RESEARCH SYNTHESIS / ENGINEERING RECONSTRUCTION Integrated Lens-First Transparent Photonic Eyewear Electromagnetic Architecture, Multimodal Optical Sensing, Binocular Field Delivery and Fabrication Verification D: RGB source/phase C: world detector E: NIR source/detector Primary lens-first construction — plan view Interleaved grid is routing geometry. Every function requires distinct usable optical modes. Unified research and engineering monograph Based on two supplied studies 4 and 9 November 2026 | Revised 11 November 2026 96 indexed and repaired work instructions | Reproducible reduced-model calculations | Original vector figures Evidence state. This document describes a proposed optical system, internally reproduced mathematical examples, source-audited work instructions, and selected verified published precedents. It does not report successful physical fabrication, a qualified full-wave model of an assembled lens, human testing, or laser safety certification. Abstract A lens-integrated photonic eyewear architecture is reconstructed from two technical source PDFs into one coherent research and manufacturing dossier. The proposed system places three separately calibrated optical interfaces in each transparent spectacle lens: phase-controlled visible-field delivery to its pupil, collection of ordinary external-scene light, and near-infrared eye illumination and return. The primary embodiment retains a square grooved crystalline carrier containing physically crossed drawn microfibres and a lens-bound peripheral region for sources, actuators, detectors, and first electrical interfaces. Patterned-guide and alternate crossing routes are kept as separately named comparators. An original reduced-model computation package reproduces crossing strain, accumulated route loss, optical phase sensitivity, receiver rank, photon statistics, ideal depth curvature, illustrative pupil power, and thermal/- electrical budgets. Engineering analyses identify material guiding, repeated curvature, independent optical control, weak natural-light reception, joint-channel crosstalk, source heating, and moving-pupil delivery as decisive physi- cal research questions. The 96 supplied manufacturing instructions are retained in an indexed annex and reordered so that pair-level measurements follow two individually verified one-lens builds. Selected primary references and official standards are checked against publisher or authoritative records. The result is an executable research specifi- cation and a reproducible documentary artifact; fabrication performance, human use, and product compliance remain unverified. Keywords: integrated photonics; transparent optical interfaces; drawn microfibres; square grooves; optical phased ar- ray; computational imaging; near-infrared eye tracking; holographic field delivery; binocular displays; optical metrol- ogy; process verification. i Editorial and provenance notice This manuscript consolidates a 106-page theoretical monograph and a 119-page engineering dossier uploaded by the requester. The original documents contain proposed calculations and laboratory instructions, not a set of fabricated lens measurements. All figures in this revision are newly drawn and labeled by inference level. The numerical pack- age was executed locally; a separate full-wave solver, original author code archive, qualified manufacturing equip- ment, measured material lots, and human optical exposure laboratory were not supplied. Source facts, independently checked publication claims, calculations, new design choices, and unresolved physical parameters are distinguished in the text. No scientific authorship, institution, funding declaration, ethics approval, or commercial device certifica- tion is assigned beyond the evidence provided. ii Contents Abstract i Editorial and provenance notice ii 1. Research proposition, evidence state, and architecture 1 1.1. The central scientific proposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2. Evidence taxonomy and the level of inference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.3. Two-source reconciliation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.4. Observable first-generation success . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2. Electromagnetic foundations and photonic geometry 4 2.1. Guiding in a physical core, sheath, and carrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.2. The square network is not an independent-pixel array . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.3. Woven crossing geometry and bending demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.4. Accumulated guided loss and crossings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 3. Phase-controlled display and field delivery 8 3.1. Optical superposition and active control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 3.2. RGB spectral operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 3.3. A real virtual image and the moving pupil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.4. Interferometric amplitude, source splitting, and output power . . . . . . . . . . . . . . . . . . . . . . 9 4. Outward scene reception and computational imaging 11 4.1. Receiving functions and information content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 4.2. Natural light, coherence, and timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 4.3. Photon and noise accounting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 5. Inward eye sensing and binocular geometry 13 5.1. Eye-return optical channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 5.2. Left and right lens registration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 5.3. Pupil-access, latency, and failure behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 6. Shared operation, power, calibration, and stability 15 6.1. Joint D/C/E operation and interference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 6.2. Calibration and computational feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 6.3. Electrical and thermal accounting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 6.4. Direct-view transparency and optional correction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 7. Manufacturing architecture and physical research gates 17 7.1. Passive coupon: P0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 7.2. Active single lens: P1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 7.3. Second lens and matched optical pair: P2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 7.4. Accessories and product development: P3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 7.5. Material, process, and interface release register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 8. Verification, fault response, and scientific conclusions 19 8.1. A test must make a false interpretation fail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 8.2. Optical exposure and independent source shutdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 8.3. What the present integration establishes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 9. Reviewed literature and remaining source coverage 21 9.1. Verified primary and official sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 9.2. Material limitations that remain open . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 9.3. Document provenance and visual repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 iii INTEGRATED LENSFIRST PHOTONIC EYEWEAR RESEARCH / ENGINEERING 10.Reproducibility dossier and numerical checks 23 10.1. Executable model scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 10.2. Reproduced cases and their status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 10.3. Independent limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Verified external references 24 A. Correction ledger and source reconciliation 27 B. Source-to-unified-section map 30 C. Controlled fabrication and verification instructions 31 C.1. Common requirements and operating status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 C.2. Repaired sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 C.3. Procedure index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Final engineering status 130 DOCUMENTARY RECONSTRUCTION | NO FABRICATED LENS iv / 130 List of Figures 1.1. The three permitted optical directions. The directions and the source/receiver identity are physical specifications. Their co-location is proposed and not a measurement of a completed device. . . . . . . 1 1.2. Physical evidence required for each gate. Documentary completion and a component precedent do not advance a gate that requires measured integrated specimens. . . . . . . . . . . . . . . . . . . . . . 3 2.1. Carrier, mode-guiding material, active field-control element, and package have different jobs. Alter- native material families require separate process identification. . . . . . . . . . . . . . . . . . . . . . . 4 2.2. Reconstructed lens plan, showing three optically distinct path families and a lens-bound peripheral active zone. The diagram is schematic, and does not establish that the illustrated intersections possess independent optical control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.3. Proposed centerline elevation for a single over–under crossing. The horizontal spacing, rise, and fiber diameter are example inputs; actual profile and optical surroundings require measurement. . . . . . . 6 2.4. Recomputed geometric bending proxy over routing pitch for three proposed outer diameters. The diagram is not a fracture-strength prediction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.5. Conditional crossing-count losses. The 6 dB line is a planning allocation, while propagation and cross- ing values are taken from proposed source targets rather than actual lens measurements. . . . . . . . 7 3.1. Original vector plot of the evaluated scalar finite-array model. Element radiation patterns, coherent source stability, real couplers, and pupil optics are deliberately absent. . . . . . . . . . . . . . . . . . . 8 3.2. Paraxial optical-path curvature for two commanded virtual depths. The plotted 5 mm effective aper- ture is an ideal parameter study, not a fabricated display. . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.3. Recomputed conditional example of optical energy remaining after each specified stage. Use the model solely for bookkeeping until actual source, route, splitter, and pupil transfer measurements exist. 10 4.1. Singular values of a reproducible 16-by-16 synthetic receiver matrix. It has full nominal rank for the chosen 16-bin task but is not a measurement of the proposed lens, and it cannot encode 1,024 unconstrained unknowns in 16 static readings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 5.1. World-scene and eye-return data flows are separate inverse problems and have different optical and calibration models. Every detector stage is real hardware; software alone cannot stand in for the optical mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 5.2. Separate lens and model-eye coordinate systems must be experimentally related before a shared virtual-point claim is accepted. The lines are conceptual and do not represent measured ray paths. . . 14 6.1. Direction of principal sources and receivers relevant to simultaneous-operation leakage. The diagram displays tests to run, not a measured leakage matrix. . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 6.2. Conditional phase sensitivity for the stated path and material coefficient, with the wavelength depen- dence shown directly. Measured composite-stack coefficients are needed before a thermal controller can be released. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 8.1. An independently testable source-disable path is required in the proposed control architecture. This schematic is not a Class 1 label or safety qualification. . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 v 1. Research proposition, evidence state, and architecture 1.1. The central scientific proposition The investigated device is a pair of transparent optical assemblies in which each spectacle lens carries three distinct directional operations: delivery of deliberately synthesized visible light to its associated pupil; reception of ordinary external-scene light through world-facing collection ports; and illumination and observation of the eye using sepa- rately addressed near-infrared paths. Each lens incorporates a square network of supported microscopic optical paths. The physical reference design consists of drawn optical microfibres placed into grooves in a transparent crystalline carrier, with a three-dimensional over–under centerline at selected orthogonal crossings. The sources, phase actua- tors, receivers, and first electrical interfaces belong to the lens or its bonded peripheral region. The spectacle frame has a mechanical and electronic service role, but does not replace an intended in-lens camera or optical image-forming device. This proposition is a system design , not an experimentally completed system. A compatible field theory exists for the individual optical operations, and several constituent mechanisms have been reported on different laboratory platforms. The two supplied source documents do not contain a physically manufactured three-function lens, a verified spatial pair, or a qualified eye-exposure result. The present reconstruction provides a coherent and inspectable design hypothesis, independent reproduction of selected reduced-model calculations, original vector diagrams, and a logically repaired fabrication program. It reports no new physical measurements. This boundary is foundational: the document can be complete as a research-and-engineering specification while the proposed hardware remains unverified. The most decisive unanswered question is whether the full light-guiding, phase-controlling, sensing, and transparent- viewing operations can coexist in one mechanically viable lens assembly with sufficient optical information and throughput. Three lines crossing in a drawing do not create three independent optical functions. An experiment must demonstrate distinguishable source, detector, and field-control responses in a fixed physical specimen; an appearance- only photograph cannot establish that information was obtained. External world scene radiance Transparent carrier Square-groove optical network D / C / E port families Pupil and model eye C: world light in D: RGB out E: NIR launch E: eye return Three optical directions share a carrier, not a single uncontrolled ray path Figure 1.1.: The three permitted optical directions. The directions and the source/receiver identity are physical specifications. Their co-location is proposed and not a measurement of a completed device. 1.2. Evidence taxonomy and the level of inference We use five explicit evidence labels throughout the monograph. S refers to a proposition or reported parameter in the supplied documents, with its location retained; L is a primary-literature result whose bibliographic identity and 1 INTEGRATED LENSFIRST PHOTONIC EYEWEAR RESEARCH / ENGINEERING stated experimental scope were checked against a publisher or authoritative record; C is a calculation reproduced here from identified inputs; D is a new design decision, screening target, or procedural requirement; and U is a physical parameter or capability lacking the necessary measurement or design input. A sixth label, X , identifies an operation that is executable only at a suitable laboratory and was not performed in producing this document. These labels do not measure intellectual importance. A mathematically correct numerical case, a valid published component demonstration, and an untested complete lens belong to different evidence classes. The source’s references to physical laser exposure review, active-die attachment, material curing, imaging results, and manufacturing yield remain proposed tasks until actual specimen records exist. A numerical script that regenerates an ideal array factor neither measures a phase actuator nor qualifies a product. Level Established in this package Not established here Source recon- struction Two supplied PDFs analyzed; 96 source pro- cedures retained and re-indexed. Original CAD, deposited layers, inspected specimens. Analytical calculation Reproducible reduced-field, geometry, pho- ton, power, and thermal examples. Full-wave simulation of the actual assembled microfibres. Experimental precedent Selected published phased-array, sensing, waveguide, and process studies. Three-function operation on a common trans- parent lens. Manufacturing Geometry trials and measurement conditions described. Process-qualified recipe, mechanical yield, safe human use. 1.3. Two-source reconciliation Source A, the 106-page Integrated Transparent Photonic Eyewear monograph (4 November 2026), develops a broad optical taxonomy and treats literal fibres, lithographically interleaved paths, and a passive-carrier/active-rim hybrid as separate realizations. Source B, the 119-page Lens-First Integrated Photonic Eyewear dossier (9 November 2026), makes a narrower physical choice: the full optical source/actuator/detector arrangement is integrated or bonded to the substrate of each lens. The latter is the controlling architecture in this synthesis. Active items can occupy an opaque lens-peripheral zone that does not block the declared clear aperture; moving such components into a distinct frame-mounted projector or imaging module would change the object under study. The alternative patterned-guide and hybrid routes are retained as comparators , not surreptitious replacements. A fixed-height stack with orthogonal drawn fibres preserves microscale fibres and separate crossing heights but removes the repeatedly bending over–under sequence. Lithographically deposited films preserve a crossed routing graph but are not literally interwoven drawn fibres. Their performance must be measured against the same outgoing display modes, incoming scene information, eye-return information, direct-view quality, and assembly yield. A better passive- lens substitute is scientifically informative even if the literal embodiment fails; its result must be labeled by the geometry actually fabricated. 1.4. Observable first-generation success The development order is P0, P1, P2, and P3. P0 is a passive fibre or patterned-guide coupon with measured mode confinement, repeated-crossing transfer, phase stability, and optical clarity. P1 is a single lens carrying independently verified D, C, and E paths. P2 is a registered pair of successful P1 lenses that demonstrates narrow-field disparity, different commanded field curvatures, pupil-position response, and two distinguishable scene viewpoints. P3 is an ensuing wearable and human-factors program sized from the measured P2 hardware. The first engineering result is not defined as a commercial full-field display. DOCUMENTARY RECONSTRUCTION | NO FABRICATED LENS 2 / 130 INTEGRATED LENSFIRST PHOTONIC EYEWEAR RESEARCH / ENGINEERING P0 Passive coupon Modes · crossings · clarity P1 One complete lens D + C + E, joint test P2 Matched optical pair Depth · disparity · pupils P3 Wearable engineering Power · thermal · human Evidence-gated physical development sequence Failure remains attached to the measured specimen; a document or simulation does not advance a physical gate. Figure 1.2.: Physical evidence required for each gate. Documentary completion and a component precedent do not advance a gate that requires measured integrated specimens. DOCUMENTARY RECONSTRUCTION | NO FABRICATED LENS 3 / 130 2. Electromagnetic foundations and photonic geometry 2.1. Guiding in a physical core, sheath, and carrier The optical path is governed by Maxwell’s equations and the electromagnetic boundary conditions at each material interface. In an ordinary dielectric guide, the mode exists because the material and cross-sectional geometry support an electromagnetic field that remains localized around the guiding region. The actual requirements depend on the core index, surrounding index, guide dimensions, wavelength, polarization, bends, and neighboring structures. Snell’s law gives an intuitive large-scale interface picture; a subwavelength microfiber requires a vector eigenmode calculation rather than an account using repeated geometric reflection alone. Placing silica directly against visible-index sapphire risks coupling the guided field into the higher-index host. A mechanically protective coating does not automatically provide optical isolation. The design therefore identifies the drawn-fibre core, its lower-index optical sheath, any adhesive or air gap, and the crystalline carrier as separately characterized materials. The sheath must preserve the guided field through installation, curing, humidity exposure, repeated crossings, and edge attachment. If the outer mechanical diameter is one micrometre, a core-plus-sheath mode model must fit within that diameter or identify a different physical size; “one micrometre” cannot simultane- ously name incompatible core and total diameters. The nominal candidate materials have distinct functions. Sapphire retains the original crystalline-carrier interpreta- tion and raises crystal orientation and birefringence questions. Fused silica is a useful comparative optical carrier but is amorphous. Silicon nitride and alumina are candidates for lithographic guiding layers; they are not drawn sil- ica microfibres. Lithium-niobate-based active regions can change a field’s phase electro-optically, but their reported chip-level operation does not select an optical bond or electrode process compatible with a curved transparent car- rier. A transparent adhesive that changes the mode after curing is an optical design failure even if the lens remains mechanically intact. Carrier Guided core Active control Package Sapphire crystal Silica witness Drawn silica SiN or alumina Electro-optic or thermal trim Coatings, bonds and edge seal Material choices have different optical functions and qualification tests A material precedent on an opaque silicon handle does not qualify a transparent assembled lens. Figure 2.1.: Carrier, mode-guiding material, active field-control element, and package have different jobs. Alternative material families require separate process identification. 2.2. The square network is not an independent-pixel array Seven geometries must be recorded separately: total lens aperture; square-network routing pitch; guide core and sheath dimensions; crossing elevation and transition length; locally emitting aperture diameter; emitter spacing within that local aperture; and electrode/actuator footprint. A 30 mm grid at an 80 μ m route pitch has 375 nominal horizontal positions and 375 nominal vertical positions, or 140,625 geometric intersections. These locations do not 4 INTEGRATED LENSFIRST PHOTONIC EYEWEAR RESEARCH / ENGINEERING imply 140,625 phase-addressable pixels, or even that each intersection emits light. A common phase shift applied upstream of a splitter changes a common state; it does not independently command every branch after the splitter. Define an input vector a of coherent guided modes and a measured complex linear response H at a specified wave- length, polarization, actuator state, and temperature. The outgoing field is e = Ha . The number of independently useful linear output patterns is bounded by the rank of this response, not by how many lines are visible in the pattern. Nonlinear response, active changes, wavelength or time scheduling can produce additional measured operating states, but must be described in the relevant expanded model. Reciprocal optical coupling exchanges appropriately paired source and receiver modes under its applicable material assumptions; it does not restore information destroyed by loss or by measuring only an intensity sum. A local tile with a very small emitter spacing may be fed by a coarse square routing grid. The required optical control count then follows the fields the tile must generate, not the route’s line count. This separation is the strongest physically coherent reading of the woven-routing idea. It preserves the requested mechanism without treating a transport grid as a high-resolution display or camera simply because its lines intersect. D: RGB source/phase C: world detector E: NIR source/detector Primary lens-first construction — plan view Interleaved grid is routing geometry. Every function requires distinct usable optical modes. Figure 2.2.: Reconstructed lens plan, showing three optically distinct path families and a lens-bound peripheral active zone. The diagram is schematic, and does not establish that the illustrated intersections possess independent optical control. 2.3. Woven crossing geometry and bending demand The originating construction calls for a continuous raised-and-lowered fibre centerline at selected orthogonal cross- ings. A finite vertical rise over a short horizontal interval creates curvature; the outside surface of a bent strand experiences longitudinal extension. For the deliberately simplified cosine centerline used in the dossier, the small- deflection geometric outer-fibre strain proxy is ε outer ≃ π 2 dh P 2 , (2.1) where d is total outer diameter, h is centerline rise, and P is the horizontal transition length. This expression estimates a local geometric demand. It does not incorporate nonlinear stress, coatings, defects, proof strength, residual tension, contact, ageing, or optical radiation loss. Applying it to a steep profile beyond its approximations requires the exact measured curvature rather than a confident fracture prediction. At an 80 μ m transition, a 4 μ m rise and 1 μ m outer diameter give 0.617% in this geometric model. At a 20 μ m transition with the same rise and diameter the proxy becomes 9.87%, making short repeated elevations an immediate test priority. The relevant first specimen is therefore not a decorative full-lens braid. It is a controlled family of straight reference guides, isolated crossings, and 25-, 50-, and 100-crossing cascades, measured before and after placement and cure. DOCUMENTARY RECONSTRUCTION | NO FABRICATED LENS 5 / 130 INTEGRATED LENSFIRST PHOTONIC EYEWEAR RESEARCH / ENGINEERING −40 −30 −20 −10 0 10 20 30 40 Horizontal position along crossing (μm) −2 −1 0 1 2 3 4 5 6 Vertical level (μm) 4 μm rise Trial cosine elevation: 80 μm pitch, 4 μm rise; not a released groove geometry Figure 2.3.: Proposed centerline elevation for a single over–under crossing. The horizontal spacing, rise, and fiber diameter are example inputs; actual profile and optical surroundings require measurement. 20 30 40 50 60 70 80 Square-grid routing pitch (μm) 0 10 20 30 40 50 60 Geometric outer-fiber strain proxy (%) 1% marker (not a strength limit) Repeated crossing curvature is sensitive to pitch and outer diameter Outer diameter 1 μm Outer diameter 3 μm Outer diameter 5 μm Figure 2.4.: Recomputed geometric bending proxy over routing pitch for three proposed outer diameters. The diagram is not a fracture-strength prediction. DOCUMENTARY RECONSTRUCTION | NO FABRICATED LENS 6 / 130 INTEGRATED LENSFIRST PHOTONIC EYEWEAR RESEARCH / ENGINEERING 2.4. Accumulated guided loss and crossings A useful full-route budget begins with measured straight-path propagation attenuation, adds the excess contribution of every encountered crossing, and then includes launching, splitting, modulation, outcoupling, pupil overlap and other paths. In a simple scalar logarithmic account, L route = L straight ℓ + N x L x , (2.2) where the two coefficients are measured in decibels per unit length and decibels per crossing respectively. A 30 mm route at 80 μ m row pitch encounters 375 perpendicular lines if every row is physically crossed. With a hypothetical attenuation of 1 dB/cm and crossing excess of 0.02 dB, the passive contribution is 3+7.5=10.5 dB. For a separately proposed passive allocation of 6 dB, the crossing term would need to average no more than 0.008 dB in that geometry. Neither coefficient is a measured property of the proposed product. This arithmetic explains why measuring only a successful isolated intersection cannot establish a useful full-length lens. Repeated bends may also change polarization and differential phase even when total intensity remains ac- ceptable. Compare cascades against matched straight references, record cross-family leakage, and measure the full response after bonding and encapsulation. The lowest-effort causal test is a loss and phase comparison for the same strand before placement, after placement, and after curing; these differences separate successive physical failure mechanisms. 20 30 40 50 60 70 80 Routing pitch (μm) 5 10 15 20 25 30 Passive-route loss (dB) 30 mm route, 1 dB/cm straight propagation, every row crossed 0.020 dB/cross (source target) 0.008 dB/cross (conditional allocation) 6 dB passive allocation, hypothetical Figure 2.5.: Conditional crossing-count losses. The 6 dB line is a planning allocation, while propagation and crossing values are taken from proposed source targets rather than actual lens measurements. DOCUMENTARY RECONSTRUCTION | NO FABRICATED LENS 7 / 130 3. Phase-controlled display and field delivery 3.1. Optical superposition and active control The Huygens–Fresnel superposition principle expresses the propagation of each compatible optical contribution to an observation region. Sources with a common optical carrier frequency and known polarization can interfere according to their complex amplitudes, optical path phases, and propagation geometry. The controller changes electrical actua- tor states on much slower time scales than the optical carrier oscillation; it does not need to clock the optical cycles themselves. A global phase shift common to all contributors does not change the time-averaged intensity pattern, but a differential branch phase can move an interference maximum or alter the shape of a delivered wavefront. For a local scalar array used only as an initial calculation, the angular field is E ( θ ) = N ∑ n =1 A n g n ( θ ) exp { ikx n [ sin θ − sin θ 0 ] + iδ n } , (3.1) where A n is the assigned amplitude, g n the emitter radiation pattern, x n the position, θ 0 the commanded direction, and δ n any residual phase error. This model assumes a common optical frequency and omits mutual coupling, material vector response and propagation through an eye. It can illustrate angular aliases; it is not a verified full-wave model of the woven cell. The illustrative 64-site results in Figure 3.1 show how a dense 0.275 μ m local spacing and a sparse 2 μ m spacing create different angular responses at 550 nm. The arrays also differ in physical width; the comparison is about emitter sampling, not matched-aperture resolution. A route pitch of 20–80 μ m would be even sparser if it were mistaken for an evenly spaced visible-light emission array. A directional element pattern, deliberately limited steering range, or aperiodic arrangement may reduce specific unwanted orders, but no technique creates an unlimited number of useful independently commanded fields. −40 −20 0 20 40 Observation angle (degrees) −50 −40 −30 −20 −10 0 Normalized scalar intensity (dB) Illustrative 64-site field at 550 nm, commanded 10° steering 0.275 μm local spacing 2 μm sparse spacing Figure 3.1.: Original vector plot of the evaluated scalar finite-array model. Element radiation patterns, coherent source stability, real couplers, and pupil optics are deliberately absent. 3.2. RGB spectral operation Full-color synthesis is composed of separately calibrated coherent contributions in the selected red, green, and blue bands. Wavelength-specific phase maps differ because the refractive index, accumulated optical phase, coupler re- sponse, and actuator effectiveness depend on wavelength. The three colors do not require stationary mutual co- herence for an ordinary RGB percept; image perception integrates their intensities. Sequential operation needs a 8 INTEGRATED LENSFIRST PHOTONIC EYEWEAR RESEARCH / ENGINEERING schedule that includes source switching and phase settling. Parallel operation needs additional independent routes or a characterized multiplexing network. These alternatives cannot be ranked from a monochrome engine’s catalog electrical power alone. The published 2025 RGB integrated phased-array experiment established simultaneous RGB beam control and a simple RGB image on its reported foundry platform [2]. The 2025 thin-film lithium niobate study established a 32-channel visible OPA with measured steering on a different substrate [3]. The 2024 metasurface-waveguide ex- periment achieved full-color multi-depth holographic near-eye results using a phase spatial light modulator and a separate waveguide system [1]. These are relevant mechanisms, not measurements of the present on-lens woven microfiber architecture. 3.3. A real virtual image and the moving pupil An image is not merely a set of light spots on the surface of a spectacle lens. The field arriving at the pupil must possess the angular distribution and curvature corresponding to the desired virtual scene. The human eye forms its own retinal image from that incident field. A single steered beam establishes directional control, but not useful image detail, resolved points across a field, color registration, or depth-dependent focus cues. An optical engine must also deliver that field over a region of pupil positions large enough for the defined eye-relief and permitted eye motion. The first controlled-aperture experiment should therefore measure point, two-point, line, and extended patterns with a model eye, and repeat the tests over commanded angles, phase maps, wavelength bands, and pupil locations. Report the point-spread function, brightness uniformity, depth-dependent best-focus response, unwanted orders, speckle, and exposure timing. A measured small viewing field is preferable to an unsupported wide-field claim. For an ideal quadratic-wavefront comparison between virtual depths of 0.5 and 2.0 m, the center-to-edge path differ- ence changes by approximately 0.147 nm across a 28 μ m-wide effective aperture. Across 5 mm, the corresponding ideal change is 4.6875 μ m. The first value shows why a tiny proof-of-principle tile cannot, by itself, claim humanly useful accommodation control. The second illustrates the demand on a larger coherent effective aperture; it does not predict focus performance without measuring the transmitted wavefront, pupil clipping, eye optics, and control noise. 0.0 0.5 1.0 1.5 2.0 2.5 Radius from aperture center (mm) 0 1 2 3 4 5 6 Paraxial center-to-edge path (μm) Ideal quadratic phase curvature across a 5 mm effective aperture Virtual depth 0.5 m Virtual depth 2 m Figure 3.2.: Paraxial optical-path curvature for two commanded virtual depths. The plotted 5 mm effective aperture is an ideal parameter study, not a fabricated display. 3.4. Interferometric amplitude, source splitting, and output power The system may use a common source divided among multiple branches, with local phase control and optional amplitude weighting. Equal division into 64 branches gives 18.06 dB less power in each branch before other effects, but does not destroy 18.06 dB from the sum of optical powers over all 64 branches. Whether the coherent fields reinforce at an intended pupil direction depends on their phase relationships, element pattern, and losses; power is redistributed angularly and spatially. Adding the per-branch equal-split number to a separately defined total delivered-power budget without treating all branches would double-count a redistribution. DOCUMENTARY RECONSTRUCTION | NO FABRICATED LENS 9 / 130 INTEGRATED LENSFIRST PHOTONIC EYEWEAR RESEARCH / ENGINEERING The available manuscript provides a hypothetical 1 mW die-output case with 1 dB launch loss, 10.5 dB passive trans- port, 1 dB control loss, 3 dB extraction, and 3 dB pupil overlap. These assumed stages sum to 18.5 dB, leaving approxi- mately 14.1 μ W in the accepted model pupil if they describe the applicable total incident field power. This value is not a selected brightness, retinal irradiance, exposure allowance, or real-lens measurement. Every real channel budget must record whether a value is individual-branch power, a coherent measured mode, a summed radiometric power, or a pupil-limited output. Source die Launch seam 30 mm route Field controls Outcoupling Pupil overlap 10 −2 10 −1 10 0 Optical power after stage (mW) 1.0000 0.7943 0.0708 0.0562 0.0282 0.0141 Illustrative source-to-pupil ledger from a 1 mW die Figure 3.3.: Recomputed conditional example of optical energy remaining after each specified stage. Use the model solely for bookkeeping until actual source, route, splitter, and pupil transfer measurements exist. DOCUMENTARY RECONSTRUCTION | NO FABRICATED LENS 10 / 130 4. Outward scene reception and computational imaging 4.1. Receiving functions and information content World-facing couplers intercept external radiance according to their angular, spectral, spatial, and polarization ac- ceptance and couple selected field components into guided modes. Photodetectors convert received optical energy to charge; readout transforms that charge into data; reconstruction then maps the data to a scene estimate. The route does not turn into a camera solely because it transmits light. If 16 photodiode outputs are all proportional to the same incident intensity pattern, the system has one effective measurement direction rather than 16 useful independent scene measurements. In a simple static linear intensity model, y = Hx + η , where x is a declared vector of scene angular intensities, y is the vector of detector readouts, H is a calibrated response matrix, and η accounts for noise and unmodeled effects. The rank is no larger than the number of independent readouts. Sixteen static scalar readings cannot uniquely determine an unrestricted 32-by-32 scene intensity vector containing 1,024 independent unknowns. Additional optical states, exposures, diversity, known scene structure, regularization assumptions, or other information may make a narrower reconstruction task solvable, but a successful regularized image is not evidence of an unrestricted full-resolution camera. A first P1 experiment may legitimately seek a full-rank 16-bin response over a narrow declared field and reconstruct previously unseen 4-by-4 calibration targets with logged errors and signal-to-noise levels. To avoid circular validation, use held-out patterns and separate response acquisition from scene reconstruction. Change a world target while holding the receiver state fixed, and block a world-facing p