Global Discovery P2 · engineering evidence

Publish the method first. Publish numbers only when the capture can be reproduced.

This Evidence Lab separates listening preference from technical evidence. It provides the measurement protocol, level-matched A/B method, deterministic browser-generated test signals and reusable logs needed to challenge or reproduce a result.

Generate test WAVs Read the measurement method

Evidence status

What is published now—and what is deliberately not claimed yet.

Measurement protocol

Published

Reference route, level, sample rate, capture conditions, peak checks and interpretation limits are documented before any numerical result is promoted.

Level-matched A/B method

Published

Normal-impact listening and Gain Match verification are treated as different questions, with a reusable log for recording the result.

Synthetic test-signal generator

Published

1 kHz sine, logarithmic sweep, deterministic pink noise, stereo channel and polarity checks are generated locally as WAV files with a conservative default peak ceiling.

Raw product measurement datasets

Publish after capture

No frequency-response, distortion, loudness or latency number is promoted here until the source, route, software version and measurement conditions are recorded.

Video evidence markup

Waiting for a real video

No VideoObject structured data is emitted until an official public demonstration with a stable URL and accurate metadata actually exists.

External directory presence

Verifiable

ArSonKuPik has an external KVR Audio product listing. Directory metadata can lag the official release, and the listing is evidence of third-party directory presence—not an endorsement or independent review.

Evidence rule: an attractive graph is not enough. Every numerical claim should identify the exact release, source signal, level meaning, routing, sample rate, controls, comparison state and measurement tool. If those details are missing, treat the number as incomplete.

Step 1

Generate a controlled source

Use the free browser-side generator for repeatable synthetic WAV signals.
The default is -24 dBFS peak ceiling; it is not a calibrated SPL, LUFS or RMS reference.
No generated audio is intentionally uploaded by the tool.

Step 2

Record the complete signal path

Log software version, host or Standalone route, sample rate, buffer, preset and every relevant control.
Capture a reference before the processor when possible.
Keep operating-system enhancement and interface-gain conditions explicit.

Step 3

Separate impact from level bias

Listen normally first if the question is whether you prefer the intended enhancement.
Use Gain Match when the question is tonal preference at a more comparable level.
Do not turn one listening pass into a universal claim.

Step 4

Publish limits with the result

Frequency plots do not fully describe transient or spatial perception.
One interface, one host or one song does not establish universal performance.
Results should be reproducible by someone who did not create the product.

Downloadable reproducibility kit

Use the same fields every time you test.

CSV templates make it harder to forget the details that turn a screenshot into a reproducible engineering record. They contain headers only; no result is pre-filled or implied.

Measurement log CSVSignal path, version, controls, levels and objective measurement fields.A/B listening log CSVTrial order, preference, confidence and listening observations.

Open the WAV test-signal generator Read the test-signal manifest External KVR Audio directory listing Official release remains authoritative

Evidence routes

Choose the question you need to answer.

Measurement methodSignal route, level, sample rate, safety and interpretation limits.Fair A/B listeningSeparate intended impact from loudness-biased judgement.Free WAV test signalsGenerate sine, sweep, pink noise, stereo and polarity checks locally.Project evidence policyAuthorship, distribution, claims and official-source boundaries.