to plot about a dozen data points and connected them with straight lines.

The map on the right shows a tracklog recorded on an iPhone 4 with multitasking. Trails was able to frequently get new GPS data points, and the resulting tracklog is pretty accurate.

Trails causes your battery to run down faster than usual. Make sure you remember to turn it off when you’re done shooting, and periodically check that you have enough battery life to make it to the next charging opportunity in case you need to make an emergency call.

Step 2: Download the Tracklog

One of the easiest ways to get a Tracklog out of trails is emailing it to yourself. Open the tracklog from the “My Tracks” area and touch the “Export” button in the bottom left corner of the screen. One of the export options will be “Email” — enter your address and you’ll get the tracklog as both GPX and KML files.

The KML file can be opened in Google Earth. The GPX file is what you’ll need for geotagging in Lightroom.

Step 3: Sync Photos to the Tracklog

If you don’t already have it, download and install Jeffrey Friedl’s “GPS-Support” Geoencoding Plugin for Lightroom. Jeffrey makes great Lightroom plugins so be sure to leave a donation.

In Lightroom, select all the photos you wish to geoencode and click File → Plugin Extras → Geoencode…

Go to the Tracklog tab and find the first field labeled Sync to tracklog(s). Click the “Browse…” button to open the GPX file you got from Trails.

Choose your timezone from the next drop down menu, then optionally adjust the “fuzziness” and time compensation fields.

Click the Geoencode Images button and Jeffrey’s plugin will record the map data for each photo in your Lightroom Catalog’s database. This process works by checking the timestamp on each photo and looking for the closest GPS data point in the tracklog.

Now when you export JPGs and upload to publishing services from Lightroom, you can include this “GPS Shadow Data” (you’ll see new options in the Export settings window). Programs and services such as iPhoto, Picasa, Flickr and SmugMug can read this data and plot your photos on a map.

Remember: this means anyone can see where you were when you took a photo. It might not be a good idea to do this with photos taken at home or work.

Here’s the first of a new series I’m thinking about: Photoshop before & after. I’ll pick a photo and show how it looked straight out of the camera vs. the finished product, then explain what was done.

This photo of the Golden Gate Bridge was made last February from the Marin Headlands. We were extremely fortunate to have clear skies and great weather on the night we headed across the bridge the see the bridge and skyline. The sun is to my right and very close to the horizon, putting a nice even lighting across the entire frame with subtle highlights on the span of the bridge.

As usual I shot this in aperture priority… most of the time I prefer to frame up, choose an aperture that gives whatever depth of field I want, adjust the exposure compensation to make sure nothing gets blown and let the camera do the rest of the work. I’m at f/10 with the EF 135mm f/2 L lens. It’s a particularly sharp lens, and the focal length does a nice job of compressing the scene.

Here’s a large version of the image straight out of camera:

Time to come up with a plan. I want to…

• bring out the colors in the sky.
• emphasize the glowing side light on the bridge.
• brighten the foreground and middle ground without affecting the background too much.

Default Settings

The base settings I use on almost every photo are applied on import to Lightroom:

• Adobe’s default treatment: +5 blacks, +50 brightness, +25 contrast.
• Some color and clarity: +20 clarity, +25 vibrance, +8 saturation.
• A slight curve to create medium shadows and slight midtones: +6 lights, -9 darks, -25 shadows.
• HSL and profile tweaks specifically for the EOS 5D.

Basic Processing

The vast majority of what I want to do can be accomplished with just the basic exposure, brightness, contrast and color sliders.

• White Balance: I like what the camera chose, so I’ll leave it “As Shot” at 6350 and -8 tint.
• Exposure: -1.05 since the image is looking very bright.
• Fill Light: +11 to slightly bring out shadow detail.
• Contrast: +43
• Clarity: +40
• Saturation: +44

The colors and contrast in the original are very flat and faded. I’m no meteorologist but I think that’s partly due to atmospheric conditions combined with the fact that I’m shooting from far away with a telephoto lens. Adding a lot of contrast and saturation pushed the colors and contrast to where I want them.

Clarity is a great tool for emphasizing edges and adding a little local contrast. Being heavy-handed with this slider can wreck havoc on portraits, but you can get away with it on a landscape.

I want heavier shadows, which can be done with a different tone curve than what I originally chose:

• Highlights: 0
• Lights: +8
• Darks: -10
• Shadows: -50

Fine Tuning

Most of the work is done, but there are still a few minor details that I want to adjust in the sky. I would like deeper and more saturated colors there, and more details in the clouds.

I can use the HSL sliders to adjust the colors in the sky without affecting the bridge, water or hills.

• Purple Saturation: +45
• Magenta Saturation: +45
• Purple Luminance: -24
• Magenta Luminance: -14

To make the bridge span glow a little brighter I’ll also slide the red luminance to +20.

For increased cloud definition I’ll use the Adjustment Brush to paint +100 clarity in the sky.

Final Image

Do some basic sharpening and slight noise reduction, and we have the end result:

In this unprocessed photo you can clearly see a lot of dust spots. At large size there are even more… hundreds of them. Some are unmistakeable, but others are small, light and while not immediately visible they still detract from the image quality.

In this photo in particular, the dust spots will cause problems throughout processing. I want to bring out the clouds in the sky with a great deal of clarity. I’m thinking about burning certain areas, and brushing over others with +100 clarity a few times. That’s going to bring out the dust even more, so it’s important that I get rid of as much as possible.

In a case like this, instead of squinting at the image with my nose inches from the screen I’ll push the blacks slider far to the right. Since the majority of the frame is lightly colored, pumping the blacks all the way up will reveal the dust spots. Here’s what it looks like with blacks at +90:

Now it’s much easier to see all the dust and I can get blast them away with the Spot Removal tool much faster. After a few minutes I’m able to get rid of the dust, pull the blacks slider back down to a reasonable amount and proceed with processing. Here is the finished image:

In the beginning I just used labels to mark folders waiting to be processed, or just remembered all the shoots which were unfinished. But as my photo library grew this solution did not scale. When you are dealing with thousands of photos, or files of any type, trying to remember what needs to be done and constantly sorting through the pile becomes a tremendous waste of time. You need an efficient way to search. You need automation.

Adobe Lightroom’s smart collections are a great way to solve this problem. With a carefully planned set of rules and list of smart collections, Lightroom becomes a powerful self-updating dashboard for your workflow. It frees the photographer’s mind from having to think about moving files around between folders, albums or collections and makes keeping a mental inventory completely unnecessary. All that is required is to process and export photos at will.

The workflow that follows is heavily based on “Workflow Smart Collections” by John Beardsworth. Where John’s purposes necessitate tracking recent shoots, title, keyword, caption & copyright meta data, and specific exposure settings I am merely concerned with simply processed vs. unprocessed and published vs. unpublished.

Here is a snapshot of my entire system at work:

Only Lightroom Matters

Absolutely everything gets imported into Lightroom, even snaps from my iPhone. I used to keep RAW files separate from JPGs shot with camera phones or point & shoots, but this became troublesome because I was attempting to maintain two different databases of photographs. I should probably have my computer science degree confiscated for that brilliant idea.

So Lightroom is the basis for every image I produce. From here everything gets published out, to Flickr and SmugMug, to this blog, to iPhoto and to folders on my Home Theater PC for slideshows. If an image or a piece of meta data doesn’t exist inside Lightroom, it doesn’t matter. If I want to make a change to a photo, I make those changes inside Lightroom and re-export or re-publish the image.

The Culling Process

Once a batch of photos is imported I flip through them all and flag whatever I like. If I really like an image it gets a flag and one star. Everything that’s flagged will eventually be processed, and the starred photos are what I consider portfolio-worthy.

Unflagged photos are the rejects. If hard drive space were more expensive I might delete them, but it’s cheap so I keep them around and will perhaps take another look at them some day.

At this time I also usually batch-assign keywords and GPS data. Titles are given at the time an image is processed.

Now, let’s get into some smart collections.

Workflow Step #1: Sorting

The RAW Flagged, Processed smart collection shows me all the keepers I’ve finished. Photos must match all of the following rules:

• Pick Flag is Flagged
• Has Adjustments is true
• Any of the following are true:
• File Type is Digital Negative (DNG)
• File Type is RAW

The RAW Flagged, Unprocessed smart collection shows me all the flagged RAW images that I have yet to processed. Images must match all of the following rules:

• Pick Flag is Flagged
• Has Adjustments is false
• Any of the following are true
• File Type is Digital Negative (DNG)
• File Type is Raw

The RAW Unflagged smart collection shows me all of the rejected photos. Images must match all of the following rules:

• Pick Flag is unflagged
• Any of the following are true
• File Type is Digital Negative (DNG)
• File Type is Raw

The Starred smart collection shows me everything that’s portfolio-worthy. Images must match all of the following rules:

• Rating is greater than or equal to one star

The Processing Unnecessary collection is a manually maintained collection of files that I’ve flagged and want to keep, but don’t need to process. JPEGs that I think look fine straight out of a point & shoot camera or iPhone get dropped in here. Doing so allows me to build an accurate “Processing Queue” smart collection in the next section without bloating its image count with files that I will never process.

Workflow Step #2: Processing

The Finished smart collection shows me every photo that’s considered done. Images can match any of the following rules:

• Collection contains all Processing Unnecessary
• All of the following are true
• Pick Flag is flagged
• Has Adjustments is true

The Processing Queue smart collection contains every image that is waiting to be processed. It’s not really a queue since its contents aren’t processed in first in, first out order. I just randomly pick a few at a time to work on. But “queue” is descriptive enough. Images must match all of the following rules:

• Collection doesn’t contain Processing Unnecessary
• Pick Flag is flagged
• Has Adjustments is false

Workflow Step #3: Publishing

I publish photos to Flickr based on Thomas Hawk’s workflow. Images are sorted into an A list and B list. Each time I publish to Flickr, I first send a batch from the B list, then send one from the A list. Doing this ensures that contacts who only view the one latest image from each of their contacts will see my best work.

The Published to Flickr Manually is a manual collection of files that I previously uploaded to Flickr through their web form, from my phone, or from another application. This allows me to create smart collections that are aware of images that were uploaded to Flickr, but not via Lightroom’s publishing features.

The Possible Flickr A smart collection shows me all of what I think is my best work, which I might want to put on Flickr. Images must match all of the following rules:

• Rating is greater than or equal to one star
• Pick Flag is flagged
• Publish Collection doesn’t contain Photostream
• Collection doesn’t contain Published to Flickr Manually
• Collection doesn’t contain Flickr Upload Queue A
• Any of the following are true
• Has Adjustments is true
• Collection contains all Processing Unnecessary

The Possible Flickr B smart collection contains everything that’s not portoflio-worthy which I might also want to put on Flickr. Images must match all of the following rules:

• Rating is less than one star
• Pick Flag is flagged
• Collection doesn’t contain Published to Flickr Manually
• Collection doesn’t contain Flickr Upload Queue B
• Publish Collection doesn’t contain Photostream
• Any of the following are true
• Has Adjustments is true
• Collection contains all Processing Unnecessary

The manual collections Flickr Upload Queue A and Flickr Upload Queue B contain everything I’ve chosen from the “Possible Flickr” smart collections and decided to queue up for uploading at some point in the future. If you look back at the rules for both “Possible Flickr” smart collections, you’ll see that they automatically update themselves to not contain anything that’s been queued up here for uploading.

These collections are also not true queues. I just pick images at random to upload.

The Ready for SmugMug smart collection lists every finished image that has not yet been archived to my SmugMug account. Images must match all of the following rules:

• Uploaded to SmugMug is No
• Pick Flag is flagged
• Any of the following are true
• Has Adjustments is true
• Collection contains all Processing Unnecessary

Workflow Step #4: Published

Finally, I have a few smart collections that show me what’s already been published.

The Flickr’ed smart collection shows me everything in my Flickr Photostream. Images can match any of the following rules:

• Collection contains all Published to Flickr Manually
• Publish Collection contains all Photostream

The SmugMug’ed smart collection contains everything that’s been archived online to SmugMug. Image must match all of the following rules:

• Uploaded to SmugMug is Yes

The Blogged collection is a manual collection of everything I’ve featured on Totally Sweet Photos. I manually add images to this collection as I blog them.

That’s It

This seems like an incredibly complicated workflow, however since the majority of the tools involved are self-updating smart collections I put forth virtually no effort in order to keep track of thousands of images. In fact, in the past I have managed lesser amounts of photos with fewer steps in what were much more complicated and time-consuming workflows. Thanks to automation, all I really do now is work on my photos and trust Lightroom to keep the books straight.

One again, here is everything in action:

And also once again, thanks to John Beardsworth on whose workflow this is based.

Since camera vendors don’t release the exact specifications for their sensors, the only software that can natively get a camera’s raw image data absolutely correct is the vendor’s software. Although some of this software is very good, many photographers prefer Adobe’s products. Third party image editing software such as Lightroom is unable to get this image data exact due to the lack of a widely adopted open RAW spec, so it does the best job it can. In order to replicate the colors and contrast we see in-camera, we need to calibrate and tweak Lightroom’s settings on a per-camera basis.

Lightroom Camera Profiles

With the stock Adobe settings, RAW images look a bit flat. Specifically for the 5D, I find that the reds look very poor. By using a color profile calibrated for your specific camera, you can improve the image quality. Adobe includes DNG Camera Profiles that do a very good job of improving the color. These are included by default with the latest Lightroom distributions. Users with older versions will have to install them manually.

If you make frequent use of the recovery slider, you will notice that doing so tends to wreck havoc on skin tones. Jacksonville Florida Wedding Photographer Thomas Lester released some Untwisted Camera Profiles that correct this problem. I installed them and they worked well, but I thought the Untwisted Profiles gave my images a green color cast. Thomas’ Invariate Camera Profiles solve this problem — I find them to be a great middle ground of corrected recovery slider performance without the green color cast. These are the profiles I’m using on all my 5D RAW files, specifically the Invariate Camera Standard profile.

Tweaking Lightroom for the Canon 5D

The RAW images look a lot better with a proper profile, but we’re still not done. The reds and oranges look really intense, especially in tungsten light. In order to get them under control, go to the HSL panel and pull the red and orange saturation sliders down to -10, then push the red and orange luminance sliders to +10. I got this tip from Lansing and Grand Rapids Michigan Wedding Photographer Brett Maxwell. Some people also like to push the blues up a little, but that doesn’t look necessary to me.

I’ve used these same settings to process RAW .CR2 files from both the original EOS 5D and the 5D Mark II. I was happy with the results for both cameras without any additional adjustments.

Sample Images

Take a look at these sample photos. Both have the exact same development settings, including white balance and a slight s-curve. The only difference is that the one on the left has its saturation and luminance HSL sliders set to 0 and uses the stock Camera Standard calibration profile, while the image on the right uses the HSL settings and invariate untwisted profile described in this article. When you look closely you’ll notice that the image on the right has much nicer colors, especially in the tulip petals and the yellow cab.

Adobe’s upcoming “Content-Aware Fill” feature is able to dynamically analyze an image and automatically make guesses as to the best way to synthesize texture, as John Knack says. The examples get better as the video goes on, and the last scene in which Bryan O’Neil Hughes adds a mass of content to a panorama with uneven edges looks like complete and utter magic.

The RAW image shown to you by Lightroom often does not come close to matching what you saw through the viewfinder or on your camera’s LCD. Granted the image on your LCD has been run through the camera’s processor and the point of RAW is to leave the decision making up to you, but often Lightroom does not display certain colors they way they truly appear. Kerry notes that the difference in reds between his Canon 30D and Lightroom is often the most noticeable example of this problem and I agree — the reds I see when I begin editing in Lightroom are definitely not what I saw through the viewfinder of my Canon 5D.

Lightroom 2.2 addresses this issue by providing a set of pre-packaged camera profiles that help the software see the RAW image in the way your camera intended. The results of using these camera profiles are clear as day in the example picture below. The version on the left was exported with Adobe’s ACR 3.3 profile, and the one on the right had the “Camera Standard” profile applied. As you can see, the “Camera Standard” profile did a great job of correcting the colors, most noticeably in the woman’s bright red shirt.

Kerry has also posted a video “Adobe Photoshop Lightroom 2.0 Camera Profiles” which further explains the purpose of camera profiles and shows you how to apply them to your pictures.