The Drone Metadata Mystery: Why Your Photo Altitude Doesn’t Match Your Controller

krdroneworks
6 minutes ago
4 min read

By: Colonel (ret) Bernie Derbach, KR Droneworks Academy, 07 June 26

If you’ve ever pulled up the metadata of a photo shot on your drone, you might have noticed something deeply bizarre.

You place your drone on the grass, wait for a solid GPS lock, and verify the controller reads exactly 0 meters. You take off, climb to a verified 1 meter off the ground, and snap a photo. But later, when you open that photo’s properties on your computer, the metadata claims the camera was sitting at 17 meters.

By the time you fly up to 90 meters, the photo properties insist you were actually at 120 meters.

What gives? Is your drone broken? Is it a software bug?

Rest assured, your drone isn’t glitching, and you aren’t alone. This is a notorious quirk that confuses thousands of drone pilots. Here is the breakdown of why this happens across virtually all modern drones—and why it’s actually supposed to work that way.

The Root Cause: Two Different Reference Points

The mystery boils down to one simple fact: Your drone’s controller and your photo’s metadata are measuring height from two completely different baselines.

1. The Controller Measures "Relative Altitude" (Height Above Takeoff)

When your drone is sitting on the ground and you see "0m" on your screen, the drone is calibration-stamping that exact spot as its baseline. To do this, it primarily uses an internal barometric altimeter (which measures changes in air pressure).

When you fly up 90 meters, the controller reads 90 meters because the air pressure dropped by exactly 90 meters worth of atmosphere relative to your launch pad.

2. The Photo Metadata Measures "Absolute Ellipsoid Altitude"

The moment you click the shutter button, the camera embeds data into a standardized image tag called EXIF:GPSAltitude. This tag doesn't care about the patch of grass you took off from. Instead, it pulls raw data from the GPS satellites using a mathematical model of Earth called WGS84.

Here is the catch: The Earth isn't a perfect sphere; it's a lumpy, irregular ellipsoid. Because of this, the mathematical "ellipsoid sea level" used by GPS satellites rarely matches actual physical ground level.

In many parts of the world, this mathematical model sits 15, 30, or even 50+ meters below or above the actual dirt you are standing on.

Doing the Metadata Math

If your specific takeoff location happens to sit 16 meters above the invisible WGS84 mathematical line, the drone will automatically record 16m into the photo properties while resting on the ground.

When you fly into the air, the photo metadata simply adds your flight height to that invisible baseline:

Base Ellipsoid Height (16m)} + Actual Flight Altitude (90m)} = Photo Metadata (106m)

Note on drifting numbers: If your metadata reads 120m instead of the mathematically expected 106m, don't worry. Consumer GPS modules have a notorious vertical accuracy variance (often fluctuating by plus or minus10 to 15 meters), which regularly blends with barometric drift during a climb.

Is This True for All Drones?

Yes. Whether you fly a DJI Mini, an Autel enterprise rig, or a Skydio, they all do this.

International camera standards dictate that the primary EXIF:GPSAltitude tag must reflect a global coordinate system. If a drone manufacturer changed this tag to show "height above your specific takeoff spot," the photos would become incompatible with global software like Adobe Lightroom, Apple Photos, and Google Maps.

The Good News: The Real Data is Hidden Inside

If you are a mapping pilot or a photographer worried about accuracy, don't panic. Drone manufacturers actually record both numbers; standard photo viewers just don't show you the hidden ones.

If you open your image in an advanced metadata viewer or photogrammetry software, you will find custom manufacturer tags:

XMP:RelativeAltitude: This holds the exact, real-world height relative to your takeoff point (matching your controller).
XMP:AbsoluteAltitude: This tracks height above true Mean Sea Level (AMSL) or the ellipsoid, depending on the drone model.

Mapping software like DroneDeploy, Pix4D, or ArcGIS natively ignores the confusing standard GPS tag and reads the RelativeAltitude tag instead, ensuring your 3D models remain perfectly accurate.

How to Fix It

If the mismatched numbers bother you or mess up your workflow, you can easily overwrite the confusing global data.

Using a free, industry-standard command-line tool called ExifTool, you can batch-command your computer to copy the hidden relative altitude over the standard GPS altitude tag for an entire folder of photos instantly. Because drone brands house this data in slightly different custom directories, targeting the broader wildcard tag is the most reliable method:

Bash

exiftool "-GPSAltitude<*RelativeAltitude" -GPSAltitudeRef=0 /path/to/your/photo/folder

(Note: -GPSAltitudeRef=0 ensures the software knows the new altitude is "Above Sea Level" so it doesn't accidentally read it as a negative/below-sea-level value).

Summary

The next time your photo properties tell you that you were flying at a terrifyingly illegal altitude, take a deep breath. Your drone is operating perfectly. It’s just speaking the complex mathematical language of global satellite networks, while your controller is telling you what you actually need to know to fly safely.

References

EXIF Standards: JEITA CP-3451D / Exchangeable image file format for digital still cameras: Exif Version 2.31. Established by Japan Electronics and Information Technology Industries Association (JEITA).
WGS84 Coordinate System: National Imagery and Mapping Agency (NIMA) Technical Report TR8350.2, "Department of Defense World Geodetic System 1984: Its Definition and Relationships with Local Geodetic Systems."
ExifTool Documentation: Harvey, Phil. ExifTool by Phil Harvey (v12+). Available at: [https://exiftool.org/](https://exiftool.org/)
Drone Mapping Workflows: Pix4D Knowledge Base. "Vertical Accuracy in Drone Mapping and Geoid Models." / DroneDeploy Support. "Understanding Drone Imagery Metadata and GPS Elevations."