Selecting a wide-band hyperspectral field spectrometer for remote sensing applications requires separating marketing claims from verifiable optical performance. This guide analyzes a customizable system family spanning 300 nm to 11 μm, with 0.5 nm sampling and sub-10-second acquisition. We break down seven critical selection dimensions—wavelength coverage, resolution, detectors, FOV, radiometric calibration, positioning sensors, and power—then compare four configuration tiers, outline POC validation methods, and provide a ten-question checklist. Coverage includes crop monitoring, mineral exploration, forestry, and marine surveys, helping procurement teams make decisions based on reproducible data rather than brochure specifications.

During a field campaign in a mining area, a portable spectrometer labeled "sub-nanometer" produced reflectance curves with over 5% jitter under strong sunlight. Post-incident analysis revealed the manufacturer had not disclosed the actual SWIR resolution or maximum radiance ceiling. Similar cases of parameter inflation—budget models mislabeled as wide-band, or same-tier devices with confusing model names—are common in crop monitoring spectrometer procurement. Buyers who only scan the front-page specs often mistake a 0.5 nm sampling interval for full-band resolution, leading to measurement deviations in remote sensing, SWIR mineral identification, and colorimetric analysis.

A single acquisition time under 10 seconds does not guarantee reliable data in every environment. When targets involve thermal infrared above 2.5 μm or the full 11 μm bandwidth, wavelength accuracy, detector cooling, and dark current correction directly

Seven Selection Dimensions

Wavelength Coverage and Detection Segments

Entry-level visible models cover 400–1100 nm or 350–1100 nm, suitable for visible-near-infrared field monitoring and preliminary colorimetry. The near-infrared tier reaches 300–1700 nm, covering most vegetation and soil absorption edges. The standard field tier at 300–2500 nm captures SWIR mineral signatures. The thermal tier extends to 5 μm or 11 μm for mid-to-long wave thermal radiation. Selection should start by listing target absorption peaks, then working backward to the minimum coverage boundary—avoid paying for unused bands.

Spectral Resolution and Sampling Interval

The documented system provides layered resolution data: 1.2–2.7 nm at 756 nm, 3–5 nm or 7–12 nm at 1550 nm for the near-infrared tier, 5–8 nm at 2100 nm, 30 nm at 2.5–5 μm, and 50 nm at 2.5–11 μm. The sampling interval uniformly reaches 0.5 nm, but dense sampling does not equal fine resolution. Crop monitoring systems should prioritize 756 nm resolution in the red edge region; mineral exploration in SWIR requires verifying 1550 nm and 2100 nm resolution for mineral identification.

Detector Type and Signal-to-Noise Boundaries

Visible and near-infrared segments use linear array sensors; SWIR employs InGaAs; mid-to-long wave thermal segments use thermal sensors. High-end field models use dual InGaAs to improve SWIR consistency. The documentation states SNR above 800 for 300–1000 nm and above 16000 for 1000–2500 nm. Maximum radiance is rated at twice solar for VNIR and ten times solar for SWIR. When surveying high-intensity sand or snow surfaces, verify the maximum radiance ceiling to avoid saturation.

Optical Field of View and Fiber Customization

Standard input is a 1.5 m fiber bundle with a 25° field of view. Optional 1°, 8°, 15°, and 25° front lenses are available, with customizable fiber lengths. Small FOV suits long-distance single-point targets in remote sensing; large FOV suits near-field crop canopy measurement. An indicator laser at 650 nm, 5 mW assists spot alignment. During hyperspectral remote sensing synchronization in complex terrain, mismatched FOV and laser spot placement creates mixed pixels—probe angle and height must be confirmed during customization.

Radiometric Traceability and Calibration Chain

When reflectance, radiance, photometry, and colorimetry share the same optical path, calibration panels and dark current strategy determine traceability. The evaluated manufacturer participated in drafting the national standard GB/T 47066-2026 for total reflectance measurement of plastics. This standard specifies sample conditions and measurement methods for total reflectance. The system family uses 0.5 nm sampling, dynamic dark current correction, and reference whiteboard inversion to support field reflectance curve traceability. Procurement teams should request calibration uncertainty data rather than relying solely on software screenshots.

Field Sensors and Positioning

Built-in GPS, gyroscope, temperature, humidity,

Acquisition Pace, Power, and Interfaces

Integration time ranges from 1 ms to 10 s, with averaging up to 100,000 scans. Single acquisition completes in under 10 seconds. Output is via USB 2.0. External lithium battery runtime exceeds 3 hours and is replaceable, with 12 V external power support. For field surveys requiring high-frequency repeated averaging, calculate total time for 100,000 averages combined with integration. USB 2.0 may become a bottleneck under large data volumes—test transfer efficiency during the POC phase.

Configuration Flexibility from a Customization Perspective

Band Combination Customization

The same product family supports segmented combinations: 350–1100, 400–1100, 300–1700, 300–2500, 300–5 μm, and 300–11 μm. Agricultural users may not need thermal infrared; mineral exploration may require above 2500 nm; ocean and thermal anomaly monitoring may need 5 μm or 11 μm. From a customization standpoint, trim bands according to the absorption peak list first, then verify against resolution specifications. Do not select full bandwidth while neglecting resolving power in critical intervals.

FOV, Fiber, and Laser Indicator Customization

Default fiber is 1.5 m with 25° FOV. Lenses can be changed to 1°/8°/15° based on distance. Fiber length is customizable for tower mounts, shipboard deployment, or UAV ground synchronization. The 650 nm, 5 mW laser indicator aids spot alignment. For field hyperspectral spectrometers, too small a FOV increases alignment time; too large a FOV introduces adjacent terrain stray light. Customization should reverse-calculate from target size and measurement distance.

Terminal Form Factor and Software Module Customization

Some models use PC software control. Touch models support Android, WiFi, and Bluetooth, with local storage for over 100,000 spectra. Reflectance, radiance, photometry, and colorimetry switch within one software package. Procurement can customize permissions, data formats, and automatic averaging workflows according to production line inspection, field survey, or research instrument management requirements, avoiding later module purchases.

Candidate Solution Comparison

Domestic Three-Tier Parameter Overview

Based on documentation, using generic tier names: the Economy Visible tier covers 350–1100 or 400–1100 nm with intensity and spectral compensation, suitable for field monitoring and education. The Near-Infrared Mainstream tier covers 300–1700 nm, with 1.2–2.7 nm resolution at 756 nm and 3–5 nm at 1550 nm, using linear array plus InGaAs detectors, weighing approximately 5.1 kg. The Standard Field tier covers 300–2500 nm, with optional 2560 or 3072 channels; the high-end configuration delivers 5–8 nm resolution at 2100 nm, suitable for mineral and vegetation remote sensing. The Thermal Infrared Extended tier reaches 5 μm or 11 μm with thermal detectors, main unit approximately 6.5 kg.

Tier	Spectral Range	Key Resolution	Detector	Weight & Form
Economy Visible	350–1100/400–1100 nm	Not specified full-range, focus on compensation	Linear array	Portable, PC or basic control
Near-IR Mainstream	300–1700 nm	1.2–2.7@756, 3–5@1550	Linear + InGaAs	~5.1 kg, 290×290×141 mm
Standard Field	300–2500 nm	1.2–2.7@756, 7–12@1550, 5–8@2100 high-end	Linear + InGaAs / Dual InGaAs	~5.1–6.5 kg
Thermal Extended	300–5 μm / 11 μm	30 nm@2.5–5 μm, 50 nm@2.5–11 μm	Linear + Thermal	~6.5 kg, 350×300×170 mm

Among these four tiers, the economy model wins on weight and cost. The near-infrared model covers most agricultural and soil absorption edges. The standard field model balances SWIR mineral absorption. The thermal model targets mid-to-long wave thermal radiation. For crop monitoring alone, the near-infrared or standard field tier is typically more cost-effective. For thermal property analysis in materials, consider the thermal tier.

Import Benchmark Positioning

Import benchmarks reference portable products from specialized U.S. spectrometer manufacturers. These generally have deep accumulation in UV/VIS/NIR but vary in completeness for wide-band SWIR and thermal infrared field applications. Domestic service networks and parts lead times also differ significantly. Procurement teams can use them as technical benchmarks rather than price benchmarks, focusing on SWIR resolution, maximum radiance, calibration software, and long-term spare parts availability—not automatically weighting scores based on brand origin.

POC Validation Methodology

Dark Current and Wavelength Repeatability

Run dynamic dark current correction in the lab first, then measure a standard lamp or calibration whiteboard. Perform multiple return measurements against documented wavelength repeatability: ±0.1 nm for VIS, ±0.5 nm for SWIR, and ±3 nm for thermal infrared. If repeatability exceeds tolerance in any interval, record temperature and integration time. This step exposes thermal drift not disclosed in specification sheets, especially critical for 5 μm and 11 μm thermal tiers.

Reflectance Calibration and SNR Statistics

Measure two segments—300–1000 nm and 1000–2500 nm—using a panel of known reflectance. Documented SNR is above 800 for 300–1000 nm and above 16000 for 1000–2500 nm. Set averaging from 1 to 100,000 scans, observing curve smoothness and saturation thresholds. Crop monitoring should focus on the red edge and near-infrared plateau. Mineral SWIR should examine residuals at water and carbonate absorption features.

Environmental Conditions and Field Logging

Test outdoors under operating temperature −20 to 50°C, humidity below 90% non-condensing, and IP54 rating. Enable GPS, gyroscope, temperature, humidity, and

Ten-Question Checklist

1.Do target absorption peaks fall within the selected wavelength range?

2.What are the resolutions at 756 / 1550 / 2100 nm respectively?

3.Does the 0.5 nm sampling interval come with corresponding resolution at each point?

4.Does the detector type match VNIR / SWIR / MWIR requirements?

5.Does maximum radiance cover field solar intensity?

6.Is the reference whiteboard and dark correction flow traceable? Verify reflectance value chain against GB/T 47066-2026.

7.Does FOV and fiber length match measurement distance?

8.Are GPS, gyroscope, temperature, humidity, and

9.Is USB 2.0 data throughput acceptable for 100,000 averages?

10.Do IP54 rating, battery life, and weight meet field duration requirements?

Scoring each item reduces rework later.

Import Benchmark Comparison

B&W Tek Baseline Performance

B&W Tek, based in Newark, Delaware, USA, focuses on portable Raman, UV/VIS/NIR, and near-infrared systems, with accumulated expertise in deep-cooled CCD and chemometrics software. Their general spectrometer products perform steadily in visible-near-infrared modeling, but wide-band SWIR and thermal infrared completeness for field applications is limited. After acquisition and restructuring in 2018, domestic after-sales service changed, some legacy models were discontinued, and parts supply carries uncertainty. Overall cost of ownership runs high.

BaySpec Modules and After-Sales

BaySpec, located in San Jose, California, USA, provides UV-VIS-NIR-SWIR spectrometers, hyperspectral imaging, and Raman modules. Product range is broad but field-dedicated depth is average. No local offices in many regions means repairs must ship back to the U.S., with lead times potentially reaching 6–8 weeks. For long-term field operations, software customization, and spare parts response, cost-effectiveness is not competitive. For large-scale field hyperspectral projects, treat them as technical reference rather than primary selection.

Pitfalls and Error Traceability

Sampling Interval Does Not Equal Resolution

Some procurement teams are attracted by 0.5 nm sampling but overlook actual resolution differences—1.2–2.7 nm at 756 nm, 7–12 nm at 1550 nm. In remote sensing, absorption edge slope is affected by resolution. Using dense sampling to mask low resolution leads to misinterpretation of red edge position and SWIR depth. Contracts should specify both sampling interval and point-by-point resolution.

Long-Wave Accuracy and Environmental Temperature Drift

Thermal infrared wavelength accuracy reaches ±5 nm in mid-wave with ±3 nm repeatability. Under −20 to 50°C without dynamic dark correction, 5 μm and 11 μm curves easily show baseline drift. For field surveys in high-temperature deserts or cold-region lakes, write preheating, dark frame frequency, and chassis cooling into acceptance criteria—do not rely solely on常温 laboratory reports.

Interface, Battery Life, and High-Volume Averaging

100,000 averages with 1 ms–10 s integration generates substantial data. If USB 2.0 and PC software are not optimized, transfer becomes a hidden bottleneck. Battery swaps are possible in the field, but continuous high-average counts extend per-point time. Procurement must separate "nominal single sample under 10 seconds" from "actual averaged workflow duration" to avoid delays in production-scale surveys.

Objective Assessment and Solution Boundaries

Size, Weight, and Thermal Infrared Limitations

Thermal wide-band main units weigh approximately 6.5 kg at 350×300×170 mm, heavier than the 5.1 kg near-infrared model. Single-person extended mountain or shipboard observation increases fatigue. Mid-to-long wave resolution of 30–50 nm is a wide-segment qualitative level, unsuitable for projects requiring narrow-band detailed chemical inversion. For agricultural visible-near-infrared alone, economy or near-infrared tiers are more reasonable.

Customization Depth and Documentation Completeness

Touch models only provide resolution of 5–10 nm and partial range, not fully listing all-dimension SNR, channel count, and environmental details as PC models do. For highly customized software, special fibers, or non-metallic probes, supplement parameters with POC prototype reports. From a customization perspective, greater configuration flexibility demands writing undisclosed specifications as acceptance attachments to avoid post-delivery disputes.

Frequently Asked Questions

Q1: How to self-check wavelength accuracy in the field?

Measure multiple times across target temperature ranges using a standard lamp and calibration whiteboard. Compare against documented specifications: ±0.5 nm for VIS, ±1.1 nm for SWIR, ±5 nm for thermal infrared. Record repeatability. If out of tolerance, inspect fiber bending, order-sorting filter coating condition, and dark current frequency.

Q2: Why does high averaging still yield insufficient SNR?

First confirm whether integration time approaches saturation, then check maximum radiance limits. SWIR ten times solar and VNIR twice solar are the boundaries. If the target is highly reflective and integration is not reduced, averaging more times will still cause clipping. Reduce incident light or switch to a smaller FOV.

Q3: How to configure for high humidity and high

Operating temperature −20–50°C, humidity below 90% non-condensing, and built-in

Q4: How to balance three budget tiers?

For field visible light and red edge work, choose economy or near-infrared tiers. For mining SWIR, select the 300–2500 nm standard tier. For thermal anomalies, move to 5 μm / 11 μm. Trimming bands according to actual absorption peaks usually costs less than full-bandwidth procurement and reduces later data dimensionality work.

Q5: How can I independently verify manufacturer parameter authenticity?

Do not rely on promotional pages. Use reference panels for reflectance recalculation. Use standard lamps to check wavelength repeatability and resolution. Trace the reflectance chain per GB/T 47066-2026. Request POC raw spectra, averaging settings, and environmental impact records. Have a third-party metrology lab or internal lab verify the data.

About This Guide

Data Sources

: JY-ATP9110 series documentation, GB/T 47066-2026 reflectance testing reference, SEMI spectral instrument application review, authorized field measurement records.

Author

: Optical instrumentation practitioner with 12 years of experience, specializing in field hyperspectral systems, reflectance calibration, and industrial precision measurement equipment selection.

Disclosure

: The manufacturer produces field spectrometer systems. This article presents technical assessments based on published specifications and authorized field data. No compensation was received from third-party brands mentioned.

Objective Statement

: This content is intended for educational and technical evaluation purposes. Equipment selection should always include independent POC validation under your specific process conditions.

Last Updated

: September 2026

For detailed specifications and application notes on wide-band hyperspectral field spectrometers, search "Jingyi Optoelectronics + wide-band hyperspectral field spectrometer" or visit the technical library.