START HERE
Record a PC reference output and the phone output on the same recording timebase, then compare matching audio events. This measures playback mirroring lag relative to the PC output. Network RTT and buffer counters are different measurements. Just want a quick ear-and-eye Bluetooth or headphone delay check right now? Open the free in-browser audio sync & delay check.
Define the number before collecting it
“Latency” can describe several intervals. A useful result names its start and end points, the route, and the measurement method. Otherwise a precise-looking number may answer the wrong question.
| Metric | What it tells you |
|---|---|
| Playback mirroring lag | How much later the phone output arrives than a documented PC output |
| Network RTT | Round-trip time over the selected network/media path |
| Playback buffer | Time represented by a particular queue, not the entire route |
| Microphone return delay | Time from physical microphone input to the chosen Windows receiving boundary |
A PC output has its own buffering. Comparing it with a phone gives a relative result, which can even be negative for some reference paths. It is not automatically total capture-to-ear latency. Dividing RTT by two does not turn it into a measured one-way audio delay.
1. Capture both outputs on one clock
- Use a repeatable, distinctive test signal with identifiable events. Avoid relying on a continuous repeated tone whose cycles are hard to match.
- Record the PC reference and phone output simultaneously on separate channels of one audio interface or another validated shared-timebase setup.
- Document the reference output and both recording channels. If using microphones, record their positions and account for sound travel and channel skew.
- Keep source level moderate and avoid clipping. Keep the same geometry for every comparison.
Do not directly subtract timestamps from independent PC and phone clocks without a validated synchronization method. Starting two recorder apps at roughly the same time is not synchronization. Electrical connections also need appropriate inputs and levels; do not connect amplified speaker outputs to microphone inputs.
2. Compare matching events
Find the same event in both recorded channels and preserve the direction of the offset. For a sample-based measurement, the conversion is offset in milliseconds = sample offset ÷ sample rate × 1,000.
For example, an artificial 2,400-sample offset at 48,000 samples per second equals 50 ms. That is an arithmetic example, not a MicDrop result. A correlation tool should be checked against known artificial offsets before using it to analyze a real recording.
Measure many events over time, not just the best-looking pair. Report sample count, typical delay, spread, larger outliers, and dropout count. A small set of events cannot support a convincing high-percentile result. State uncertainty from recorder alignment, acoustic placement, and event detection.
3. Keep the comparison fair
Record the app versions, Windows and phone OS versions, source device, phone model, headphones, connection type, output route, and whether microphone sending was enabled. Compare speaker, wired, and Bluetooth outputs as separate cases.
When comparing configurations, use the same network, source, and physical setup. Report stability alongside speed: a smaller buffer that drops audio is a tradeoff, not an unqualified improvement. Keep startup and recovery times separate from steady playback measurements.
What if you do not have measurement equipment?
A listening comparison can still identify whether a problem appears only with Bluetooth or after enabling duplex. Label it as an observation. You can also run the browser audio sync & delay check on wired vs. Bluetooth output to compare relative lag on a repeating flash-and-click beat—keep in mind that display refresh and browser audio scheduling make visual checks relative comparisons, not calibrated millisecond measurements. A clap/video check can reveal obvious synchronization issues in a recording, too, though frame rate and device buffering limit precision. For aligning camera and audio in recordings or diagnosing broadcast delay, see fixing OBS audio and video sync and Twitch stream delay vs. audio sync.
MicDrop's diagnostics help explain network and pipeline health, but do not establish physical latency on their own. The project provides a documented measurement method and analysis tools; its engineering targets are not published benchmark results. Keep raw observations and the method with any number you share.
Sources & version notes
Checked against the documentation below on November 13, 2026. App labels and platform support can change; use matching MicDrop builds and check the current release notes. These are setup guides, not new hardware benchmark results.