A windshield HUD must be evaluated with its actual laminated glass, not only with a PGU on a bench. Reflections from two optically effective surfaces can create a displaced secondary image; a suitably designed wedge can reduce that separation at intended eye positions. This guide separates ghosting from distortion and stray light, then sets out system-level checks without claiming an ALPHA OPTIK vehicle result.
GLOBAL IMAGING BRIEF
For an OEM or Tier 1, a bright PGU image does not establish a readable HUD. Glass curvature, wedge, assembly position and driver eye location all affect where the virtual image and its ghost appear. Supplier review should therefore use the target windshield and a measured eye-box grid, with separate acceptance criteria for double-image separation, distortion, brightness and thermal drift.
Full perspective
The picture generation unit, or PGU, is the image source of an automotive head-up display. The HUD's mirrors and other optics direct light toward the windshield, which reflects a useful portion into the driver's eyes so that information appears as a virtual image ahead. The windshield is therefore part of the optical system, not a neutral cover. ALPHA OPTIK's public HUD PGU page identifies LCoS and DLP architectures and discusses brightness, contrast, field of view, stray light, ghosting, solar load and packaging as development concerns. It does not provide a measured ghost-separation map or qualification result for a particular vehicle.
A laminated windshield has more than one optically effective reflecting interface. The inner surface commonly forms the intended reflection while another path associated with the outer side can produce a fainter, angularly displaced image. This is a different problem from a blurred PGU source, geometric distortion of a single image or a bright dashboard reflection. A buyer should ask for those defects to be measured separately. The analysis below draws on general HUD optics and primary technical explanations from windshield and optical-engineering sources; it does not imply that ALPHA OPTIK supplies a specific wedge-glass design.
A ray from the HUD can reflect at the cabin-facing glass interface, while another portion travels through the laminated structure and returns from a second effective interface. The two paths do not generally send identical angular bundles to the eye. The result can be a primary symbol and a displaced, lower-intensity ghost. Its apparent separation depends on incident angle, local glass thickness and wedge, curvature, image location and eye position. Treating the windshield as one flat mirror misses the very variable the vehicle team must control.
Ghost separation should be reported in an agreed angular or image-plane metric for the actual viewing region, together with the relative ghost brightness. A secondary image that nearly overlaps may still affect symbol contrast, while one farther away may become a clearly readable duplicate. Neither a photograph taken from one eye position nor a single central-field number describes the entire eye box. The test target should contain fine lines and symbols at multiple positions so both offset and legibility can be reviewed.
HUD windshields may vary the laminate thickness so the two effective reflecting surfaces are not parallel. A suitable wedge changes the relative directions of the two reflected paths, bringing the primary and secondary images closer together for a defined design eye position. It does not work by simply sending the ghost outside the useful field. Because the glass is curved and the eye may move, an adjustment that aligns one point may leave residual separation elsewhere. The wedge must therefore be considered with the HUD ray bundle and the vehicle's eye-box specification.
A glass supplier's wedge measurement and an installed HUD image test answer different questions. The former checks a manufactured property of the windshield; the latter measures what a driver actually sees after the PGU, mirrors, glass and dashboard are assembled. AGC describes evaluating both image distortion and double-image magnitude with the actual HUD image. That distinction is useful for assigning responsibility when a windshield variant or replacement part is introduced: component conformance does not automatically establish system conformance.
Virtual-image distance and apparent position are set by the combined PGU, relay optics, windshield and viewing geometry. Glass curvature contributes optical power and distortion, but thickness alone does not determine where the virtual image appears. The design team should keep a map of virtual-image location across the displayed field and eye box, separate from the map of double-image separation. A displaced whole image can be an alignment or distortion problem even when no ghost is visible; conversely, a ghost can remain in a correctly placed primary image.
The selected virtual-image distance is a vehicle-program decision linked to display content and human-factors assessment. A far virtual image can reduce the driver's need to refocus relative to a near dashboard screen, but it does not make refocusing or head movement disappear for every driver and situation. Avoid inserting a universal distance range into a supplier brief. Specify the target distance, allowed deviation and measurement method for the intended display modes instead.
An eye box is the region from which the required HUD content remains usable. Measure the image from a grid of positions representing seat adjustment and driver stature, not only the nominal central eye. At each position, record ghost separation, brightness, uniformity, focus impression and distortion for symbols at the center and outer field. Include both eyes where the vehicle's human-factors method requires it. A correction that improves a central camera image may be inadequate if the ghost becomes conspicuous when the driver's head moves.
Repeat the grid under representative ambient light and with the target windshield rather than a generic glass coupon. Sunlight entering the HUD aperture can heat internal parts or create unwanted reflections; night scenes can reveal weak ghosts against a dark background. Keep those phenomena separate in the report. Optical ghosting from the windshield, stray light within the module, solar loading and dashboard reflections may demand different remedies and different ownership between suppliers.
The system stack runs from microdisplay or PGU through relay optics and mounting hardware to the installed windshield and eye position. A tolerance review should consider PGU placement, mirror angle and figure, glass wedge and curvature, dashboard aperture and vehicle assembly variation. The output is not a single 'strict alignment' instruction but a sensitivity map: which change moves the primary image, which separates the ghost and which reduces brightness or adds stray light. That map helps choose adjustment points and inspection stages.
Thermal cycling, vibration and sunlight exposure belong in a reliability plan, but a component pass is not proof of vehicle-level image quality. Measure the same eye-box and field criteria before and after stress and compare the installed system with its baseline. If a replacement windshield is supplied, verify that its optical variant and calibration method match the HUD system. The official ALPHA OPTIK page describes development directions; it does not publish a specific vehicle's tolerance budget, test outcome or qualification certificate.
Ask for the target windshield geometry and optical sample used in the proposal. Request separate maps of double-image separation, ghost-to-primary brightness, virtual-image position, distortion, luminance and stray light across the eye box and display field. Ask how the proposed wedge or alternative combiner treatment is evaluated with the PGU and relay optics, and who owns recalibration after a glass change. A render or a PGU-only bench photograph cannot answer these installed-system questions.
ALPHA OPTIK's public HUD PGU material identifies optical and integration topics that can begin a project discussion, but it is not evidence of a qualified windshield–HUD combination. A prospective OEM or Tier 1 can bring its field-of-view, eye-box, virtual-image, package and environmental requirements to an inquiry and agree what must be demonstrated on the target vehicle. The outcome should be a testable joint scope, with manufacturer claims kept distinct from independently observed system performance.
Original source
Original technical analysis drawing on ALPHA OPTIK's HUD PGU page and primary explanations from AGC and Ansys Optics. The windshield double-reflection and wedge discussion describes general HUD engineering; ALPHA OPTIK has not published a vehicle-specific ghost map, approved windshield combination or measured performance in the cited page. No certification or production deployment is asserted.