The MODAVIS ecosystem

Multimodal Organological Data Analysis and Virtualization Systems connects acquisition, interpretation, public access and preservation. The framework coordinates components; it is not an additional acquisition application.

Establishes the relational structures that other components use. Runs in the research environment; a public web service is not offered.

Preserves source-native evidence and prepares documented transfers. Canonical reconciliation and projection belong to the Processor workflow.

Its twelve modules keep source observations, candidate assertions, review decisions and accepted projections distinct. The cohesive 0.2.0 package is a development candidate, not a verified public release.

Search results remain provisional. Prospector does not write directly to canonical research records.

Exposes research projections, provenance and citations. Public browsing is separate from administrative operations.

Release preparation includes metadata, rights and fixity checks. Packaging a candidate does not itself authorize publication.

An independent observer of explicitly permitted status sources. It does not execute jobs, repair services or change canonical data.

Connects a specialist biographical source to research inspection workflows. It is an internal application, not a publicly deployed discovery service.

Passing a transfer schema establishes structural validity; it does not constitute acceptance into the canonical corpus.

Retains original observations. Review and canonical projection remain explicit subsequent decisions.

Canonical projection requires owner authorization and frozen dependencies; it is not an unrestricted import shortcut.

Quotations and derived statements retain exact document evidence and provenance.

Preserves ambiguity, temporal qualifications and coordinate accuracy rather than silently treating a place guess as a precise location.

Uncalibrated scores remain qualified, and insufficient evidence can result in abstention.

Unknown dates, pages and other fields remain uncertain. Candidate parsing does not write canonical records.

Fetching is opt-in and source evidence is retained for replay and review.

Provenance-Linked Evidence Network for Uncertainty Modeling considers transmission and evidence independence. Its scores are not calibrated probabilities or universal source-trust ratings.

Reports pitch-control qualifications. It does not establish an instrument’s identity, acoustic range or performed sound.

Segmentation is class-agnostic and requires interpretation. A mask does not by itself classify a pipe, keyboard or stop.

Point clouds are a complete supported output. A textured mesh is not guaranteed. This photographic workflow is distinct from parametric pipe reconstruction in VAO-Pipe.

The 0.3.2 release adds independent audio analysis with waveform, spectrogram, pitch and partial views, plus documented exports. It supports VAO 0.5 export and a verified offline POD 1.6 projection; individual source identities remain explicit.

Map locations and source qualifications should be read together; geographic display does not imply survey-grade precision.

Available experiences depend on the package’s declared capabilities, the viewer version and browser support. Not every package offers playback, XR or simulation.

Retains image crops and raw OCR for manual review. Corpus matches are suggestions, not automatic substitutions. The archived 0.1.0 workflow uses POD 1.5; this is distinct from later POD releases.

Records rendering choices and paradata so that a derived audio file can be related to its MIDI and virtual-instrument inputs. No fixed public release was verified in this audit.

A web application and FastAPI backend developed around the Landshut Propstglocke case, including 3D Tiles. Modelled propagation and synthesis depend on declared assumptions; no fixed public software release was verified.

The report covers 18 open labial pipes from the Cuntz, Silbermann and Welte instruments. Measured dimensions drive procedural geometry. Passive eigenmode analysis is not a complete aerodynamic model of a sounding pipe.

A methods architecture and research roadmap. The thesis does not establish an implemented, end-to-end field-validated software pipeline.

A reproducibility and review environment for five backend components, not a deployment of the complete live research system.

The archived 0.2.0 release targets VAO 0.4. Newer 0.3.1 development supports the final 0.4 and 0.5 lines. Bounded remote retrieval, integrity checks and local caching remain distinct from depositing a release.

The current 0.6.0-rc.4 development line targets Unity 6, final VAO 0.4 and a pinned 0.5 candidate. Support is profile- and platform-dependent; this does not imply blanket compatibility with every final 0.5 package.

Supports final VAO 0.4 and a pinned 0.5 candidate. Modern package inspection is distinct from the fuller legacy interaction path for VAO 0.2.2.

A reference library and tool family accompanying the specification. VAOM and the independently versioned VAO CLI should not be treated as the same release.

An independent implementation embedded in the OrgRec project, not a separately verified application release.

Specifies layers, conformance profiles, bootstrap and preservation carriers, distributions and fixity. A standard version and a particular instrument-package version are separate identifiers.

Uses OWL, SKOS and SHACL to make identities, assertions and constraints explicit.

The cited deposit is the OMARO reference dataset, including Hornbostel–Sachs classifications. Its dataset release should not be confused with the entire MODAVIS Ontology Network.

The release contains 215,769 canonical organs. These are corpus counts, not a census of all surviving organs.

This is the older 1.5.0 preservation package. It does not establish a version-identical reviewer edition of public POD 1.6.0.

A curated research dataset about external software and algorithms; inclusion does not make those third-party projects MODAVIS software.

A research guide to source coverage and methodological differences; it is separate from the integrated POD corpus.

A systematic comparative research corpus, not a directory of software authored by MODAVIS.

Brings together a methodological taxonomy, systems catalogue and supporting research corpus.

Listed in the thesis publication register. The cited deposit was not retrievable through the public metadata registry during the 26 September 2026 audit, so no verified download is offered here.

An inventory and availability assessment; it does not grant rights to every listed sample set.

Connects the instrument’s documented structure and media in a versioned package; also contributes measured pipes to VAO-Pipe.

The package release is independently versioned from the VAO standard and contributes measured pipes to VAO-Pipe.

The Kinoorgel record supplies a versioned research object and contributes measured pipes to the pipe evaluation.

The thesis lists a data-package deposit, but its public metadata could not be verified during the audit. A download is not implied.

A documented research case rather than a separately verified public software release or VAO deposit.

A preceding toolkit in the development of the project’s deposition workflows.

Uses stateful .hdpc project files for replayable processing.

HPC-Link is an integration within this tool family. It delegates compute-intensive 3D and audio tasks; it is not an additional MODAVIS core component.

Retained as development references. The older standalone Ingestor is distinct from the current Processor Ingestor module; these helpers are not presented as maintained public applications.

Community membership is a discovery aid; cite a specific version record when reusing an artifact.

Discovery through a community does not replace verification of a package’s version, rights and fixity.