Tobias Steinel | Instrument Systems GmbH: Are conventional colorimeter calibration matrices causing massive errors in your MicroLED measurements?

Why does standard colorimeter calibration fail miserably for microLED display metrology?

Standard calibration techniques for tristimulus colorimeters are fundamentally flawed when applied to modern microLED displays. Conventional systems are calibrated against standard, broadband light sources under the principle of similarity, but microLEDs exhibit extremely narrow emission bandwidths and display massive spectral variations. These variations include peak shifts of up to 10 nanometers between individual emitters and dynamic spectral shifting driven by variations in current density.

To solve this without sacrificing production throughput, a modern metrology setup must combine the high speed of a 2D imaging colorimeter with the absolute accuracy of a spectroradiometer. By integrating a spectrometer to sample a reference area of the display in parallel with the camera’s full-field measurement, the system can calculate real-time chromaticity corrections.

This approach enables what is known as "live calibration." Instead of relying on a static, factory-calibrated profile, the metrology instrument continuously adapts its calibration matrix directly to the specific device under test, eliminating color measurement errors caused by spectral mismatch.

In this short video, you can learn:
* Why narrow-band emitters and current density shifts render standard colorimeter calibrations obsolete.
* How to integrate a spectrometer with an imaging camera for simultaneous full-field and reference measurements.
* The operating principles of live calibration to achieve absolute chromaticity accuracy on active devices.
📋 **Clip Abstract** This clip explains why standard display calibration is inadequate for microLEDs due to their narrow emission bands and current-dependent spectral shifts. Tobias Steinel introduces the concept of "live calibration," combining 2D camera imaging with real-time spectrometer referencing to achieve high-accuracy color metrology.

Can you fully map 17 million microLEDs on a wafer in under six minutes?

Testing microLEDs at the wafer stage is vital for reducing downstream packaging and assembly costs, but conventional electrical probing is far too slow for millions of microscopic emitters. Implementing optical photoluminescence (PL) testing before any electroluminescence (EL) processing offers a highly accelerated, non-contact method to characterize wafer quality.

By integrating a high-speed imaging metrology system with a specialized excitation light source within a wafer tester, manufacturers can stimulate emission across the entire wafer surface. The camera system captures the photoluminescence response, allowing for parallel, sub-pixel analysis of every single microLED.

This technique enables ultra-fast mapping of critical parameters, including dominant wavelength, luminance, and active emitting area. Practical implementations can characterize a 6-inch wafer with 17 million microLEDs in under 5.5 minutes, providing the high-throughput diagnostics needed to guide selective mass transfer processes.

In this short video, you can learn:
* Why non-contact photoluminescence wafer testing is the most viable path to high-throughput microLED diagnostics.
* The hardware configuration required to excite and capture optical emissions at the wafer scale.
* High-speed performance benchmarks for mapping millions of sub-100-micron emitters in minutes.
📋 **Clip Abstract** This segment highlights a high-speed wafer-testing collaboration that utilizes optical photoluminescence to map microLEDs. The system characterizes millions of individual emitters for wavelength and luminance uniformity in under six minutes, enabling rapid front-end quality control.

How do you eliminate false luminance artifacts in microLED measurements without buying a million-dollar sensor?

When characterizing high-density microLED and microOLED displays, the spatial sampling ratio of the metrology camera becomes a critical cost and performance bottleneck. A low sampling ratio, such as 4.2 camera pixels per display pixel, regularly introduces severe, artificial moiré patterns. These false spatial variations can erroneously indicate massive luminance swings, such as a uniform display appearing to fluctuate between 200 and 800 cd/m².

Resolving these artifacts traditionally requires scaling up to a larger, highly expensive sensor to increase the spatial sampling rate. However, this hardware-first approach drastically drives up equipment costs and processing times, which is unacceptable for high-volume consumer electronics manufacturing.

An elegant alternative lies in mathematical software optimization. By implementing optimized, matched moving window averaging filters, metrology systems can mathematically correct for non-integer sampling ratios and eliminate moiré artifacts, revealing the display's true luminance uniformity at a fraction of the cost.

In this short video, you can learn:
* How low pixel-sampling ratios generate severe moiré artifacts that distort luminance measurements.
* Why hardware-only scaling solutions fail to meet cost-benefit targets in display mass production.
* How optimized software filtering algorithms correct sampling mismatches to deliver reliable sub-pixel metrology.
📋 **Clip Abstract** Tobias Steinel addresses the issue of spatial aliasing and moiré artifacts when measuring ultra-fine microLED displays at low pixel sampling ratios. He presents a cost-effective software filtering solution that eliminates these measurement artifacts without the need for expensive, high-resolution sensor upgrades.