Every mapping quote you write rests on a number most clients never ask about: the ground sampling distance. It decides how much detail the finished map can show, how high you fly, how many photos you take and how long processing runs. Too coarse and the client can't read what they paid for; too fine and you burn batteries and processing hours on detail nobody needed. This guide walks through the ground sampling distance drone pilots work with every day: what GSD is, the formula, a worked example, and how to pick a GSD that fits the deliverable.
What ground sampling distance actually measures
Pix4D's documentation on ground sampling distance in photogrammetry (checked Oct 28, 2026) defines GSD as "the distance between two consecutive pixel centers measured on the ground." The same page notes that "the bigger the value of the image GSD, the lower the spatial resolution of the image and the less visible details." Its example: a GSD of 5 cm means one pixel represents 5 cm on the ground, so each pixel covers 25 square centimeters.
In plain terms, GSD is the size of one pixel on the ground; detail smaller than that gets averaged away. If you need a refresher on how individual photos become one stitched, scaled image, our explainer on what an orthomosaic is covers it.
GSD is resolution, not accuracy
This is the point clients most often get backwards, so it's worth saying plainly: GSD tells you how fine the detail is, not how correct the positions are. A map with a 1.5 cm GSD can still sit well away from its true location on the earth if the job had no ground control, no RTK or PPK corrections, or a weak processing result. A coarser map built on good control can land closer to the true coordinates.
You can't reliably measure something smaller than a pixel, so the two are related, but a small GSD is not an accuracy promise. When a client asks "how accurate is it?", answer from how the job was controlled and checked.
The GSD formula
Pix4D's article on computing the flight height for a given GSD (checked Oct 28, 2026) gives the relationship using four inputs:
- Sw: real sensor width, in millimeters
- FR: real focal length, in millimeters
- imW: image width, in pixels
- H: flight height, in meters
Pix4D's equation (4) is written to solve for height: H [m] = (imW × GSD × FR) / (Sw × 100), with GSD in centimeters per pixel. Rearranged to solve for GSD, the same equation reads:
GSD [cm/pixel] = (Sw × H × 100) / (FR × imW)
The same Pix4D page also gives the footprint width of one image on the ground: Dw [m] = (imW × GSD) / 100. That footprint is what your flight planner uses, together with your overlap settings, to space the flight lines. For how much overlap to plan and why, see our guide to drone mapping overlap settings; this post sticks to the resolution side.
Use the real focal length, not the 35 mm equivalent
Camera spec sheets often list a 35 mm equivalent focal length. Pix4D is explicit that "it is not the 35 mm equivalent but the real focal length that should be used." For a 4:3 sensor, its conversion is FR = (F35 × Sw) / 34.6. Plugging the 35 mm equivalent number straight into the formula is one of the most common reasons a planned GSD doesn't match what the processing report shows.
A worked example
Pix4D's own example. The vendor page works the problem in the height direction: a 5 cm/pixel target, a real focal length of 5 mm, a real sensor width of 6.17 mm and an image width of 4,000 pixels. The result is H = (4000 × 5 × 5) / (6.17 × 100) = 162.07 m. Run it backward through the GSD form and you get the same answer: (6.17 × 162.07 × 100) / (5 × 4000) ≈ 5.0 cm/pixel.
A second example, with illustrative numbers. The following camera values are example numbers used to show the math. Look up your own camera's real sensor width, focal length and image width from its manufacturer before planning a job.
- Example sensor width: 13.2 mm
- Example real focal length: 8.8 mm
- Example image width: 5,472 pixels
At 120 m above the ground (about 394 feet):
GSD = (13.2 × 120 × 100) / (8.8 × 5472) = 158,400 / 48,153.6 ≈ 3.3 cm/pixel (roughly 1.3 inches per pixel). Each image covers about 5472 × 3.29 / 100 ≈ 180 m across.
At 60 m (about 197 feet), with the same example camera:
GSD ≈ 1.6 cm/pixel (roughly 0.65 inches), and each image covers about 90 m across.
Two things fall out of that math. GSD scales directly with height: halve the altitude and you halve the GSD. And the footprint shrinks by the same factor, so covering the same area at the same overlap takes roughly four times the images, with matching jumps in flight time, storage and processing.
Part 107 generally caps altitude at 400 feet above ground level (14 CFR 107.51(b), checked Oct 28, 2026), which limits how coarse a GSD a given camera can reach.
What changes GSD in the field
- Height above the ground, not above takeoff. The H in the formula is distance to the surface being mapped. On a site with a ridge or a deep pit, a flight at constant altitude above the launch point produces finer GSD over the high ground and coarser GSD over the low ground. Pix4D notes that "even when flying at a constant height, the images of a project may not have the same GSD" because of terrain differences and camera angle, and that an average GSD is computed for the orthomosaic. Terrain-following flight helps keep it consistent.
- Focal length. At the same height, a longer real focal length gives a smaller GSD and a smaller footprint, which means a faster image rate to hold overlap.
How to pick a GSD for the deliverable
Pix4D's guidance is to decide on the GSD "before starting the image acquisition," flying lower when detail matters and higher when large areas don't need it. Turning that into a client conversation works best when you start from the deliverable, not the drone:
- Ask what the smallest thing is that they need to see or measure. A crack, a curb edge, a stake, a stockpile toe, a row of seedlings. A feature that spans only one or two pixels is hard to identify reliably, so plan a GSD where that feature covers several pixels.
- Ask how the map will be used. A progress overview and a detailed surface inspection are different jobs, even on the same property.
- Check the contract or spec. Some clients, especially engineering and public-sector clients, specify a GSD or pixel size. If they do, plan to it and report the GSD from the processing results, not just the planned value.
- Price the trade-off honestly. Finer GSD means more photos, flight time and processing. If a coarser GSD meets the need, say so.
- Write the planned GSD on the job. Record the target GSD, flight height and camera settings with the mission so you can compare plan to result and repeat the job the same way next time.
GSD and your processing budget
Image count is where GSD shows up on your hardware. Halving the GSD can roughly quadruple the photo count for the same area, and processing time and disk use climb with it. If you process locally, that's a real planning factor: our guide to processing drone maps on your own hardware covers what drives those times and how to size a workstation for the jobs you actually fly.
Pilot Ledger's SkyView Desktop turns RGB drone photos into orthomosaics on your own Windows computer, with no per-map processing fee, and delivers the results to your clients through Pilot Ledger alongside your quotes, invoices and client records.
Quick reference
- GSD is the size of one pixel on the ground. Smaller is more detail.
- GSD [cm/pixel] = (Sw × H × 100) / (FR × imW), from Pix4D's flight-height equation.
- GSD is resolution, not accuracy. Accuracy comes from control, corrections and checks.
- Choose GSD from the smallest feature the client needs, then plan the flight around it.
Ready to keep quotes, flights and client delivery in one place? See Pilot Ledger plans and start a 14-day trial.