Two ways to time something absurdly fast
Light crosses a meter in about three nanoseconds, and a Time of Flight sensor has to resolve differences far smaller than that. The industry solved the problem twice. Direct ToF (dToF) actually clocks the pulse: fire, start timer, catch the return, stop timer. Indirect ToF (iToF) sidesteps the stopwatch entirely: it sends light whose intensity oscillates continuously and measures the phase shift between what went out and what came back, which encodes the distance. Both produce depth; they fail and excel differently.
What is the difference between dToF and iToF?
dToF measures the round-trip time of discrete light pulses directly, usually with single-photon avalanche diode (SPAD) detectors sensitive enough to register individual returning photons. iToF measures the phase offset of a continuously modulated signal on a conventional sensor array. dToF handles long range and bright ambient light better and its precision holds with distance; iToF historically delivered finer lateral resolution at short range and a simpler sensor stack, at the cost of range ambiguity and multipath sensitivity.
Where each design shows its character
Range behavior. iToF's phase measurement repeats: beyond a certain distance the phase wraps around and a far target can masquerade as a near one (the aliasing problem, managed with multiple modulation frequencies). dToF has no wrap-around; a longer flight is simply a longer time.
Sunlight. dToF's photon-counting detectors and pulsed operation let it pick its own signal out of ambient infrared more effectively, one reason it dominates outdoor-capable designs. The broader sunlight problem is covered in ToF in sunlight.
Multipath. Light that bounces off a wall before returning corrupts iToF phase measurements more than dToF pulse timing. Corners and shiny floors are where iToF depth maps warp.
Resolution and cost. iToF rode standard image-sensor manufacturing to dense, cheap depth arrays, which made it the default in early consumer depth cameras. SPAD arrays for dToF were exotic once; consumer devices have since pulled them into volume production, and the gap keeps narrowing.
What a buyer should actually do with this
For most purchasers of finished systems, the dToF/iToF choice is already baked into the product, and the right move is to test outcomes rather than debate architectures: verify accuracy in your real geometry and lighting using a ground-truth method like the accuracy test methodology. Where the split matters is edge cases: strong sun through glass frontage, highly reflective interiors, unusually high mounting. If your site has one of those, ask the vendor which family the sensor uses and how the design mitigates that specific failure mode. A vendor who cannot answer is answering.
For the fundamentals both designs share, start at the pillar explainer. In people counting geometries (an overhead sensor a few meters above a doorway), both families deliver accuracy far beyond what the counting task needs, which is exactly why the technology took the category over.
---
