Download audio test files for a repeatable speaker or headphone check

If the device you need to check cannot open WaveLocus, download a portable audio test pack on a browser that can. The pack contains generated stereo WAV signals, a versioned recipe, a manifest, a short README, and SHA-256 checksums. Play the same file on each device at a comfortable volume, record the full output path, and change one thing at a time. Matching files make the source repeatable. They do not make two devices acoustically identical.

Know when a portable file helps

A local browser tool is convenient for an immediate test, but a comparison may involve a media player, another computer, or an output that cannot run the site. A downloaded reference travels with you. It is useful for checking channel routing, an audible interruption, or whether a problem follows a particular playback path. It is not a calibration file, a hearing test, or an electrical measurement of the device that plays it.

The current search results offer many downloadable tones and channel files. A file alone leaves important questions unanswered: which signal settings were used, whether the left and right channels are distinct, and whether a later copy is exactly the same file. WaveLocus's portable test-pack tool includes those settings and integrity information beside the WAVs. That extra context matters when you return to the test weeks later or share a reproduction with support.

Create the pack in your browser

1. Choose one job. Select a built-in recipe that fits your question, such as a left/right reference, or a local custom recipe you have already safety-checked in Recipes.
2. Choose a sample rate. The export tool currently offers 44.1 kHz or 48 kHz. Record your choice rather than assuming every player will use the file unchanged.
3. Download the ZIP. WaveLocus renders the recipe locally and creates 16-bit stereo PCM WAV files. It does not include microphone capture or personal recordings.
4. Inspect the contents. Keep the WAVs together with recipe.json, manifest.json, README.md, and SHA256SUMS.txt. Do not pass around only a renamed WAV if the method needs to be reproduced.
5. Play one file at low physical volume. Confirm the intended output and normal listening position before raising the volume to a comfortable level.

WaveLocus uses the browser's offline audio rendering path to produce the signal files. MDN's OfflineAudioContext documentation describes how this API renders audio into a buffer rather than a live output. The resulting WAV is a generated reference, not a captured recording of your equipment.

What is in the ZIP and why

A SHA-256 checksum is an integrity check, not proof that the player reproduced the waveform perfectly. If a copied WAV has the same digest, its bytes match the original checked file. The checksum does not verify system volume, resampling, Bluetooth processing, cable wiring, or what came out of a speaker. MDN's SubtleCrypto digest reference explains the browser hash operation used where Web Crypto is available.

Compare two devices without inventing precision

Use the same WAV and note the file name. Keep the listening position, headphones or speakers, and player controls as similar as practical. When switching devices, record which device, player, connection, endpoint, and physical volume setting changed. Do not expect two different volume controls to produce equal sound pressure at the same displayed number. The digital level recorded in the manifest is dBFS, not physical dB SPL. The dBFS guide explains that boundary.

For a left/right question, play the left-only and right-only files separately. Listen from the normal position and write down which physical side responds. If a channel swaps on one path but not another, you have a useful contrast. You have not yet proven whether the player, operating system, connection, or endpoint made the change. Run a controlled swap in guided diagnosis before naming a faulty component.

For an interruption or distortion, repeat the same short file a few times on each path. Note whether the event occurs at the same point or moves between plays. A pattern that repeats in one path and not another can guide the next test. Do not interpret a listening file as a precise latency, frequency-response, or dropout measurement. Those claims need a suitable return path and an eligible measurement method.

Handle file conversions and sharing carefully

If a player imports and converts the WAV to another format, your new file is not byte-identical to the packed reference. Record the conversion, or use a player that accepts the original WAV. Copying a whole ZIP is easiest when you need its method record. If a support team requests evidence, share the generated test pack only when appropriate and describe your observed result separately. The pack contains generated signals and recipe settings, not a recording of the fault or a diagnostic verdict.

Do not use an unfamiliar adapter, electrical loopback cable, or high playback level just to make this comparison. If the signal is uncomfortable, stop. A pack can make a source reproducible while the device and room remain uncontrolled. It cannot certify a speaker, establish hearing thresholds, or prove that two playback chains deliver the same sound.

Next step

Open portable test packs, export one relevant recipe, and keep the ZIP intact. After testing, write one clean comparison: the same named WAV, the two documented paths, and the single variable you changed. Continue in guided diagnosis if the problem follows one path.

Sources and method

This guide reflects WaveLocus's current versioned recipe and local ZIP exporter. MDN documents offline audio rendering and SHA-256 digest support. The article does not claim that a deterministic file guarantees identical acoustic playback, calibrated level, or a hardware diagnosis.