Since my trip to Alaska in 2015, geotagging has become something that I’ve become increasing interested in. There were a number of images that in review I would love to know where I was when I took them. Unfortunately, barring some specific circumstances, one fjord looks largely like another, and tagging after the fact in Lightroom isn’t really feasible for a lot of the images.
Of course, the easy way to geotag is to have a camera that’s capable of handling it on its own. This was one of the big selling points I saw in the Canon EOS 1DX mark II, and one of the features I was disappointed in both the Canon 80D, as well as the newest Nikon bodies the D5 and D500.
Barring a built in GPS, a compatible external GPS unit that can feed data directly into the camera is worth a thought — perhaps even a real long hard thought.
At $250 Canon’s GP-E2 is rather expensive for what it is, and while it can communicate with most modern Canon cameras, including my 5D mark III, through the hot shoe that means it has to be in the hot shoe.
There are two huge positive points for using a GPS that can tag images directly in the camera. First, is you don’t have to worry about timing at all. When you take a photo, the camera reads writes out the current coordinates to the image and it’s properly tagged. There’s no messing with time zones or time offsets or keeping clocks in sync, and honestly this alone might be worth the price of entry.
The second positive point, at least in the case of Canon’s GP-E2 is that it has a digital compass, and logs not only location by direction. Admittedly Nikon users with the GP–1A units won’t have that feature. And honestly I’m not sure how important compass data really is in the grand scheme of things.
Moving even further away from automated in camera tagging is running a GPS unit that does track logging. Using an external logging unit requires the most amount of fiddling but has the potential to be the cheapest option as well. And of course, with an external device, the whole time syncing situation crops up — a problem that’s exacerbated by the decisions made by both standards bodies (EXIF) and software developers.
The final, and potentially cheapest option is to use a smart phone and some logging software to do the GPS logging. This of course, hash the advantage of using hardware that you probably already have (your smart phone).
That said, there are two downsides to using a smart phone instead of a dedicated GPS unit. First, is the simple fact that it’s going to drain your phones battery. If you have a big smart phone (like an iPhone 6s Plus) that may not be all that big of a deal, they have big batteries. If you have a smaller phone, like my iPhone SE, the battery drain can be significant, and all day use may be out of the question.
The second issue is one of ruggedness. Most stand alone GPS units are designed to be used in harsh environments. They’re usually water resistant to some degree, and often are build in rubberized hardened case that can take some bumps and drops. For a GPS receiver, where having an unobstructed view of the sky is kind of needed, drops and bumps become more of a potential problem.
Phone Battery Life
Battery life is the always present concern for this, either as a stand alone device, an on camera device, or in this case a repurposed iPhone 4. I do have to note, I replaced the battery in my iPhone a little over a year ago, so even though the phone is old the battery is basically new.
Prior to having the phone available, my top choice for a GPS unit was the Garmin eTrex 20. The rated battery life is 25 hours on a pair of AA batteries. Outdoor Gear Lab seems to have gotten similar performance but they didn’t specify how they were using it though. I suspect, if you’re running a track log, the performance will probably be worse.
The question regarding battery life though is just how much do you actually need? I mean if you’re only logging for 5–6 hours a day, 25 hours would give you multiple days of use, but then you have to track where you are so you don’t end up with a dead battery in the middle of the day.
I’ve started doing some basic testing using pretty absurd sampling rates in GPS4Cam Pro — including 1 minute time based sampling and 10 meter position based sampling.
In both cases I was seeing about 4% per hour of battery drain, though of course there are a lot of caveats to this. At 4% per hour, running nothing but GPS4Cam Pro and keeping the screen off, using 90% of the iPhone 4’s battery should yield around 22.5 hours of continuous GPS logging. Given that that’s all I intend use this for, that’s way more than I reasonably expect to use in a day.
In a second more realistic test I ran GPS4Cam Pro on my iPhone 4 for just over 6–1/2 hours, using position based capture settings (capture every 50 m, conveyance automatic, accuracy filter 20 m), and consumed 35% of my phones battery. Over that time, I drove some and I walked some, and all told captured 207 track points (an average of about 31 points an hour). As far as battery life goes, that works out to about 5.3% per hour in a more realistic scenario, pro about 17 hours of runtime on a 90% battery consumption.
All told, these kinds of power consumption numbers aren’t something I would necessarily want to see if I was using my primary phone to do the GPS track logging. However, for a dedicated old phone that’s no longer has service, 17–20 hours of logging isn’t a horrible proposition. Moreover, things may improve even more if I pull the SIM card — which I’ve been lead to believe should completely disable the cellular radios.
That said, more testing is warranted.
Intervals and Accuracy
Actually that’s a good point to segue into talking about just what kinds of intervals and accuracy is really needed for geotagging images and for what uses. This is of course, a personal decision and it’s going to reflect what you want to do with your geotagging and your photography.
Actually you can even expand this to compass data — something the GPS4Cam doesn’t provide, but a built in GPS unit or GP-E2 can. Do you really need to know which way you were pointing when you took the picture?
There are two kinds of accuracy that need to be discussed here.
First there’s the accuracy of the GPS point itself. There’s not much you can do about this it’s limited by the GPS unit. In practice, you’re looking at somewhere between 3 and 10 meters of accuracy (call it 10 to 32 feet). You can’t change this, so there’s no much point in worrying about it.
The more pressing value for a photographer is the accuracy of the track over time, and by extension the interval between points.
GPS tracks are made using a series of track or way points taken at intervals, either distance or time based.
But what about pictures that are taken between those update intervals?
Most Geotagging software will interpolate a position between the two closest logged points based on a straight line approximation.
For example, you can setup GPS4Cam to record a track point every 5 minutes. If you’re just walking at the typical speed of 3.1 MPH (5 km/h) that will result in a track point taken every something like 1360 feet or so (assuming you walk in a straight line). If you’re driving at 60MPH, the space between those track points will be 5 miles.
An alternative is to use distance; say every 50 meters. In this mode, as you move faster, the GPS logs more points at roughly the same distance, as opposed to bigger gaps between the points. However, the price for this kind of accuracy is power consumption.
So the obvious question is what mode should you choose? And beyond that, how frequently should you sample for the points?
Okay, I have to come clean here. I haven’t figured this out yet either.
Time based intervals, especially relatively long ones, have a huge advantage of not using tons of power. At the same time, they run the risk of not having an accurate position either. They’re also going to be much more subject to certain kinds of positional errors.
For example, if you’re driving at say 60 MPH, with 5 minute logging interval, and your drive 2 miles down a road, come to an intersection, get out take a picture, and then drive 2 miles down the road to the right. The interpolated track log will, on average at least, put you 1.4 miles out in the middle of nowhere not even close to the road. At which point we go back to the question, does that matter to you?
Time and Clocks
As I noted in the beginning of this, the big problem with external GPS logging is matching times. All GPS units record position information in UTC (Coordinated Universal Time). At the same time, cameras following the EXIF standard record time as the local time, and with no provision to store time zone offsets.
Moreover, for the most part, it’s pretty clear that camera industry and the post processing software people haven’t really thought through all the hard stuff with respect to times and time zones.
For example, Lightroom assumes that that the time offset for the camera is the same as the computer’s current time zone. There’s a mechanism to adjust that as that won’t always be right, but it’s still somewhat problematic.
What even more frustrating nit for me is if you’re somewhere that has you crossing the time zone boundary. Las Vegas, for example is in pacific time, but the Grand Canyon and Zion are in Mountain time. The OCD part of me wants to have the local time correct in both time zones, but the realist in me is saying that there’s no way I’m going to remember to stop and restart track logs at the time zone boundary just so I can get the right time stamps to work in Lightroom — Lightroom doesn’t have an apparent mechanism for dealing with tracklogs that cross time zone boundaries.
Even setting the camera to UTC doesn’t really help, as I would both have to adjust the time on every picture to reflect the local time, and I would have to offset all the GPX tracks to UTC because of Lightroom’s time zone assumptions.
GPS4Cam has their own “solution” to this problem which involves photographing a QR code when completing the tracklog and using their free desktop software to set the coordinates on your RAW images as you download them from your card. Though personally this is not a step in my workflow that I’m really interested in adding.
The more time I spend looking at digital photography the more holes I continue to find in both the software and workflows. Geotagging, especially with an external GPS unit, is absolute no exception to that rule. In fact, the only real good Geotagging situation is to have a camera integral GPS unit that tags at capture. Then there’s no questions about time zones and syncing images.
As much of a boon as a camera integrated GPS unit would be in terms of simplifying the whole post processing logging process. I’m still committed, at least for the moment, to doing this on the cheap. Mostly for two reasons, first, I don’t do a lot of GPS logged trips. More importantly, there’s a very real possibility that my next camera may have an integral GPS unit. Either one, and especially both, make buying a GP-E2 a poor idea to me.
A standalone GPS doesn’t really solve any problems that I have with reusing an old phone, except maybe it gets better battery life under some conditions. But the problems relating to clocks is still an issue.
This is definitely a topic I plan on revisiting in the future. One of the biggest problems that I see in trying to work out a GPS workflow is testing edge cases without having to actually having go out and get a real world GPX log and images that say cross timezones. I have some ideas for doing that, but I’m not sure that they’ll be practical to get done before I actually have to use the GPS in the field.
This has become a lot more rambling than I had intended, so I’m going to cut this off here and hopefully will be able to come back to more specific parts of this topic in the future.