It's a practical question with an awkward answer: the number your planner shows most prominently — the waypoint count — often has almost nothing to do with it.
You need the photo count for three reasons. It sets how much card space to bring, how long processing will take, and whether the camera can physically keep up with the flight. That last one is the interesting failure, and it's the one that quietly damages datasets.
Waypoints are not photos
How the two relate depends entirely on how the camera is triggered.
Photo at each waypoint. The aircraft stops (or passes) each waypoint and fires the actions attached to it. Here photo count really is waypoint count multiplied by the number of photo actions per waypoint — and if you never added a photo action, it is zero, however elaborate the route looks.
Interval or distance-triggered shooting. This is what most mapping missions use. Waypoints mark only the ends of each flight line, and between them the camera fires automatically every few metres. A grid with 100 waypoints can easily produce a few thousand photos. The waypoint count tells you essentially nothing.
A test survey of ours reported 100 waypoints and 2,548 photos — roughly 27 GB. Planning card space from the waypoint number would have under-provisioned by a factor of twenty-five.
Working out the real number
For a distance-triggered grid, the photo count is total flown line length divided by photo spacing. Both of those come out of your overlap settings and camera footprint:
Worth being careful about which footprint dimension goes where. At nadir, flying nose-along each line in the usual landscape orientation, the long edge of the frame lies across the flight direction. So side overlap (between lines) works on the long edge, and front overlap (between consecutive shots) works on the short edge. Getting these the wrong way round produces entirely plausible-looking numbers that are wrong in both directions at once — tighter lines than you need, and less front overlap than you asked for.
A worked example. A Mavic 4 Pro at 120 m has roughly a 154 m across-track by 103 m along-track footprint. At 75% front and 65% side overlap:
Over a 40 ha square block (about 632 m a side), that's 12 lines of 25 photos — around 290 photos, before you add the margin you should be flying beyond the boundary. At roughly 11 MB per JPEG that's about 3 GB, or around 9 GB if you're shooting RAW.
Sizing card space
File size tracks the sensor and the output resolution together, not the megapixel count alone. A 20 MP frame off a Micro Four Thirds sensor is a substantially heavier file than a 48 MP frame off a phone-sized sensor, because the pixels themselves carry more information. Estimating from megapixels alone will mislead you in both directions.
| Factor | Effect on card space |
|---|---|
| RAW instead of JPEG | Budget roughly 3x |
| Full resolution vs binned | Larger files and a slower sustained capture rate |
| Boundary margin | Extra lines outside the area of interest |
| Crossing grid for 3D | Roughly doubles the mission |
Then bring more card than the estimate. Running out mid-mission costs you the flight; carrying a spare card costs nothing.
The trigger-rate trap
This is the part that catches people, because nothing warns you at the time.
Photos are triggered by distance, not by time. So flying faster does not give you more photos — the spacing is fixed by your overlap. What flying faster does is shorten the time between exposures:
A frame every 1.6 seconds is faster than many DJI cameras sustain, particularly at full resolution and especially writing RAW. When the camera cannot keep up, the aircraft does not slow down and it does not report a problem. It simply misses the shots it could not take and carries on.
The result is a dataset with lower real front overlap than you planned, distributed unevenly along the lines, with nothing in the flight log or the planner to indicate it. You find out during processing — or worse, you don't find out, and the model is subtly poorer than it should be. If a survey has been unexpectedly hard to reconstruct, compare the photo count you actually got against the count you planned. A shortfall is the tell.
For most survey settings, 3–4 m/s is a safe range. If you need to fly faster, widen the photo spacing (accepting less front overlap), or drop to a binned capture mode the camera can sustain. Check the trigger interval your plan implies before you launch — it is one division, and it is the difference between a clean dataset and a puzzling one.
Binned or full resolution?
Many recent DJI cameras offer a high-megapixel mode alongside a standard binned output. For mapping, the binned mode is usually the better acquisition choice: smaller files, a capture rate the camera can actually sustain, and pixels that are cleaner per unit area. The high-resolution mode is worth reaching for when you specifically need the detail and can slow the flight down to earn it.
Whichever you choose, make sure your planner is modelling that mode — resolution changes GSD, file size, and sustainable capture rate together, not just the first of those.
DronePoint shows the estimated photo count and card space alongside the waypoint count in the preview bar, and warns when your speed and spacing imply a trigger interval the camera is unlikely to sustain.
Related: front and side overlap sets the spacing all of this derives from, and ground sample distance sets your resolution ceiling.
Follow DronePoint
See more of these notes in your Google Search results.
Add DronePoint as a preferred source