LiDAR vs TLS vs Photogrammetry

If you’ve searched “LiDAR survey” for your building project and gotten back results ranging from drone mapping to handheld scanners to Revit deliverables, you’re not confused - the industry genuinely conflates three distinct technologies. Getting the wrong one doesn’t just hurt accuracy; it produces deliverables you cannot use downstream and blows budget recovering from the mismatch. This guide is written from the field, not a spec sheet.

---

Why the Terminology Is Genuinely Confusing (And Why It Matters)

The root of the confusion: LiDAR is a measurement physics, not a product category. Any system that fires pulsed or modulated laser light and times the return qualifies. That covers a scanner bolted to a tripod inside a hospital corridor, a sensor pod on a helicopter mapping a pipeline right-of-way, a backpack unit navigating a warehouse, and a $300 accessory clipped to an iPhone. All four are “LiDAR.” None of them are interchangeable.

The terminology is further muddied by vendors who use “LiDAR scan” as a marketing shorthand for any 3D capture service, and by clients who see “1-inch accuracy” claims from consumer-grade mobile apps and assume the same number applies to professional instruments.

Two specific distinctions that matter on every project:

TLS (Terrestrial Laser Scanning) is a stationary, tripod-mounted scanner. The instrument does not move during a scan; the laser sweeps a hemisphere and builds a dense point cloud from a fixed origin. This is what we use on every building interior job, because it is the only platform class that consistently achieves 1-3mm field accuracy in confined built environments.

Photogrammetry uses an entirely different physics. There is no laser. Structure-from-Motion (SfM) algorithms triangulate 3D coordinates from overlapping 2D photographs. It can achieve 2-10mm accuracy under favorable conditions - but it fails silently on the exact surfaces most common in building interiors: painted drywall, glass curtainwall, polished concrete, and dark metal. A drone with a mirrorless camera can produce a point cloud that looks geometrically complete but contains 20-50mm errors in smooth areas the SfM algorithm could not resolve, and those errors won’t surface until a modeler tries to snap walls in Revit.

Why the platform mismatch is a real project risk: A building owner searching “LiDAR survey” for a 15,000 sq ft office renovation may receive a quote for UAV LiDAR - which delivers 3-5cm absolute accuracy and cannot see inside the building at all. For interior MEP clash detection, the required accuracy threshold is 2-3mm. UAV LiDAR at 30-50mm is roughly one order of magnitude short of that threshold. Specifying the wrong technology doesn’t produce a slightly worse deliverable; it produces a deliverable that is disqualified for the end use.

---

How TLS (Time-of-Flight Variants) Actually Works

TLS instruments in the current market use time-of-flight measurement principles - either pulse-based or high-speed digital variants - and the specific implementation affects scan time, range, and accuracy on any project.

Time-of-Flight (ToF) fires a discrete laser pulse and measures elapsed round-trip time. Distance = (travel time × speed of light) / 2. Because the measurement is a direct time interval, ToF handles long ranges well. The Leica RTC360 is a well-known ToF instrument in the industry and reaches 130m with a published accuracy of 1.9mm at 10m, degrading to 5.3mm at 40m as range noise increases.

High-speed digital time-of-flight, as implemented in the Trimble X7, achieves point acquisition speeds that enable efficient interior scanning at the same time-of-flight physics. This enables point acquisition speeds up to millions of points per second. The tradeoff common to all ToF-based TLS at short-to-medium range: scan density and acquisition rate must be balanced against station count. At short-to-medium range - the 5-40m band that covers most commercial interiors - the X7’s auto-registration and auto-leveling features keep total station time competitive.

Target-Based vs. Target-Free Registration

This is a field decision that affects both registration accuracy and setup time, and it never comes up in marketing materials.

Target-based registration uses retroreflective targets - HDS flat targets (black-and-white coded discs) or spherical reflectors - placed in the overlap zone between adjacent scan positions. The scanner actively locates and ranges to the center of each target; those center coordinates are used by registration software (Trimble Business Center, Leica Cyclone 360, FARO Scene) to solve for the rigid-body transformation between scans. Achievable registration accuracy: 1-2mm RMS with good target geometry (targets spread at varying distances and elevations, not all co-planar).

Target-free (cloud-to-cloud) registration uses the SfM-like overlap of raw geometry between adjacent scans. Software identifies matching geometric features - wall corners, column edges, beam flanges - and solves the transformation without any physical targets. The Trimble X7’s onboard auto-registration preview uses this method for field QC. Accuracy under favorable conditions (rich geometry, high overlap): 2-4mm RMS. In environments with low geometric complexity - long straight corridors, open floor plates, parking garages with identical bays - cloud-to-cloud registration degrades significantly, often to 6-10mm RMS or worse. For those environments, we always place targets regardless of what the scanner’s onboard system reports.

Our rule on the project: We use target-based registration on all LOD 300+ BIM projects and any project with contractual accuracy requirements. Target-free registration is acceptable for rapid survey work where LOD 200 is the deliverable and 3-5mm is contractually acceptable.

Our Scanner in Detail

Trimble X7 - Our primary terrestrial scanner. High-speed digital time-of-flight, 80m range, 2mm range accuracy, 4mm 3D point accuracy at 10m, IP55-rated. The onboard dual-axis compensator auto-levels to within 5 arc-minutes, which eliminates the manual tripod-leveling step and removes approximately 0.5-1.5mm of setup-induced tilt error per station. On a typical 10,000 sq ft commercial floor, we run 4-6 scan setups per floor in open areas, 8-12 setups when MEP is dense.

Scan density and capture time matter operationally. The X7 at quarter-density (approximately 6mm point spacing at 10m) completes a scan in roughly 2 minutes per station. Full density (approximately 1.5mm spacing at 10m) takes approximately 8 minutes per station. On a 40-station project, that’s a 4-hour difference in field time - a decision we make based on deliverable spec, not habit. LOD 300 BIM typically requires no less than half-density; LOD 350 with MEP coordination gets full density.

For detailed or close-range work - components, reverse engineering, tight mechanical assemblies - we deploy handheld scanners including the Creaform MetraSCAN, which operate at sub-millimeter accuracy in a completely different tier from the terrestrial scanner.

Registration workflow: After field capture, we stitch scans in Trimble Business Center or compatible registration software. The registered cloud exports as .RCP for Revit workflows, .E57 for cross-platform exchange, or .LAS for survey/GIS hand-off. Our maximum acceptable registration residual is 3mm RMS across any project. If a scan pair exceeds that threshold, we return to the field to add a setup - we do not deliver from a point cloud with known gap errors.

What TLS cannot do: It cannot see through vegetation canopy. It cannot see around a corner without re-positioning the tripod. And it is not efficient for horizontal extents beyond a single building - scanning a 50-acre campus with a tripod scanner becomes uneconomical past roughly 5 acres of open ground. That is where UAV LiDAR steps in.

---

Airborne and Mobile LiDAR: When Terrestrial Is the Wrong Tool

Airborne LiDAR mounts a sensor on a fixed-wing aircraft or helicopter alongside a high-grade IMU and GNSS. Vertical accuracy: 5-15cm; horizontal: 10-30cm. That is entirely sufficient for topographic surveys, floodplain mapping, and utility corridor documentation - and completely inadequate for as-built interior work or MEP clash detection. We do not offer airborne LiDAR capture directly, but we incorporate airborne datasets as site context for large civil projects.

UAV LiDAR brings the sensor much closer to the surface. Current offerings in the high-accuracy UAV LiDAR market include the DJI Zenmuse L2 (integrated with the Matrice 350 RTK; official DJI specs state 5cm horizontal / 4cm vertical accuracy at 150m with RTK fix and post-processing, with a recommended operating altitude of 30-150m AGL), the Hesai AT128 (a 128-channel hybrid solid-state lidar gaining traction on autonomous vehicle and survey platforms), and the YellowScan Mapper Ultra (a dedicated survey payload with 2.5cm accuracy at nadir from 50m height, with a recommended maximum flight height of 240m AGL). Accuracy improves to 1-2cm with ground control points. For roofs and building facades on sites larger than 2 acres, UAV LiDAR is typically the right tool - it delivers 1-2cm accuracy with GCPs at a cost-per-acre that terrestrial re-setups cannot match. For drone as-built surveys for large commercial sites, UAV LiDAR often delivers the best cost-per-acre.

Mobile LiDAR (NavVis VLX, Leica BLK ARC) mounts a scanner on a backpack or trolley and navigates indoors using SLAM (Simultaneous Localization and Mapping) or GNSS-aided positioning. Mobile LiDAR platforms can cover large interior areas per day - roughly 5-10× the daily coverage of a TLS crew. Typical interior accuracy: 5-20mm. That’s the tradeoff. For a project requiring LOD 350 BIM or fabrication-tolerance documentation, mobile LiDAR doesn’t clear the bar. For a rapid LOD 200 survey of a 500,000 sq ft distribution center, it’s the right tool.

The decision fork in direct terms: Use TLS when 3mm or better is required or the project is a single building. Use UAV LiDAR for large sites, roofs, and facades where 1-2cm with GCPs is acceptable. Use mobile LiDAR for rapid large-area coverage where 5-10mm is contractually acceptable and speed is the constraint.

---

Photogrammetry: Cameras, Overlap, and Where It Beats LiDAR

Photogrammetry processes dozens to thousands of overlapping photographs through Structure-from-Motion algorithms. Software like Agisoft Metashape, RealityCapture (now owned by Epic Games), or Bentley ContextCapture solves for camera positions by matching keypoints (SIFT, ORB, or proprietary variants) across image pairs, then triangulates 3D point positions from those matched features. The output is a dense point cloud and, typically, a textured mesh.

Processing time is a real project variable, not a footnote. On a mid-range workstation (RTX 4090, 128GB RAM), 500 drone images at full resolution process in RealityCapture in approximately 45-90 minutes for alignment and dense cloud generation - RealityCapture’s GPU-accelerated pipeline is genuinely fast. The same dataset in Agisoft Metashape at “High” quality takes approximately 3-5 hours on equivalent hardware, though Metashape’s alignment is often more robust on difficult datasets with repetitive geometry (think tiled facades or uniform grassland). For large aerial missions - 2,000+ images - budget 8-12 hours in Metashape or 3-5 hours in RealityCapture. These are real compute cycles on capable hardware; a laptop with integrated graphics is not a processing platform for photogrammetry.

Where photogrammetry wins: Cost and color. A DJI Mavic 3E with RTK runs approximately $5,000 body-only - a legitimate entry point for site mapping at 1-3cm accuracy. A PPK-equipped survey drone (DJI Matrice 350 RTK with P1 camera, or a WingtraOne) suitable for 5mm GCP work starts at $10,000-$25,000 for the system. For exterior UAV surveys on civil projects, photogrammetry dominates because the cost-to-accuracy ratio is hard to beat for terrain models and site context.

Accuracy with ground control points and calibrated cameras: 2-10mm on exterior surfaces. With careful GCP placement and high-overlap flight plans (80%+ front, 70%+ side overlap), facade surveys achieving 5mm accuracy are achievable. UAV photogrammetry for site surveys typically lands at 1-3cm absolute, down to 5mm with dense GCPs.

Where it fails - and this is the field-critical failure mode: Photogrammetry requires texture contrast to find matching keypoints between images. Painted drywall, glass curtainwall, polished concrete, and dark or shiny metal provide insufficient texture for the SfM algorithm. The software will produce a point cloud that looks geometrically complete - there are no obvious holes - but the coordinates in those featureless regions are interpolated or poorly constrained, not measured. Photogrammetry datasets of office interiors are known to show 20-40mm deviation from TLS ground truth on smooth partition walls, with no visual indicator in the dense cloud that the data is unreliable. That is a silent failure. LiDAR is indifferent to surface texture; a time-of-flight laser reflects off a white drywall partition just as precisely as off a brick facade.

Hybrid TLS + photogrammetry workflows: We use photogrammetry selectively for exterior facade texture overlays draped onto TLS-derived point cloud geometry. The process: capture TLS from the exterior for accurate geometry, fly a photogrammetric mission for color mesh, then register and drape the photogrammetric texture onto the TLS geometry in Leica Cyclone 3DR or Agisoft Metashape (which supports point cloud import and mesh draping natively). The typical accuracy degradation in the draped result depends on registration quality - with 4+ GCPs shared between both datasets, a 5-10mm deviation in the combined product is typical compared to 1-3mm for TLS-only geometry. That’s acceptable for historic preservation documentation but not for MEP coordination.

Critical downstream note: Photogrammetry outputs a mesh or orthophoto (.OBJ, .FBX, .TIFF). TLS outputs a raw point cloud (.E57, .RCP, .LAS) that feeds directly into Revit or AutoCAD. A photogrammetric mesh requires significant processing to produce a BIM-ready point cloud, and the result carries accuracy gaps in exactly the locations - smooth interior surfaces - where AEC coordination is most demanding.

---

Accuracy Comparison: A Straight Numbers Table

AEC takeaway: The 2-3mm threshold is where clash detection, LOD 300/350 BIM, and fabrication tolerances become achievable. Photogrammetry and UAV LiDAR are fully appropriate for site context and LOD 200 massing - but specifying either for interior as-built work requiring clash detection is a specification error with real cost consequences downstream.

---

Why Professional 3D Scanners Cost So Much (And What You’re Actually Paying For)

The Trimble X7 lists at approximately $75,000. The Leica RTC360 is around $90,000. A FARO Focus S350 runs approximately $60,000. A DJI Mavic 3E with RTK for photogrammetry: approximately $5,000 body-only, or $10,000-$25,000 for a survey-capable PPK system. That’s a 3-18× cost differential. Here’s what the scanner price actually covers.

Precision laser modules and wavelength stability: Time-of-flight measurement requires that the laser pulse timing and detection circuitry maintain consistent calibration across the full scan duration and across ambient temperature changes. Achieving that stability requires factory-calibrated optical components and thermal compensation circuitry that has no counterpart in consumer electronics. Field recalibration of the optical path is a factory procedure, not a firmware update.

Encoder angular resolution: The scanner head rotates continuously to build a hemispheric point cloud. The angular encoder must resolve position to ≤1 arc-second consistently across tens of thousands of rotations per scan. At 10m range, 1 arc-second of angular error translates to approximately 0.05mm of lateral position error - which is why the encoder spec is the floor on achievable point accuracy, not a cosmetic specification.

Dual-axis compensator: The Trimble X7 auto-levels to within 5 arc-minutes using its internal dual-axis compensator. That eliminates approximately 0.5-1.5mm of tilt-induced position error per station compared to manual tripod leveling, which compounds significantly across 30-50 setups on a large project.

Ruggedization - specifics: The Trimble X7 carries an IP55 rating, meaning the housing is protected against dust ingress sufficient to prevent interference with operation, and against water jets from any direction. In practice: a scanner sitting in a concrete slab pour area during decking operations is exposed to concrete dust, water mist from curing operations, and vibration from pneumatic tools. Scanning active mechanical rooms at 95% relative humidity, unconditioned parking structures at 18°F, and industrial processing facilities with airborne particulate is routine work for a professional scanner. The IP55 housing keeps the optical window, encoder, and electronics alive in those environments. A mirrorless camera body - typically IPX4 at best - is not the same class of protection.

Software stack: Trimble Business Center ($8,000-$15,000/year), Leica Cyclone 360 ($6,000-$12,000/year), FARO Scene ($5,000-$10,000/year), Autodesk ReCap Pro (~$600/year), Revit (~$3,300/year). These are not optional line items; they are the pipeline that converts raw scans into BIM-ready deliverables.

Field labor: A certified scanning technician - typically two on larger sites for safety and production - plus target placement, site coordination, and on-tablet QC review: 4-8 hours per 10,000 sq ft.

Cost benchmark: A TLS-based as-built for a 10,000 sq ft commercial floor typically runs $2,500-$6,000 all-in depending on complexity, occupancy constraints, and deliverable type. Photogrammetry-only exterior facade work runs $800-$2,500. For the full cost breakdown by project type, see what it actually costs to scan a building.

Consider the risk math: a single RFI caused by a missed or incorrect dimension on a $5M renovation can result in significant change-order cost. Projects without accurate as-builts commonly encounter avoidable change orders that a TLS as-built at $4,000 would have prevented - the cost of working from verified dimensions is modest relative to that exposure.

---

Laser Scanning vs Traditional Measuring Tools: When to Make the Switch

The traditional as-built toolkit: tape measure (±5mm, operator-dependent), laser distance meter (±1.5mm point-to-point, no geometry capture), total station (±2mm, excellent precision but slow on dense geometry), hand-sketched field notes.

Speed differential: An experienced drafter with a tape measure and laser distance meter captures roughly 500 sq ft per day to produce a usable floor plan. Our TLS crews cover 20,000-50,000 sq ft per day at higher accuracy. That is a 40-100× productivity gap on a per-square-foot basis.

Error propagation - the silent failure mode of manual measurement: A 5mm error introduced in room 1 does not stay in room 1. Every subsequent dimension taken from that wall carries the error forward. By the time you’ve traced three column lines and four room sequences, a single bad baseline measurement has infected a significant portion of the floor plate. TLS registers to an absolute control network, meaning each scan position is solved against that network independently - errors do not accumulate room to room.

Handheld structured-light scanners (Artec Leo: 0.1mm accuracy; Creaform MetraSCAN: 0.025mm) operate in a completely different tier - sub-millimeter accuracy, purpose-built for small manufactured parts and component reverse engineering. They are not building tools. Maximum effective scan volume per setup is roughly 0.5m³; they are not designed for open environments or for navigating through a doorway while scanning.

---

Choosing the Right Technology: A Decision Framework for Project Owners

Step 1 - Define required accuracy. Fabrication tolerances and coordination drawings: 1-3mm (TLS, target-based registration). Clash detection: 2-3mm (TLS). Site context and massing: 1-3cm (UAV LiDAR or photogrammetry). Do not over-spec, but never under-spec for the contractual deliverable - the cost to re-mobilize because a dataset is unusable exceeds the cost difference between technology tiers.

Step 2 - Define project extent. Single room to multi-floor building: TLS. Large campus or industrial site: mobile LiDAR or UAV. Roof or exterior facade only: UAV photogrammetry or UAV LiDAR.

Step 3 - Define the deliverable and its downstream use. BIM model in Revit (.RVT) for construction coordination: TLS point cloud → ReCap Pro → Revit scan-to-BIM workflow. If the Revit model will be used for MEP clash detection, the LOD requirement is almost certainly 300 or 350 - which means TLS with target-based registration, not mobile LiDAR and not photogrammetry. 2D CAD drawings for permit submission: TLS or photogrammetry depending on interior vs. exterior. Orthophoto or terrain model for civil: UAV photogrammetry. For the complete picture, see our full scan-to-BIM workflow from field to Revit.

Step 4 - Assess surface conditions. Glass curtainwall, polished concrete, shiny dark metal: TLS with intensity filtering (time-of-flight scanners can filter out returns below a reflectivity threshold to reduce noise from specular surfaces). Rough textured masonry, heavily detailed exterior ornament, or complex organic forms: photogrammetry is competitive on accuracy and significantly more economical per surface area.

Step 5 - Confirm regulatory and contractual requirements. ALTA/NSPS land surveys require a licensed land surveyor to certify the deliverable regardless of capture technology. Structural documentation submitted for permit may require a licensed professional engineer’s stamp on the deliverable drawings. BIM deliverables on public projects sometimes specify LOA (Level of Accuracy) per USIBD Guide, not just LOD - confirm which standard the specification references before scoping the capture. What we provide is accurate, timestamped existing-conditions documentation - point clouds and models that the client’s own design team and licensed professionals then use for their permitted and engineered work. If the spec says LOA 30 or higher (±3mm positional accuracy), that is a TLS requirement.

---

Point Cloud to Deliverable: How the Data Flows After Capture

The point cloud is an intermediate product - raw material, not the deliverable. What matters is what gets built from it, and how the client verifies that what got built is accurate. For a detailed look at how point cloud registration works, that post covers the stitching process in depth.

TLS workflow: Scan in field → register in Trimble Business Center or compatible registration software → generate registration QC report (residuals per scan pair, RMS across project, maximum single-pair error) → export .E57 or .RCP → index in Autodesk ReCap Pro → link into Revit → model to LOD/LOA spec → QC model against point cloud → deliver .RVT and .DWG. For specifics on importing a point cloud into Revit, we’ve documented the full indexing and linking workflow.

Photogrammetry workflow: Photos → SfM alignment in RealityCapture or Metashape (45-90 min for 500 images in RealityCapture on RTX 4090; 3-5 hours in Metashape on equivalent hardware) → dense cloud generation → mesh reconstruction → orthophoto or textured mesh export (.OBJ, .FBX, .TIFF). The path to BIM is significantly less direct than TLS. Feeding a photogrammetric dense cloud into Revit requires re-indexing through ReCap Pro, and the resulting snap-to-point-cloud experience is degraded compared to a clean TLS registration because photogrammetric clouds have higher noise on featureless surfaces. Budget additional modeling time: a 10,000 sq ft floor plate takes roughly 2-3 days to model from a clean TLS cloud; the same floor from a photogrammetric cloud on interior surfaces can take 4-5 days due to noise filtering and gap-filling.

BIM model QA/QC against point cloud: Delivering a Revit model is not the end of the quality workflow. After modeling, we run a cloud-to-model comparison using Autodesk ReCap Pro or Leica Cyclone 3DR: the modeled surfaces are compared against the registered point cloud, and deviations are output as a color deviation map (typically: green = ±3mm, yellow = ±6mm, red = >10mm). This deviation color map is included in every LOD 300+ deliverable we produce. Walls that look correct in a 3D view may carry 8-12mm systematic deviation if the scan-to-BIM operator was working from a sparse cloud section - the deviation map catches those before the client does. We specify ±6mm as the maximum permissible deviation for LOD 300 elements (walls, slabs, major structural) and ±3mm for LOD 350 MEP-coordinated elements.

File format guide:

LOD implications: LOD 200 (massing geometry, approximate dimensions) can be modeled from UAV photogrammetry on exterior elements. LOD 300 (precise geometry, accurate dimensions, correct location) requires TLS. LOD 350 (full coordination including connection details, clearances, equipment rough-ins) requires TLS at 1-3mm with target-based registration - there is no photogrammetry path to LOD 350 for interior building systems. For understanding laser scanning deliverable specifications, our specs post walks through LOD, LOA, and what each requires from the source point cloud.

---

What We Use and Why: Named Gear, Real Workflow

Our primary terrestrial scanner is the Trimble X7. On a typical commercial interior, we station it per 400-600 sq ft in complex spaces (mechanical rooms, restrooms, tight corridors) and per 800-1,000 sq ft in open floor plates. At half-density (~4 min per station), a two-person crew runs 60-80 setups per 8-hour field day, covering 40,000-60,000 sq ft of open space or 25,000-40,000 sq ft of complex MEP space. That productivity calculation is what determines the field day count in our project quotes.

The X7’s auto-leveling saves 2-3 minutes per station compared to manual tripod leveling. On a 50-station project, that is 100-150 minutes recovered - roughly half a field day at scale. Onboard auto-registration preview evaluates cloud-to-cloud overlap in the field before we move the tripod. If a station pair shows poor overlap or a high initial residual, we add a setup then - not after we’ve demobilized and discovered the gap in the office.

For detailed component work, reverse engineering, and close-range part documentation, we deploy handheld scanners including the Creaform MetraSCAN. These instruments operate at sub-millimeter accuracy and are purpose-built for manufactured parts and small assemblies rather than open building environments.

Our field protocol:

• Minimum 20% point cloud overlap between all adjacent scan positions
• HDS flat targets or spherical reflectors at every setup for redundant target-based registration
• Maximum registration residual: 3mm RMS across the full project; any scan pair exceeding that threshold triggers a field re-shoot before demobilization
• Every scan reviewed on-tablet before breaking down the setup
• Cloud-to-model deviation map included on all LOD 300+ deliverables, with ±6mm tolerance for LOD 300 elements and ±3mm for LOD 350

Where photogrammetry fits in this work: Exterior facade texture overlays draped onto TLS-derived geometry in Metashape or Cyclone 3DR. We do not substitute photogrammetry for TLS where 2-3mm accuracy is contractually required - the failure modes on smooth interior surfaces are too consistent and the downstream BIM impact too significant. Our professional 3D laser scanning services page covers how we scope each project type.

---

FAQ

Is LiDAR the same as 3D laser scanning?

LiDAR is the parent physics - any system using pulsed or modulated laser light to range distance. Terrestrial laser scanning (TLS) is one specific platform type within that category: a stationary, tripod-mounted instrument that sweeps a hemisphere and produces a dense point cloud from a fixed origin. The others - airborne LiDAR at 5-15cm accuracy, UAV LiDAR at 3-5cm, mobile SLAM LiDAR at 5-20mm - are also LiDAR but produce fundamentally different accuracy and coverage profiles. When a building owner says “LiDAR survey,” they almost always need TLS. Confirming that before scope is written saves a re-mobilization.

What measurement principle does the Trimble X7 use and how accurate is it?

The Trimble X7 uses high-speed digital time-of-flight measurement - it fires rapid laser pulses and times the return. Per the official Trimble datasheet, range accuracy is 2mm and 3D point accuracy is 4mm at 10m, with an 80m range and IP55 environmental rating. With target-based registration and good control geometry, project-level registration accuracy of 1-3mm is achievable across a typical commercial floor. Accuracy degrades past 60-70m as range noise increases. Dark, specular, or transparent surfaces (glass, black rubber, polished stainless) absorb or scatter the laser, creating returns that are noisy or absent - scan angle and target density matter more in those environments.

Can photogrammetry replace LiDAR for building as-built surveys?

For exterior facades and LOD 200 massing on buildings with textured exterior surfaces: sometimes, and the cost savings can be real. For interior as-built work, BIM requiring LOD 300 or higher, or any project with clash detection requirements: no. Photogrammetry fails silently on featureless interior surfaces - painted drywall, concrete, glass - producing point clouds with 20-50mm geometric errors in those regions that have no visual indicator in the output. The downstream path from a photogrammetric mesh to a Revit model also requires significantly more manual intervention than the TLS → ReCap → Revit pipeline, and the modeled result will contain accuracy gaps in exactly the areas where MEP coordination is most sensitive.

Why are professional 3D scanners so expensive?

Three compounding factors: precision components, calibration infrastructure, and software. The laser and timing circuitry must maintain consistent calibration across the full scan duration and across ambient temperature changes - any drift introduces systematic range error. The angular encoder resolves position consistently across tens of thousands of rotations per scan, and at 10m range, angular encoder error is the floor on achievable point accuracy. The dual-axis compensator auto-levels to within 5 arc-minutes, eliminating a major setup error source. The IP55 housing keeps optics and electronics functional in a dusty concrete pour or a freezing parking structure. On top of hardware: Trimble Business Center, Leica Cyclone 360, FARO Scene, and ReCap Pro collectively run $20,000-$55,000 per year in licensing. The business case is straightforward: projects without accurate as-builts commonly encounter avoidable change orders that a TLS as-built would have prevented. On a $5M renovation, even a modest change-order rate produces cost exposure that dwarfs a $4,000-$8,000 TLS as-built.

When should I stop using tape measures and switch to 3D scanning?

Under 500 sq ft with simple geometry and no BIM or clash-detection requirement: tape and laser distance meter are fine. At 500-2,000 sq ft with complex MEP, curved walls, or any BIM deliverable: the case for TLS is strong. At 2,000 sq ft and above with a BIM or permit-drawing requirement, the ROI is unambiguous - our TLS crews cover 20,000-50,000 sq ft per day versus roughly 500 sq ft per day for a manual measuring team producing the same deliverable spec. The break-even point on field cost alone is roughly 2,000-3,000 sq ft.

The change-order risk math is straightforward: projects without accurate as-builts commonly face avoidable change orders, and the cost of a TLS as-built is modest relative to that exposure. At $1M+ renovation value, the expected value of scanning is unambiguous.

What point cloud file format should I ask my scanner to deliver?

Request .E57 as the primary deliverable - it is the hardware-neutral open standard supported by virtually every downstream application including Revit, AutoCAD, CloudCompare, and all major registration platforms. If your team works in Revit or AutoCAD, also request .RCP/.RCS (Autodesk’s indexed format for direct ReCap and Revit import - ReCap Pro indexes the .E57 into .RCS tiles that Revit links natively without re-processing). For survey or GIS workflows, .LAS or .LAZ is the standard. Mesh outputs from photogrammetry export as .OBJ or .FBX and are not interchangeable with point cloud formats for BIM purposes - a Revit user cannot snap to a mesh the way they snap to a registered point cloud. For a full breakdown of deliverable specifications, see understanding laser scanning deliverable specifications.

---

Ready to Stop Guessing Which Technology Fits Your Project?

We deploy the Trimble X7 terrestrial scanner and handheld scanners including the Creaform MetraSCAN as daily tools on commercial, industrial, and historic preservation projects; based in the New York metro area, we travel nationwide. We publish our field protocols, our maximum registration tolerances, and our QC deliverables - because clients who understand what they’re buying make better project decisions, and projects with better decisions don’t generate the change orders that erase margin.

Every project scope we write identifies the specific capture technology, the registration method, the maximum acceptable residual, and the deliverable format - before we invoice a deposit. If a photogrammetric workflow is genuinely the right call for your project, we will say so and price it accordingly. If it isn’t, we’ll show you why with numbers.

Request a free scope consultation or send us your project specs for a same-day quote. Review our professional 3D laser scanning services for scope types, deliverable options, and what to expect from our field process.