{"schema_version":1,"research_id":"eoa_inverse_innovation_exp05_external_evaluation_20260803","source_assessment_id":"invariant_mode_decomposition_design__art_aesthetics:P1:v0","cell_id":"invariant_mode_decomposition_design__art_aesthetics","search_queries":["multi projector art installation calibration seam color brightness artist intent calibration research","projection mapping automatic geometric photometric registration multi projector blending research paper","museum immersive installation projector calibration challenges color uniformity seams","multi-projector display color calibration perceptual uniformity ICC standards","site:tate.org.uk conservation projection installation calibration projector artwork colour","site:moma.org media art conservation projector calibration installation artwork","site:si.edu time based media art projection calibration conservation projector","museum projection artwork calibration artist intent time-based media conservation guidelines","eigenmode projector calibration iterative correction multi projector display","system identification iterative color calibration projector transition matrix eigenvalues","PCA multi-projector calibration color uniformity eigenvector","closed loop iterative photometric calibration projector camera feedback control","site:numpy.org/doc stable numpy linalg lstsq eig official documentation","site:docs.scipy.org scipy linalg eig least squares documentation","NIST least squares system identification eigenvalue uncertainty linear model guidance","FDA experimental design perturbation regression held out validation official"],"sources":[{"source_id":"S1","title":"Matters in Media Art","publisher":"The Museum of Modern Art","url":"https://production-gcp.moma.org/research/conservation/matters-in-media-art","source_class":"OFFICIAL_ORGANIZATION_DATA","publication_date":"n.d.","accessed_at":"2026-08-03","claims_supported":["Time-based media installations are complex systems posing custodial challenges.","Artists may specify installation methods and display technologies closely.","Museums identify needs for new skills, collaboration, and practical care guidance; MoMA, SFMOMA, Tate, and New Art Trust are identifiable institutional stakeholders."]},{"source_id":"S2","title":"Media Lab","publisher":"Smithsonian American Art Museum","url":"https://americanart.si.edu/art/conservation/center/media-lab","source_class":"OFFICIAL_ORGANIZATION_DATA","publication_date":"n.d.","accessed_at":"2026-08-03","claims_supported":["SAAM has an identifiable media-conservation function caring for video, film, and computer-based art.","Its conservator balances technical possibilities, work-specific particulars, and artist intentions to maintain artwork identity.","A museum media lab is a credible partner or adopter context, although it does not request modal calibration."]},{"source_id":"S3","title":"Multiple Projector Adjustment Overview","publisher":"Epson","url":"https://files.support.epson.com/docid/cpd6/cpd61935/source/adjustments/tasks/multi_projector_tasks/multi_projector_matching_overview_ebpu2113w_2220b.html","source_class":"OFFICIAL_PRODUCT_DOCUMENTATION","publication_date":"n.d.","accessed_at":"2026-08-03","claims_supported":["Established workflows match projectors automatically to a master or manually.","The workflow separately addresses color calibration or uniformity, brightness and tint, point correction, edge blending, and black level.","Epson warns that periodic light-source calibration may change color adjustments and that large geometry corrections may reduce image quality or complicate calibration.","Warm-up and fixed test patterns are recognized workflow controls."]},{"source_id":"S4","title":"VESA Updates Multiple Projection Common Data Interchange (MPCDI) Standard to Include New Color Correction Parameters","publisher":"Video Electronics Standards Association","url":"https://vesa.org/featured-articles/vesa-updates-multiple-projection-common-data-interchange-mpcdi-standard-to-include-new-color-correction-parameters/","source_class":"STANDARD","publication_date":"2015-06-01","accessed_at":"2026-08-03","claims_supported":["MPCDI standardizes interchange of multi-projector alignment, warping, blending, pixel-mapping, intensity, and color-correction data.","Version 2.0 supports 3D lookup tables and color mappings intended to mitigate inter-projector color, brightness, and intensity differences.","Standardized device-level correction is established prior art but does not describe correction-cycle eigenmodes or artist-scored outcomes."]},{"source_id":"S5","title":"Spatially Uniform Colors for Projectors and Tiled Displays","publisher":"German Research Center for Artificial Intelligence (DFKI) / Journal of the Society for Information Display","url":"https://www.dfki.de/en/web/research/projects-and-publications/publication/7296","source_class":"PRIMARY_RESEARCH","publication_date":"2007-06","accessed_at":"2026-08-03","claims_supported":["Spatial color non-uniformity within and between projectors is described as a major setup issue.","An iterative, camera-measured correction-table algorithm and inter-projector shared-gamut method were experimentally validated on 48 projectors.","Iterative multivariate photometric calibration is substantial adjacent prior art, but the reported objective is uniformity rather than artist-defined composition or modal stability."]},{"source_id":"S6","title":"Projector Warp and Blend: Complete Guide to Multi-Projector Calibration","publisher":"Scalable Display Technologies","url":"https://www.scalabledisplay.com/projector-warp-and-blend/","source_class":"COMMERCIAL_FIRST_PARTY","publication_date":"2025-01","accessed_at":"2026-08-03","claims_supported":["A commercial product already automates structured-light measurement, warp meshes, blend weights, per-projector color LUTs, verification, stored calibration, and drift correction.","The vendor identifies seams, brightness bands, color variation, aging, and drift as operational problems and identifies museums and entertainment installations as users.","The vendor reports approximately 45–90 minutes for initial calibration of 6–12 projectors and much longer manual technician time; these are first-party claims, not independent cost evidence."]},{"source_id":"S7","title":"Breaking the Scalability Limit of Multi-Projector Calibration with Embedded Cameras","publisher":"arXiv / IEEE-CVF CVPR authors","url":"https://arxiv.org/abs/2604.24024","source_class":"PRIMARY_RESEARCH","publication_date":"2026-04-27","accessed_at":"2026-08-03","claims_supported":["Contemporary research treats sequential projection-capture cycles and scaling with projector count as material calibration constraints.","A 2026 method estimates multi-projector parameters using embedded-camera measurements and reports accuracy comparable to conventional methods with nearly constant capture-cycle scaling.","Advanced multi-projector estimation is active prior art, but this paper addresses geometric parameter estimation rather than aesthetic proofing through correction-cycle modes."]},{"source_id":"S8","title":"Linear Algebra (scipy.linalg)","publisher":"SciPy Project","url":"https://docs.scipy.org/doc/scipy/tutorial/linalg.html","source_class":"OFFICIAL_PRODUCT_DOCUMENTATION","publication_date":"2026","accessed_at":"2026-08-03","claims_supported":["Maintained software directly supports linear least-squares fitting, residual outputs, eigenvalue/eigenvector decomposition, SVD, norms, and Schur decomposition.","The core numerical operations needed for a small transition-matrix prototype are readily implementable without novel numerical software.","Tool availability does not establish identifiability, adequate sample size, stationarity, eigenvector conditioning, or aesthetic validity."]}],"problem_evidence":{"support":"MODERATE","rationale":"Independent research, manufacturer documentation, a standard, and a commercial system all establish that multi-projector installations exhibit spatial color variation, brightness mismatch, seams, drift, and coupled calibration tasks. Museum sources establish that installed media works present technically consequential, artist-intention-sensitive stewardship problems. No opened source demonstrates the proposal's central prevalence claim: reproducible growing, reversing, or oscillating error combinations caused by successive technician correction cycles in art installations. The visible calibration problem exists; the proposed modal mechanism remains unverified.","source_ids":["S1","S2","S3","S4","S5","S6"]},"stakeholder_evidence":{"support":"MODERATE","rationale":"MoMA's multi-museum consortium and SAAM's Media Lab identify real custodians, conservators, technicians, artists, and institutions responsible for maintaining technically complex media works and artist intent. Epson and Scalable document operational calibration workflows used in museum or entertainment contexts. These sources support credible adopter and authorizer classes, but none expresses demand for transition-matrix eigendecomposition, modal sensitivity ranking, or this specific pilot.","source_ids":["S1","S2","S3","S6"]},"prior_art":{"proximity":"ADJACENT_PRIOR_ART","closest_analogues":[{"name":"Epson master-projector matching and manual multi-projector adjustment","similarity":"Uses test patterns and iteratively adjusts color uniformity, brightness, tint, point corrections, edges, and black level; it closely matches the proposal's coordinate-wise baseline.","remaining_difference":"It does not estimate a pass-to-pass state transition, classify eigenmodes by stability, rank corrections by protected aesthetic outcomes, or impose modal residual and spectral-gap vetoes.","source_ids":["S3"]},{"name":"VESA MPCDI color-correction and alignment standard","similarity":"Standardizes multi-projector warping, blending, intensity, color mappings, and 3D LUT interchange across systems.","remaining_difference":"It specifies correction-data representation rather than a method for identifying repeated-cycle modes or preserving artist-scored focal hierarchy.","source_ids":["S4"]},{"name":"Pagani-Stricker iterative camera-based color-uniformity calibration","similarity":"Uses measured, iterative correction to address within-projector luminance non-uniformity and derive a common multi-projector gamut; it demonstrates multivariate closed-loop calibration at scale.","remaining_difference":"Its optimization target is display uniformity, not the invariant directions and stability of technician correction cycles, and it does not include artist-defined protected outcomes or randomized preset review.","source_ids":["S5"]},{"name":"Scalable Display Manager automated warp, blend, color calibration, and drift correction","similarity":"A deployed first-party product measures projector responses, computes warp/blend/color corrections, verifies results, stores calibration, and supports recalibration after drift.","remaining_difference":"The disclosed workflow directly corrects geometry and photometry; it does not disclose eigenmode analysis of correction-cycle dynamics or rejection based on aesthetic sensitivity, mode coupling, reconstruction residual, and spectral separation.","source_ids":["S6"]},{"name":"Embedded-camera simultaneous multi-projector calibration","similarity":"Uses controlled projection and capture to estimate multi-projector parameters and explicitly reduces calibration-cycle burden.","remaining_difference":"It estimates geometric parameters for scalable calibration, not local correction-cycle dynamics or artist-authorized aesthetic proofing.","source_ids":["S7"]}],"distinctive_claim_remaining":"For one fixed artwork and installation regime, a low-dimensional linear transition estimated from replicated, bounded calibration perturbations will contain resample-stable action-relevant modes, and selecting one reversible correction by modal sensitivity will predict held-out signed changes and preserve preregistered seam and focal-hierarchy outcomes at least as well as coordinate-wise correction; the claim fails if an independent-coordinate model matches or beats prediction with lower complexity, modes rotate or are ill-conditioned, no spectral separation exists, structured residual exceeds tolerance, or blinded aesthetic review is worse.","confidence":"MODERATE"},"implementation_evidence":{"support":"MODERATE","rationale":"Existing workflows provide test patterns, camera measurements, per-zone corrections, reversible settings, stored calibration, and recalibration, while SciPy supplies the required least-squares and decomposition routines. Technical construction is therefore plausible. Feasibility is not yet established for the proposed state dimension and one-session sample budget: rank deficiency, warm-up drift, non-normality, near-degenerate eigenvalues, correlated measurement error, multiple viewing positions, and variable artist ratings could invalidate the model. Workflow and authority are manageable if the artist approves bounds and presets remain nonpublic and reversible. No special legal barrier was found; equipment safety still depends on vendor limits and venue procedures.","source_ids":["S1","S2","S3","S5","S6","S8"]},"scores":{"meaningful_impact":{"score":3,"rationale":"Avoiding compositionally damaging recalibration could matter for technically complex projection artworks, but affected prevalence and achievable effect size are unmeasured.","source_ids":["S1","S2","S3","S6"]},"stakeholder_pull":{"score":2,"rationale":"Institutions express need for technically faithful, artist-intention-sensitive care, but no source requests modal proofing or commits resources to it.","source_ids":["S1","S2"]},"incremental_advantage":{"score":2,"rationale":"The approach adds diagnosis and falsification around iterative dynamics, but established automatic calibration already provides fast measurement, correction, verification, and drift handling; comparative benefit is unknown.","source_ids":["S3","S5","S6","S7"]},"distinctiveness_plausibility":{"score":3,"rationale":"The eight-source bounded search found adjacent calibration methods and products but no direct match for correction-cycle eigenmodes combined with protected aesthetic outcomes and modal vetoes. This is not a world-novelty finding.","source_ids":["S3","S4","S5","S6","S7"]},"technical_implementability":{"score":3,"rationale":"Instrumentation, reversible controls, least-squares fitting, and eigendecomposition are available, but identifiability and stable modal interpretation under a small field sample are unresolved.","source_ids":["S3","S5","S6","S8"]},"adoption_authority_feasibility":{"score":4,"rationale":"Artist approval plus technician execution and museum constraints form a credible authority structure consistent with museum practice; an actual partner and approval have not been secured.","source_ids":["S1","S2"]},"evidence_readiness":{"score":2,"rationale":"The comparator and falsifiers are unusually explicit, but the decisive evidence requires a live installation, repeated measurements, proprietary presets, and human aesthetic review.","source_ids":["S3","S6","S8"]},"safety_net_benefit":{"score":4,"rationale":"Duplicated presets, nonpublic testing, bounded perturbations, residual vetoes, and rollback to the approved preset strongly limit aesthetic and operational harm, assuming equipment limits are enforced.","source_ids":["S1","S2","S3"]},"scalability":{"score":2,"rationale":"The method requires artwork-specific re-estimation, artist participation, repeated perturbations, and controlled conditions; modern calibration research and products show technical scaling is possible, but not for this aesthetic workflow.","source_ids":["S6","S7"]}},"score_confidence":"MODERATE","costs":{"first_evidence":{"band_2026_usd":"10K_TO_50K","scope":"One preregistered, nonpublic single-installation session covering protocol design, a reduced state vector, technician and artist time, camera/color measurement, replicated perturbations, held-out analysis, randomized duplicated-preset review, and reporting.","confidence":"LOW","assumptions":["Existing projectors, control software, venue, and approved reference media are already available.","Two to four specialist person-days plus preparation and analysis are required.","Measurement equipment is already owned or rented; no projector purchase is included.","The estimate is resource-equivalent, not a vendor quote."],"source_ids":["S3","S6","S8"]},"initial_deployment_startup":{"band_2026_usd":"50K_TO_250K","scope":"Develop a reusable prototype, measurement schema, safe-control interface, resampling and conditioning diagnostics, preset/rollback tooling, documentation, and validation on two or three installations at one institution.","confidence":"LOW","assumptions":["A part-time scientific programmer, projection technician, conservator or producer, and participating artists are needed over several months.","Commercial calibration components may be integrated but their undisclosed license pricing is excluded or represented as labor-equivalent contingency.","No new permanent projection array or gallery construction is included."],"source_ids":["S1","S2","S3","S6","S8"]},"operational_launch":{"band_2026_usd":"50K_TO_250K","scope":"Institutionalize the workflow for a venue or touring production, including staff training, standard operating procedures, measurement hardware, integration with calibration presets, governance review, and several supervised launches.","confidence":"LOW","assumptions":["Launch covers an existing small-to-medium projector fleet rather than a new immersive venue.","Artist review remains project-specific.","The range includes integration and contingency for failed or repeated sessions but excludes projector replacement and construction."],"source_ids":["S1","S2","S3","S6"]},"annual_recurring":{"band_2026_usd":"10K_TO_50K","scope":"Periodic recalibration checks, analyst and technician time, software maintenance, sensor verification, documentation, and one to four artwork-specific proofing exercises per year.","confidence":"LOW","assumptions":["Existing equipment remains serviceable.","Re-estimation is required after meaningful hardware, geometry, content, or lighting changes.","Commercial first-party claims indicate routine recalibration can be short, but modal data collection and artist review add substantial labor.","No public-program downtime or equipment depreciation is monetized."],"source_ids":["S2","S3","S6"]}},"verified_pipeline_gates":{"externally_supported_problem":{"status":"YES","reason":"Multiple independent and official sources establish visible color non-uniformity, seams, drift, calibration complexity, and artist-intention-sensitive installation requirements. The narrower modal-cycle mechanism is still an evidence gap rather than a prerequisite for recognizing the underlying problem.","source_ids":["S1","S2","S3","S5","S6"]},"externally_credible_adopter_or_authorizer":{"status":"YES","reason":"MoMA's consortium and SAAM's Media Lab identify museum conservators, technical managers, artists, and custodial institutions with relevant responsibilities. They are credible adopter or authorizer classes, although none has agreed to this pilot.","source_ids":["S1","S2"]},"distinct_testable_incremental_claim":{"status":"YES","reason":"The candidate can be contrasted against coordinate-wise and automatic calibration through held-out prediction, model complexity, mode stability, residual structure, seam metrics, and randomized artist review, with explicit falsifiers.","source_ids":["S3","S5","S6","S8"]},"bounded_next_evidence_step":{"status":"YES","reason":"A single nonpublic installation, reduced state vector, reversible controls, held-out trials, two comparators, and preregistered stop rules bound the next step.","source_ids":["S1","S2","S3","S8"]},"no_unresolved_safety_or_authority_stop":{"status":"YES","reason":"The artist can retain aesthetic authority while technicians enforce equipment limits; duplicated presets, nonpublic testing, small perturbations, immediate halt conditions, and restoration of the approved preset provide a credible safety structure. Site-specific approval is still required before fieldwork.","source_ids":["S1","S2","S3"]},"credible_cost_scope_and_range":{"status":"UNCERTAIN","reason":"The scopes and exclusions are explicit, and a first-party source supplies calibration-time anchors, but no opened source gives independent labor, software-license, measurement-hardware, or museum-overhead pricing. The bands are planning estimates with low confidence.","source_ids":["S3","S6"]}},"next_evidence_step":"Partner with one artist and one museum or production venue for a preregistered, nonpublic pilot. First reduce the state to no more than 8–12 outcomes and controls so replicated perturbation trials outnumber fitted parameters; fix two scenes, one measured viewing position, ambient light, projector warm-up, safe bounds, and an untouched approved preset. Randomize replicated small perturbations and reserve at least 25% as held-out trials. Compare (A) the proposed regularized transition model with bootstrap mode-direction and eigenvalue intervals, (B) an independent-coordinate transition model, and (C) the venue's ordinary technician correction. Primary technical endpoints are held-out signed-change error, structured residual, condition number, resample mode-angle stability, and retained-to-dropped spectral separation. Only if the modal model beats B by a preregistered prediction margin and passes all conditioning and residual rules should the highest-ranked modal correction and ordinary correction be rendered as randomized duplicated presets for blinded artist review and a small proofing group. Falsify further development if baseline cycles show no reproducible coupled structure, B performs equivalently with lower complexity, individual modes rotate or are ill-conditioned, the spectral gap is absent, protected seam or focal outcomes worsen, or the artist rejects the modal preset. The study decides only whether a larger within-installation replication is warranted.","blocking_evidence":["No external source establishes that repeated manual calibration of multi-projector artworks commonly produces stable growing, reversing, or oscillating coupled error modes.","No partner installation, artist authorization, technician protocol, equipment limits, or proprietary preset access has been secured.","The number of safe perturbation trials needed for an identifiable transition matrix is unknown relative to available proofing time and state dimension.","Measurement repeatability after warm-up and ambient control, including black-level and seam measurements, is unverified for the target installation.","No live data establish eigenvector conditioning, spectral separation, stationarity, residual tolerance, or mode stability across resamples.","No blinded comparison establishes that a modal correction preserves artist-defined focal hierarchy or seam visibility as well as existing practice.","Independent 2026 cost and licensing evidence is absent."],"research_disposition":"PARTNERED_RESEARCH_PROGRAM","world_novelty_boundary":"This evaluation found adjacent prior art, products, workflows, and a multi-projector standard within an eight-source bounded search. It did not measure world novelty, patentability, freedom to operate, market size, realized impact, or exhaustive literature and product coverage. Absence of a direct modal-art-calibration match in these sources is only a remaining contrastive hypothesis, not a novelty conclusion.","arm":"COMPLETE_PROPOSAL_PORTFOLIO","candidate_version":0,"controller_recommendation":{"action":"STOP_EMPIRICAL_RESEARCH_NEEDED","repairable":false,"material_progress_observed":true,"progress_targets":["Secure one artist-authorized, nonpublic installation and document venue and equipment constraints.","Preregister a dimensioned experimental design whose replicated perturbations support identification and held-out testing.","Measure whether baseline calibration cycles contain reproducible coupled structure before fitting or interpreting eigenmodes.","Benchmark against both an independent-coordinate model and ordinary or automatic calibration using held-out prediction, complexity, seam metrics, and blinded aesthetic review.","Quantify conditioning, mode-angle stability, spectral separation, structured residuals, warm-up drift, and rollback performance.","Obtain independent labor, hardware, and software-license quotations to replace low-confidence cost bands."],"reason":"Bounded web research supports the underlying calibration problem, credible institutional actors, implementable mathematics, and substantial adjacent practice, but cannot answer the decisive questions: whether stable correction-cycle modes exist in an artwork, whether they are identifiable from safe trials, and whether modal selection improves or at least preserves protected aesthetic outcomes relative to simpler methods. Those questions require live field measurements, proprietary preset access, and artist review, so further web search cannot close the evidence gap."},"proposal_index":1}