Bluetooth Latency Test
Acoustic mode plays a short chirp through your selected system output and listens for it with your microphone. Use speakers or place an earbud near the microphone at a low comfortable volume. The result includes output, air path and input delay; it does not isolate Bluetooth latency. Echo cancellation/noise suppression are requested off but may be imposed by the device. Use the same physical setup when comparing paths.
Browser-reported mode needs no microphone. Its processing and output estimates are shown separately and are not a round-trip measurement. Unsupported values are reported as unavailable.
Use the controls to calculate or measure a result.
Related tools
- Sound Latency Test — Measure and interpret complete acoustic paths or separate reported timing.
- AV Sync Test — Choose a subjective visual offset relative to estimated audio presentation.
Compare the timing of manually selected wireless and wired audio paths. Bluetooth Latency Test offers acoustic round-trip probes and separately labelled browser-reported timing. Its A/B workflow helps you describe a path change without pretending to isolate a codec’s delay.
Connect the device and select it as the system output before measuring. The tool does not pair headphones, choose a Bluetooth codec or automatically route audio to the device named in your label.
Build a useful wired/wireless comparison
- Choose a baseline route, such as built-in or wired output, while keeping the intended microphone fixed.
- Arrange for the output to reach that microphone at low comfortable volume.
- Select Acoustic round-trip probe, enter a clear path label and choose one or three repetitions.
- Measure successfully and select Save comparison A.
- Switch to the wireless route, preserve the input and geometry where possible, update the label and measure again.
- Select Compare as B and read B minus A alongside both medians.
An earbud may need to be near the microphone for an acoustic loop. Testing headphones while wearing them normally can leave an external microphone unable to hear the chirp. A failed match does not mean the wireless device has zero delay.
Understand what the milliseconds include
The acoustic result contains output processing, sound travel and microphone capture. Wireless transport can be part of that path, but it is not the only part. Browser and device processing can influence the measurement too.
Suppose baseline A has a median of 82 ms and wireless path B has a median of 146 ms. The reported difference is +64 ms for the two tested round trips. That is a path comparison, not a direct measurement that a codec alone adds 64 ms.
If switching to Bluetooth also changes the active microphone or device operating mode, you have changed more than output transport. Record that change rather than treating the experiment as a controlled codec comparison.
Repeat before trusting a single observation
The tool offers one or three probes per run. Three successful observations provide a median, mean, minimum, maximum and spread. One successful probe provides no evidence about run-to-run consistency, even when its displayed spread is zero.
Use the same repetition count for A and B when possible. If results fluctuate, repeat the unchanged route and inspect the setup. Avoid selecting one unusually small value as the device’s definitive latency.
Correlation values describe matching the short chirp to microphone samples. They are not codec-quality scores or confidence percentages. A reliable match still describes the whole captured path.
Use browser-reported mode as separate evidence
Browser-reported mode does not require microphone permission. It shows base/processing and output-device estimates when the browser exposes them. An unsupported value is unavailable, not a measured zero.
These figures can change with route and browser, but they are not an independently measured wireless round trip. Keep their method label when writing them down. Do not combine a reported number from one route with an acoustic number from another and call the difference input latency.
The Sound Latency Test explains those distinctions in more detail. This Bluetooth page uses the same acoustic foundation with a focused comparison workflow; it is not an independent laboratory validation.
Check the symptom you actually notice
For a rhythm game or instrument-monitoring task, the response to a new action matters. For a video, a player may use its own audio/video compensation. A video that looks aligned does not establish that every interactive application has the same response timing.
Use the AV Sync Test to compare a flash with a click by eye and ear. Its offset is a subjective alignment in that page, not an automatic measurement of Bluetooth transport.
Recheck the application where the problem occurs. A browser test cannot promise the same timing as a different audio API, game engine or recording application, and it does not change their settings.
Keep a path record that can be repeated
Record output, input, connection, browser and physical arrangement. Update the path label before the next measurement, not after saving a result from a different route. A label describes your setup; it does not prove which hardware the system actually used.
Copy result and Download CSV save numeric information rather than an audio recording. A/B data is held in the current session. Preserve a meaningful summary if you want to compare the same setup later.
The microphone probe uses temporary in-memory samples and needs HTTPS, permission and AudioWorklet support. It requests microphone processing off, but devices may impose it. Stop, hide or leave the page to end active measurement. No microphone audio is uploaded for server analysis by this tool.
Frequently asked questions
Can I detect SBC, AAC or another codec here?
No. Delay and a user-entered label do not establish the active codec. The page does not inspect or choose codec settings.
Does B minus A give Bluetooth-only latency?
It gives a difference between complete measured paths. Other changes in input, processing or routing can contribute, so it should remain labelled a round-trip difference.
Why does the chirp test fail with headphones?
The selected microphone must hear the selected output. Closed headphones worn normally may not form that loop. Check routing and arrangement at comfortable volume before repeating.
Will a low reported estimate guarantee no gaming delay?
No. Browser reporting and a game’s interaction path are different evidence. Confirm the practical response in the application and configuration you intend to use.