spinorama.org is a free, independent database of loudspeaker and headphone measurements. It collects measurements of 1000+ speakers from many sources, puts them in one place, and presents them all in the same standardized form so that speakers can be viewed, compared, and analysed on equal terms.

The goal is simple: help you choose a speaker from measured evidence instead of relying on commercial information or internet buzz. It can help you make an informed decision when buying a speaker — and there are enough measurements to do statistical analysis if you want to.

A spinorama is a standardized set of loudspeaker frequency responses intended to make directivity and tonal behaviour easier to compare. The method is defined by the US CTA in its Standard Method of Measurement for In-Home Loudspeakers (the CEA2034 spinorama).

Each spinorama gives you a few curves:

• On Axis — the response directly in front of the speaker. A smooth response is generally easier to equalize.
• Listening Window — an average of responses near the on-axis direction, representing small changes in listening position better than a single trace.
• Early Reflections — an estimate of the first reflected energy in a typical room, usually smoother and sloping downward with frequency.
• Sound Power — an average over the full measurement sphere.
• ERDI and SPDI (Directivity Indices) — the differences between the Early Reflections (or Sound Power) curve and the Listening Window. Smooth, predictable directivity is generally easier to equalize; equalization cannot repair large directivity errors, because it cannot make off-axis sound follow the on-axis correction.

For a walkthrough, see Understand Spinorama measurements in the documentation.

• Browse and search speakers and headphones by brand and model, and filter and sort them by, among others, price, shape/type, and measurement quality.
• View a full measurement set for each speaker — spinorama curves, directivity and other derived graphs, pictures, metadata, and a link back to the original review or source.
• Compare speakers side by side, curve against curve, instead of comparing graphs drawn to different scales by different reviewers.
• Use summary numbers as a first filter, in particular:
  • a preference score, an aggregate computed from the spinorama data where higher generally indicates a more preferred response under the model;
  • bass extension (LFX), an estimate of where the sound-power response falls below a reference level — lower is generally better;
  • flatness, the variation in dB around the average on-axis SPL between 300 Hz and 3 kHz — lower is better.
• Download the data and generated EQ — measurement data, graphs, and auto-generated equalization (AutoEQ) settings for speakers and headphones, ready to load into common EQ/DSP tools.
• Query everything programmatically through the read-only REST API (interactive OpenAPI docs at api.spinorama.org/docs).

The measurements are evidence, not a guarantee, and the summary scores are a starting point, not a verdict:

• A preference score does not describe distortion, compression, maximum output (SPL), or reliability, and small score differences should be treated cautiously.
• Measurements come from different rooms, equipment, and processing; always read the source and quality fields on a speaker page together with the graphs, and compare like with like.
• What you hear also depends on your room, placement, listening distance, and how loud you listen. No spinorama predicts every room.

Spinorama does not measure most speakers itself; it aggregates and normalizes measurements published by others, each with its origin, format, and quality recorded in the metadata:

• independent reviewers, notably Audio Science Review and Erin's Audio Corner — please consider supporting them;
• research sources such as the Princeton 3D3A directivity database;
• manufacturer and vendor data, including GLL files (a loudspeaker data format developed by AFMG) from brands such as Genelec, Neumann, JBL, Meyer Sound, d&b audiotechnik, and many others. Vendor data may be smoothed, simulated, or measured — the individual measurement's metadata is the source of truth.

Roughly, high quality means a Klippel NFS or large anechoic-chamber measurement, medium a smaller or less controlled setup, and low vendor or heavily processed data or measurements without a fully anechoic setup.

Everything that builds the site is open source:

• spinorama-org/spinorama — the main repository: the Python library to load, compute, score, and plot spinoramas (including computing graphs from your own measurements), the website generator, the metadata, EQ generation (AutoEQ), and the REST API. Note that raw measurements and real-data fixtures live in a private submodule, so cloning the public code repository alone does not grant access to that data.
• Documentation — spinorama.org/docs covers reading the graphs, installing the software, computing spinoramas from your own measurements, adding metadata, digitizing a graph image, and using the REST API.
• API — api.spinorama.org exposes speaker and headphone metadata and precomputed measurements as JSON.

This .github repository only holds organization-wide GitHub profile and community files; the code and its issues live in the spinorama repository above.

• Found a wrong price, a broken graph, or a missing speaker? Open an issue in spinorama-org/spinorama.
• Have original measurement files to share? Preserve them in the repository's measurement layout and record the origin, format, quality, and review URL — see the documentation. Please do not redistribute copyrighted source files without permission.
• Want to work on the code? Start from the development setup in the spinorama README (Python 3.12, Node.js, and optionally Rust/cargo + maturin for the scoring extension).

• Software: GNU GPLv3 (see LICENSE in the spinorama repository).
• Measurement data: unless otherwise stated for a specific measurement, data distributed by spinorama.org — repository data, website downloads, and API responses — is licensed under Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0). Credit the original measurement creator and source, link to the license, and indicate changes. Commercial use requires separate permission from the relevant rights holder(s). Measurement-specific licenses and source notices take precedence.

Links: Website · Documentation · API · Main repository · Issues · Contact: pierre@spinorama.org