The sensor datasheet says 1.5 MHz pulse repetition rate. The quote says 'high-density capture with our 1.5 MHz sensor'. The deliverable lands at 22 ppm² — perfectly fine for the spec, materially below what the PRR number suggested. The arithmetic that links the published sensor throughput to what actually lands on the ground passes through altitude, speed, swath width and overlap, and at some altitudes the conversion is brutal.
If you've ever compared two LiDAR quotes that both claim "high-density capture" — one with a 550 kHz sensor flown at 100 m AGL, one with a 1.5 MHz sensor flown at 150 m AGL — and assumed the 1.5 MHz quote will produce more points per square metre, you've met the PRR-vs-density confusion. The sensor's pulse repetition rate sets the maximum throughput at the laser; what arrives at the ground depends on how the sensor is flown.
Higher PRR genuinely matters. So does the flight parameter envelope — altitude, speed, scan angle, overlap. The two interact in ways that mean a 1.5 MHz sensor at 150 m can produce lower on-the-ground density than a 550 kHz sensor at 80 m. Without doing the arithmetic, the PRR comparison is just sticker shock.
This article is the arithmetic. The chain from published PRR to delivered density, the gotchas that trip buyers up, why some operators quote PRR as a proxy for density without doing the math, and the brief language that pins density down independently of sensor spec.
Five parameters convert PRR into delivered ppm²:
1. Pulse repetition rate (PRR). Sensor-published. Pulses per second emitted by the laser. Range: typically 100 kHz for entry-level survey LiDAR up to 2 MHz+ for high-end sensors.
2. Returns per pulse. Each pulse can produce multiple returns (up to 7-15 depending on sensor), because the laser pulse can partially reflect off intermediate surfaces (canopy, wires, ground beneath). For dense canopy, average 2.5-4 returns per pulse; for bare ground, around 1 return.
3. Field of view (FOV). The angular swath the sensor sweeps. Typical drone LiDAR FOVs: 70-360 degrees. Wider FOV at higher altitude covers more ground per pass but reduces per-point density.
4. Flight altitude above ground level (AGL). Determines swath width on the ground. Doubling altitude doubles swath width, halving density at a given PRR.
5. Platform speed. Determines how quickly the sensor moves across the ground. Doubling speed halves density at a given PRR.
6. Strip overlap. Adjacent flight lines overlap; overlap area receives returns from both strips. Typical overlap 30-50%; doubles density in the overlap zone.
The arithmetic that links them:
density (ppm²) = (PRR × returns_per_pulse × overlap_factor) / (swath_width × ground_speed)
Where:
Worked example for a typical project:
Density = (1,500,000 × 2.5 × 1.5) / (112 × 8) ≈ 6,300 points per second per square metre = ridiculous.
Wait, the units. Let me redo: density is points per square metre across the whole flight line. The correct expression:
density (ppm²) = (PRR × returns_per_pulse × overlap_factor) / (ground_area_swept_per_second)
Where ground area swept per second = swath_width × ground_speed.
So: (1,500,000 × 2.5 × 1.5) / (112 × 8) = 5,625,000 / 896 ≈ 6,278 ppm². Still implausible — real deliverables come in at 20-200 ppm², not 6000.
The issue: the formula above counts every pulse but doesn't account for the per-pulse footprint overlapping with adjacent pulses. The actual delivered density depends on the pulse spacing on the ground — adjacent pulses don't independently contribute to density if they hit the same point.
The realistic formula uses pulse spacing along-track and across-track:
along-track spacing = ground_speed / pulse_rate_per_line
across-track spacing = scan_angle / pulses_per_scan_line
For most drone LiDAR sensors, the published density arithmetic uses an empirical or sensor-vendor- specified ground-density formula. Conservative working approximation for engineering-grade drone LiDAR:
Delivered density (ppm²) ≈ (PRR/1000) / (altitude × speed × 0.02)
For the example: (1500 / 1000) / (80 × 8 × 0.02) = 1.5 / 12.8 ≈ 117 ppm² before overlap, 175 ppm² with 50% overlap.
The formula above is approximate — real density depends on sensor-specific scan patterns — but it's close enough to show that altitude and speed are roughly as important as PRR in determining deliverable density.
Concrete example. Same sensor, same speed, same overlap, different altitudes:
| Altitude AGL | Approximate density (ppm²) at 1.5 MHz PRR, 8 m/s, 50% overlap | |---|---| | 60 m | ~230 | | 80 m | ~175 | | 100 m | ~140 | | 120 m | ~115 | | 150 m | ~95 | | 200 m | ~70 |
The same sensor delivers 3× density at 60 m AGL vs 200 m AGL. The PRR is identical; the ground footprint is utterly different.
(See point density article for what density numbers actually mean for different deliverable types.)
Same altitude, same speed, same overlap, two sensors with different PRRs:
| PRR | Approximate density (ppm²) at 80 m AGL, 8 m/s, 50% overlap | |---|---| | 100 kHz | ~12 | | 250 kHz | ~30 | | 550 kHz | ~65 | | 1.0 MHz | ~115 | | 1.5 MHz | ~175 | | 2.0 MHz | ~230 |
PRR matters directly. Doubling PRR roughly doubles density. The point: at the same altitude and speed, higher PRR genuinely delivers higher density. The confusion only arises when altitude or speed varies between the comparison.
Quote A: "1.5 MHz sensor, high-density capture" Quote B: "550 kHz sensor, engineering-grade capture"
Without altitude and speed parameters, you cannot compare these. Quote A at 150 m AGL and 12 m/s produces about 75 ppm². Quote B at 60 m AGL and 6 m/s produces about 130 ppm². The "lower PRR" quote is denser on the ground.
Fix: ask for the delivered density (ppm²) target, not the sensor PRR. The two operators should be able to commit to a number.
Quote: "Average density 100 ppm²".
Sounds clean. The "average" hides the distribution. A capture averaging 100 ppm² might be 150 ppm² in the interior and 50 ppm² at the edges, or 200 ppm² on bare ground and 30 ppm² under canopy.
Fix: specify density per category that matters — project-area average, interior (excluding 50 m edge buffer), under-canopy (if vegetation present), overall minimum. The QA pack should report all four.
(See edge effects article for why perimeter and interior density differ structurally.)
Three reasons PRR shows up on quotes instead of density:
1. PRR is sensor-spec; density is operator-spec. PRR is published in the sensor datasheet. Density depends on how the operator flies. Some operators quote PRR because it's a fact they can verify; they may not have committed to a specific density target.
2. Density depends on conditions. Under-canopy ground density depends on canopy density, which varies across the project. Operators sometimes quote PRR because committing to a density number requires accepting variability they can't fully control.
3. Marketing simplicity. "1.5 MHz" is a single impressive number. "150 ppm² minimum in interior, 50 ppm² minimum at edges, 30 ppm² ground returns under closed canopy" is harder to put on a marketing page.
None of these is dishonest — but the PRR number doesn't substitute for the density commitment for buyer purposes. Worth requesting density explicitly.
A rough mapping from density to typical deliverable quality:
| Density | Suitable for | |---|---| | Under 10 ppm² | Broad reconnaissance, planning-grade DTM at 1 m+ resolution | | 10-25 ppm² | Standard planning DTM, 0.5 m contours on bare ground, broad asset identification | | 25-50 ppm² | Engineering-grade DTM at 0.5 m, 0.25 m contours on bare ground, vegetation classification | | 50-100 ppm² | Engineering DTM under canopy (with adequate ground returns), detailed asset features, infrastructure-corridor work | | 100-200 ppm² | High-detail capture for asset condition, complex infrastructure, marginal-canopy ground recovery | | 200+ ppm² | Specialist applications: dense canopy ground recovery, fine-feature asset extraction, change-detection at small magnitudes |
The density specified should match the deliverable. Over-specifying drives cost without value; under-specifying produces deliverables that don't support the design intent.
Three sentences in the brief eliminate the PRR-vs-density confusion:
"Target delivered point density: minimum NN ppm² across project area, minimum MM ppm² ground returns under canopy. Density to be reported in QA pack stratified by cover class and interior- vs-edge zone. PRR specification at operator's discretion provided density targets are met."
That language makes density the contractual commitment and leaves sensor selection to the operator (who knows their own kit). It also explicitly requires the QA pack to report density in the dimensions that matter.
A common confusion: density and accuracy are separate properties. Five extra points per square metre doesn't make any individual point more accurate. Density supports:
Accuracy is a separate quantity governed by sensor range precision, GNSS solution quality, IMU performance, and calibration. A dense capture can still be inaccurate; an accurate capture can still be sparse. Specify both independently.
(See accuracy article for the accuracy story.)
One subtlety worth naming: some sensors publish a "peak" PRR that's higher than their "effective" PRR.
Peak PRR is what the laser can theoretically fire. Effective PRR is what the sensor can process given memory write speed, GNSS solution update rate, and timing constraints. For some sensors these are close; for others the effective is 60-80% of peak.
When comparing sensors, ask for effective rather than peak PRR. The difference is real for density calculations.
Sensor PRR (kHz / MHz) is the maximum pulse rate at the laser. Deliverable density (ppm²) is what lands on the ground after altitude, speed, swath width and overlap have done their work.
The same sensor at 60 m AGL produces approximately 3× the density of the same sensor at 200 m AGL. The same altitude with 2× PRR roughly doubles density. Two sensors with different PRRs flown at different altitudes are not directly comparable from PRR alone.
Two common quote-comparison traps: PRR without flight parameters (incomparable), density without stratification (averages hide distributions).
Three reasons operators quote PRR instead of density: it's a verifiable sensor spec, density depends on conditions, marketing simplicity.
Brief language: specify delivered density target (stratified by cover class and edge-vs-interior), require QA pack reporting, leave PRR to operator discretion.
Density and accuracy are independent. Specify both. Watch out for peak-vs-effective PRR distinction when comparing sensors directly.
If you're scoping a project and want to specify density targets that match the deliverable rather than the sensor available, send through the deliverable specifications. We'll quote against density targets that align with your design intent and propose the sensor and flight parameters that meet them — not the other way round.
The deliverable-by-deliverable density spec — useful pairing for translating the density numbers in this article into commitments that match what the project actually needs.
The sensor-class breakdown that explains where the PRR numbers in this article actually come from — different sensor tiers, different PRR capabilities.