LMK Display Sparkle: IEC 62977-3-9 AGL Sparkle Testing

LMK Display Sparkle is a LabSoft add-on for TechnoTeam's LMK 5 and LMK 6 luminance measurement cameras. It quantifies the high-frequency luminance non-uniformity known as sparkle, which appears on displays with an anti-glare layer (AGL). The measurement procedure is compliant with IEC 62977-3-9 and was co-developed by TechnoTeam, Volkswagen and Elektrobit. The add-on is part of TechnoTeam's display package.

A measurement combines a BlackMURA-compatible alignment, an automated distance focus scan that locates the focus position at which sparkle is most strongly perceived, and a frequency-filter evaluation that separates the high-frequency sparkle signal from the periodic pixel grid. The result is reported as sparkle contrast in percent.

Yes. Sparkle measurements depend on factors such as the angular aperture of the lens or the magnification at which the display is observed, among many others. All of these need to be documented for the result, as reproducibility would otherwise be limited. LMK Display Sparkle controls these factors automatically during the measurement and records them as part of the measurement protocol.

IEC 62977-3-9 does not prescribe a single processing algorithm. It permits several different methods to separate the sparkle signal from the underlying pixel matrix, and LMK Display Sparkle uses a frequency-filter approach to do so. The standard also does not define a numerical sparkle threshold. Acceptance criteria are set by the OEM or by the end-user specification.

The add-on works with all TechnoTeam LMK 5 and LMK 6 luminance cameras from 16 mm focal length upwards. The validated lens set is 16 mm, 25 mm and 50 mm. This includes the same standard lenses already used in BlackMURA setups.

A monochrome ILMD is required. Sparkle is evaluated on a single colour channel of a green test image, and an RGB Bayer-pattern sensor would distort the high-frequency signal through spectral crosstalk and demosaicing interpolation. The LMK 6 colour camera uses a filter wheel with a monochrome sensor and is fully compatible. The LMK 5 can also be used in combination with the current LabSoft, the LMK Display Sparkle add-on and a translation stage for the distance focus scan.

The same equipment — camera and lens — is used for both measurements, and the geometric alignment follows the same BlackMURA pattern and tolerances. A BlackMURA-compliant setup is therefore also a valid sparkle setup.

What can change between the two measurements is the measurement distance and focus. For larger panels in particular, the sparkle measurement may need a different distance than the BlackMURA capture you have already optimised. The measurement procedure includes a Fourier downsampling step that was designed precisely for this case: different measurement distances within the validated sampling window produce comparable sparkle values, so you can pick the distance that fits your panel and your cell without losing reproducibility.

The sparkle pattern does not necessarily lie exactly on the display pixel plane. It sits somewhere inside the depth of focus of the camera-lens combination, at a position that depends on the AGL, its position, and the display pixel geometry. Manual focus and autofocus alone cannot reliably find this position, because the signal level looks almost identical across the depth of focus. Without a focus scan, two operators can therefore read significantly different sparkle values from the same panel, purely as a result of focus.

The distance focus scan resolves this. The camera is moved over a small distance range, a sparkle value is calculated at each step, and the maximum value is taken as the result. That maximum corresponds to the focus position at which the human eye perceives sparkle most strongly, and it is reproducible across operators and setups. The scan only has to be run when the setup or the panel type changes; for repeated measurements on the same panel it does not need to be repeated.

The distance focus scan can be performed manually, with an optional motorised linear axis, or fully automatically as part of an LMK Position 6-axis robotic system. A manual translation stage is the simplest and most economical option and is well suited to engineering labs. A motorised axis removes the manual hand-wheel and is typical for repeated measurements. The LMK Position system performs the BlackMURA-compatible alignment and the distance focus scan automatically and was also the platform used for the original setup-sensitivity studies that derived the procedure.

Sparkle can be measured at any stage of the lifecycle, because our procedure does not require physical separation of the AGL from the display. The same measurement can be applied to coated cover glass on a reference display, to a sub-assembly, or to a finished display module, which means the same procedure covers AGL development, supplier qualification and incoming inspection. The validation set used to derive the procedure included both glass-based and foil-based AGLs and displays with different PPI, so the typical real-world configurations are covered.

It does, within practical limits. Sparkle is not evaluated over the full display surface; the measurement uses a representative region — typically near the centre — where the camera resolves the pixel structure. For displays with low to moderate curvature this is unproblematic. For displays with stronger curvature, choose a setup with a larger depth of focus so that the local curvature does not push the evaluation region out of focus. The procedure itself — alignment, distance focus scan, frequency-filter evaluation — is unchanged.

Reproducibility starts with the hardware and the setup, not with software alone. LMK Display Sparkle specifies a defined range of cameras (LMK 5 / LMK 6, monochrome) and lenses (16 mm, 25 mm, 50 mm) and a defined measurement window — sampling resolution, angular aperture, alignment tolerances — that has been validated to produce comparable sparkle values across that range. When you stay within this window, you are using a setup TechnoTeam has qualified for the method.

Within that window, the measurement software smooths out the remaining differences between setups. The distance focus scan removes the operator dependence on the focus position, Fourier downsampling normalizes the result to a common reference frequency so that two different magnifications produce comparable values, and image averaging keeps the contribution of sensor noise under control. These algorithmic steps extend the practical flexibility of the method, but the foundation is still the right hardware in the right setup.

This combination was validated in a round-robin experiment with three independent labs, three different ILMDs and operators who had not been involved in developing the procedure. The same panels read comparable sparkle values across all three labs, for both glass-based and foil-based AGLs.