Webcam A/V Sync & Latency Meter: Code Examples

OBS Studio + USB Webcam + USB Mic A/V Sync Guide

Start with direct USB connections before measuring the sync offset. OBS Studio with a direct USB webcam and a direct USB microphone is the most common streaming setup in 2026. Both devices connect separately to the host computer, each traveling through its own USB pipe with its own latency characteristics. Because USB audio delivers packets in 1 ms isochronous frames while USB video delivers frames in larger bulk transfers, audio almost always arrives at OBS before the matching video frame.¹² Measuring this gap and entering it in OBS takes under three minutes.

For most users on this setup, the measured offset falls between +50 and +110 ms, with the specific value depending on the webcam model, USB host controller generation, and background system load. Both devices should connect directly to USB ports on the motherboard rather than through a hub for the most stable measurements.³

Setting up the test

Close OBS before running the 5-clap test in the browser. When OBS has the webcam open as a video source, it shares the camera feed with the browser, which can add scheduling contention and inflate the offset reading. Measure in the browser alone, then open OBS and enter the correction. After applying the sync offset in OBS, close the browser and confirm OBS captures the webcam without browser interference.

Select your USB webcam in the video input dropdown and your USB microphone in the audio input dropdown. Confirm both devices appear in the dropdowns before clicking Start Test. If either device does not appear, check OS permissions for camera and microphone access. On Windows 11, the Privacy and Security settings include separate toggles for camera and microphone access at both the system level and the individual app level, so even if the browser has permission, the OS-level toggle for the specific device type can still block enumeration.

Before clicking Start Test, confirm both devices appear in the browser dropdowns with their correct model names. An entry labeled "Default" or "Communications Device" in the audio dropdown refers to the Windows default audio device, not necessarily your USB microphone. Selecting the default device rather than the explicit USB microphone entry can produce a measurement for a different audio device than the one OBS captures, resulting in an offset value that does not match your streaming setup. Select the explicit device name for both the webcam and microphone inputs before running the five-clap test.

Matching browser dropdowns to OBS sources

The correction only helps when the browser measurement and OBS source names describe the same devices. If OBS captures a Blue Yeti X but the browser test used the webcam mic, the OBS Sync Offset value will correct the wrong timing relationship. Taking a screenshot of your browser dropdown selections before you start the test gives you a visual record that you can cross-reference against your OBS source list, eliminating the guesswork when you return to adjust the offset weeks later.

The same caution applies after you swap microphones. A new mic appears under a different dropdown name, but your saved OBS value still points at the old device unless you remeasure. Keep the screenshot next to your saved offsets so the name on the device always matches the value you paste into Advanced Audio Properties.

Interpreting and applying the result

The result panel displays the averaged offset and a pre-formatted OBS Audio Offset value. Copy this value. In OBS Studio, open Edit > Advanced Audio Properties. The table lists every audio source in your current scene; find your microphone and enter the copied value in the Sync Offset (ms) column.⁴

Building on this, if your scene uses multiple cameras with separate microphone sources, each pairing needs its own measurement. A different webcam produces a different offset even with the same microphone because each camera has its own sensor readout time, ISP pipeline, and USB transfer characteristics. Measure each webcam-microphone pair independently and apply separate sync offsets in OBS, and label each saved value with the camera model and microphone name so you can quickly identify which offset applies to which scene configuration.

Handling multiple camera sources in an OBS scene

A common OBS scene layout uses two video sources: a primary webcam for face cam and a secondary camera for a wider view or a different angle. Each camera has its own USB pipeline latency, but the scene shares a single microphone source. The Sync Offset in OBS Advanced Audio Properties applies to the audio source, not the video source, which means you cannot apply a different offset for each camera individually within the same scene without adding separate audio sources.⁴ The practical approach is to measure the offset for your primary camera and apply it to the microphone source. If the secondary camera has a significantly different offset, create a separate OBS scene with an adjusted Sync Offset for that camera's active use.

Keeping the USB topology stable after correction

Once you enter the OBS value, do not move either USB device to a different port or hub before recording. A different host controller can change video timing, and a different audio endpoint can change the audio-side buffer. If you must change the setup, run the five-clap test again and replace the old Sync Offset value with the new one.

When to use this

Use this guide when setting up a new OBS streaming PC, when changing your webcam or microphone, or when you notice lip-sync drift in a recent recording that was previously accurate.

Notes

OBS Studio applies the Sync Offset per audio source. Enter the measured value in Edit > Advanced Audio Properties > Sync Offset (ms) for the microphone source. Video sources in OBS do not have a timing offset control. The correction always applies to audio. After applying, record a 10-second test clip to verify alignment.

Examples

+72 ms typical

+88 ms typical

Try in the tool

OBS + Elgato Cam Link + XLR Microphone A/V Sync Guide

Measure the full HDMI and XLR chain before entering a delay. DSLR streamers using Elgato Cam Link 4K with an XLR microphone through a USB audio interface often have larger and more variable A/V offsets than native USB webcams. Three signal paths contribute: the camera's HDMI output, the Cam Link capture path, and the audio interface's USB audio path. Elgato describes Cam Link 4K as an HDMI-to-USB 3.0 device that appears as a webcam in applications, so its timing depends on both the camera output and the capture path rather than a single webcam sensor pipeline.¹

On this setup, measured offsets can be large enough to look visibly wrong if you guess instead of measuring. Accurate measurement is especially important here because a wrong sync offset produces a consistent lip-sync gap across every recording made with that signal chain.

Setting up the test for a Cam Link setup

Connect the camera to the Cam Link, connect the Cam Link to a USB 3.0 port, and connect the audio interface to a separate USB port.² Open a browser tab and navigate to this page. In the video input dropdown, select the Cam Link's camera entry (labeled by the Cam Link model or the HDMI source). In the audio input dropdown, select your audio interface input channel.

Set the camera to a fast shutter speed (1/100s or faster) before starting the test. Slower shutter speeds blur the clap motion across multiple HDMI frames, softening the luminance transient and causing missed or inconsistent detections.³ Fast shutter speeds produce sharp single-frame luminance spikes that the detector captures consistently.

Why shutter speed matters for Cam Link measurements

At a slow shutter speed of 1/30s, the camera exposes each frame for about 33 ms, the entire duration between frames at 30fps. Any motion during that 33 ms blurs across the exposure, spreading the clap's luminance change into the frame before and after the actual clap moment. The luminance detector measures the peak frame delta, which a blurred transient reduces significantly. At 1/100s shutter speed, the 10 ms exposure window captures the clap motion as a sharp spike in one frame, producing a much higher frame-to-frame luminance delta that the detector timestamps accurately. Fast shutter speed is the single most impactful camera setting for improving Cam Link detection reliability.

Set the shutter once and leave it fixed for all five claps. Changing the exposure mid-test mixes sharp and blurred transients, which makes the average less reliable. If your lighting forces a slower shutter, accept the wider reading spread and repeat the pass rather than adjusting exposure between claps. A consistent setting produces a tighter set of readings that you can trust as the real offset for the chain.

Applying the result in OBS

After measuring, enter the displayed OBS Audio Offset value in OBS Advanced Audio Properties for your XLR microphone source (typically labeled by the audio interface model and input channel number).⁴ Because this value accounts for both Cam Link capture latency and audio interface pipeline latency, it is the complete correction value for your full signal chain.

Building on this, if you switch cameras while keeping the same Cam Link and audio interface, re-measure. Different camera bodies produce different HDMI output timing. Keep the same ASIO or production driver configuration when comparing camera swaps so the measured difference reflects the camera path rather than a changed audio driver state.⁵

Different camera bodies connected to the same Cam Link produce different total offsets because each camera processes its HDMI output through its own imaging pipeline. Switching from one camera body to another on the same Cam Link and XLR microphone combination requires re-measuring and updating the OBS Sync Offset. Keep a record of each camera body's measured offset alongside the Cam Link's base latency so camera swaps require only a values lookup rather than a full re-measurement session.

Recording offsets for camera swaps

Keep a note for each camera body: camera model, HDMI mode, Cam Link USB port, audio interface driver, buffer size, and measured OBS value. That note prevents you from reusing a value from the wrong camera when you swap rigs between streams. Storing this information in a simple text file or spreadsheet means you can update it once and reference it every time you change your camera configuration, turning a five-clap re-measurement session into a quick lookup that takes under 30 seconds.

Keeping the ASIO buffer fixed

The audio interface buffer size is part of the measured signal chain. If you measure with a 128-sample ASIO buffer and then switch to 256 samples for a stream, the audio path changes and the old offset no longer matches. Before recording the final value, decide on the buffer size you will use for production, keep it fixed, then run the five-clap test and enter that result in OBS.

When to use this

Use this guide when setting up an OBS streaming or recording rig with an Elgato Cam Link 4K or similar HDMI capture card and a separate XLR microphone through an audio interface, or when switching camera bodies in an existing Cam Link setup.

Notes

ASIO audio drivers for the audio interface provide lower audio latency than Windows WASAPI. Lower audio latency means audio arrives even earlier relative to Cam Link video, increasing the positive A/V offset. Measure with your exact ASIO buffer size configuration active. Enter the result in OBS Advanced Audio Properties for your XLR microphone's audio interface input channel.

Examples

+135 ms typical

+158 ms typical

Try in the tool

Zoom Video Calls + USB Webcam + External Microphone A/V Sync Guide

If Zoom lip-sync looks wrong, start with the local camera and microphone pair. Zoom with an external USB microphone paired with a dedicated USB webcam is a common setup for remote workers and hybrid meeting participants. Zoom's media pipeline uses timestamps to keep audio, video, and screen-share data in order, and those timestamps are essential when combining or checking media streams.¹ Because the camera and microphone can arrive through different device pipelines, the safest workflow is to measure the local webcam-to-microphone offset before an important call rather than guessing.

Measuring the webcam-to-microphone offset before important calls takes under two minutes with the 5-clap test. Knowing your exact offset tells you whether the setup is close enough for a normal Zoom call or whether you need to verify the issue in Zoom test meeting and use a local recording or monitoring tool with a millisecond Sync Offset control.

Measuring before a Zoom call

Close the Zoom application before running the browser test. Windows cameras can only be accessed by one app at a time, and a background app such as Zoom, Teams, Skype, or a browser tab can hold the camera until it is released.² Select your webcam and external USB microphone in the browser dropdowns, run the five-clap test, and note the offset.

Zoom's media pipeline uses separate audio and video timestamps, and timestamps are essential for syncing those streams together.³ Use the browser measurement as a local diagnostic baseline, then verify the same webcam and microphone behavior inside Zoom before joining an important meeting. Zoom does not currently expose a standard millisecond microphone sync offset field in its Desktop Client settings, so the browser-visible measurement remains your best tool for understanding whether your specific hardware combination produces a noticeable timing gap before you share your camera with other participants.

Before running the browser test, close the Zoom application completely. If the browser cannot access the camera, check for another app or tab that may still be using it, close that app, and run the five-clap test again. Record the offset value, then open Zoom and verify the sync in a test call before joining the real meeting.

Choosing the external microphone entry

Select the external USB microphone entry in the browser audio dropdown, not the webcam built-in microphone. The external mic is the audio path Zoom will use if you selected it in Zoom Settings > Audio. If the dropdown shows a headset, USB audio interface, or virtual audio device, choose the same device name you choose in Zoom before the meeting.

Confirm the selection in Zoom matches the browser before you start. A mismatch between the browser dropdown and the Zoom audio setting means the measured offset describes a path Zoom is not using. When both apps point at the same external microphone name, the browser number becomes a reliable predictor of the call's timing behavior.

Verifying Zoom after the browser measurement

After measuring, open Zoom and use Settings > Audio to confirm the same microphone and camera are selected. Join a test meeting, clap once in front of the camera, and check whether the visible hand impact and clap sound line up in the preview. If the preview is unavailable, ask a trusted contact to join a quick two-person test call and watch your video for lip-sync alignment while you clap and speak.

Verifying and applying the correction

Open the Zoom Desktop Client. Navigate to Settings (gear icon) > Audio > Advanced and review the available audio options. Zoom exposes advanced audio controls such as echo cancellation and raw audio mode, but the standard Zoom Desktop Client UI should not be treated as a general millisecond A/V sync offset field for this setup.⁴

If you need a millisecond correction for local recording or monitoring, enter the measured value in OBS Studio Advanced Audio Properties for the microphone source.⁵ Then join Zoom's test meeting at zoom.us/test, clap once in front of the camera, and check whether the visible hand impact and clap sound line up in the preview.⁶ If the Zoom test meeting does not provide a review recording in your region, ask a trusted contact to join a quick two-person test call and watch your video for lip-sync alignment while you clap and speak.

When Zoom has no user-facing offset field

Zoom participants see the audio and video Zoom receives unless you route a virtual camera or virtual microphone. For local recording or monitored output, OBS can apply the measured Sync Offset; for normal Zoom calls, use the measurement to choose a stable device path and verify with a test call. Without a direct offset control, the browser measurement becomes your diagnostic tool for comparing microphone candidates before you join.

When to use this

Use this guide before an important Zoom presentation or meeting where visible lip-sync would look unprofessional, when switching to a higher-end webcam with a longer processing pipeline, or when an external USB microphone replaces the webcam's built-in mic.

Notes

Zoom's Advanced Audio settings focus on processing options such as echo cancellation and raw audio mode. If you need a millisecond A/V correction for local recording or monitoring, use a recording or streaming app with per-source Sync Offset controls, such as OBS Studio.

Examples

Measure before call

Measure before call

Try in the tool

Streamlabs OBS + USB Webcam + USB Headset A/V Sync Guide

Start with the exact headset microphone you will use live. USB headsets are the starting point for most new streamers because the microphone is built into the headset, avoiding the cost of a separate condenser mic and audio interface. Paired with a USB webcam in Streamlabs OBS, this combination still needs A/V sync calibration because the browser receives the webcam and headset microphone as separate local media devices.¹ The headset microphone travels through the USB audio input path, while the webcam travels through the video capture path, so the measured offset can differ from the webcam's built-in microphone path.

After measuring, apply the correction to the headset microphone source in Streamlabs OBS/Streamlabs Desktop audio settings rather than guessing. Streamlabs exposes audio source delay through Advanced Audio Settings with a Sync Offset column, which is different from adding a video Render Delay filter to the webcam source.²

Identifying your headset microphone in the browser

The browser media API is designed to request local multimedia devices such as microphones and video cameras, and it exposes information about the devices that produce those media streams.³ USB headsets usually appear as separate audio output and audio input entries, so the browser dropdown should show the headset microphone as an input device rather than the headphone output.

Select the microphone entry in the browser audio input dropdown. On Windows, Microsoft's microphone setup instructions also send you to Settings > System > Sound > Input, where you choose the microphone you want to use and test it.⁴ If two entries look like microphone inputs, run the five-clap test with each one and use the entry that gives the cleanest, most repeatable clap transient.

Identifying the hardware headset microphone in Windows

USB headsets present separate playback and recording paths to the OS. In the browser audio input dropdown, look for an entry that includes your headset model name, "Microphone", "Input", or a generic USB audio-device label. If two entries look like microphone inputs, test each one; a virtual audio device or processing chain can add extra delay, so the browser measurement helps you choose the path that matches your actual Streamlabs microphone source.

Write down which entry produced the cleaner reading so you can reselect it later. Windows can reorder or rename USB audio entries after a reboot or a driver update, and picking the wrong one silently measures a different path. Matching the browser entry to the source Streamlabs actually captures keeps the offset value meaningful across sessions.

Applying the correction in Streamlabs

After measuring, open Streamlabs OBS or Streamlabs Desktop. In the Audio Mixer, open the settings for your headset microphone source and use the Advanced Audio Settings area. Elgato's Streamlabs OBS delay guide describes opening the mixer settings icon, finding the Sync Offset column in Advanced Audio Settings, and setting the delay in milliseconds for the audio source.²

Enter the measured millisecond value for the headset microphone source. OBS Studio source code also shows that audio sync offset is stored and displayed in milliseconds, which is the same unit used by browser latency measurements and Streamlabs delay controls.⁵ Record a short clap test clip in Streamlabs, play it back, and adjust the Sync Offset until the clap sound lines up with the visible hand impact.

Streamlabs Desktop also provides audio filters such as Gain, Noise Suppression, Noise Gate, Compressor, Expander, and Limiter, but those filters are for audio quality rather than A/V sync.⁶ Keep the Sync Offset correction on the microphone source, then use audio filters only if you also need to reduce noise, control peaks, or shape the voice level.

Audio filters are not sync controls

Audio filters can change tone and loudness, but they do not replace the Sync Offset correction for the headset microphone. If you add a filter after measuring, re-run the five-clap test only if the filter changes the production audio path timing or you hear a new sync shift. Adding a compressor or noise gate after calibration is usually safe because those filters add negligible latency, but a linear-phase EQ or look-ahead limiter can introduce enough delay to shift the measured offset by several milliseconds.

Keeping the headset path stable

Once you measure the headset microphone, do not switch to the webcam built-in mic or a different headset during the same stream. The Sync Offset value belongs to the specific audio source you measured. If you change the microphone source, run the five-clap test again and update the Streamlabs value for the new source. Keeping a short note beside your Streamlabs profile listing the measured offset and the exact headset model prevents accidental reuse of a stale value after a hardware change.

When to use this

Use this guide when setting up Streamlabs OBS for the first time with a USB headset, when upgrading from a built-in webcam to a dedicated USB webcam, or when headset audio sounds slightly ahead of lip movement in Twitch or YouTube VOD playback.

Notes

Right-click the headset microphone source in the Streamlabs mixer, open Advanced Audio Settings, and enter the measured value in the Sync Offset column. Record a short clap test clip and verify the sound lines up with the visible hand impact before streaming.

Examples

Measure before stream

Measure before stream

Try in the tool

Elgato Facecam + Wave:3 + OBS A/V Sync Calibration

The Elgato ecosystem still needs the same A/V sync measurement as mixed-brand gear. Elgato's Facecam and Wave:3 form a popular matched ecosystem for streamers and content creators. Facecam uses a USB 3.0 Type-C connection and supports 1080p60, while Wave:3 uses USB-C and a cardioid condenser capsule.¹² Using the same brand for camera and microphone does not, however, mean the two devices are time-synchronized out of the box. The camera and microphone still travel through separate capture paths, so the offset requires measurement and correction just like any other pairing.

Measure the exact Facecam and Wave:3 combination you plan to stream with. If your OBS scene captures audio from the hardware Wave:3 device, measure that hardware microphone path; if your OBS scene captures audio from Wave Link Stream or another virtual mixer output, measure that virtual audio path instead.

Setting up the test for the Elgato ecosystem

In the browser video input dropdown, select "Elgato Facecam". In the audio input dropdown, select "Elgato Wave:3" if your OBS scene uses the hardware microphone source.² If your OBS scene uses Wave Link Stream or another virtual mixer output, select that Wave Link device instead; Windows users may see "Wave Link Stream" and "Wave Link Headset" as virtual audio devices managed through the Sound control panel.³

Run the five-clap test with the same camera resolution and audio source that your OBS scene will use. The Facecam supports 1080p60 and other 1080p/720p/540p modes, so keep the production mode active during measurement rather than switching resolution after calibration. A mismatch between the browser test mode and the OBS capture mode introduces a timing discrepancy that makes the five-clap correction unreliable, because the detector measures one frame rate while OBS records another.¹

Hardware Wave:3 vs Wave Link as the OBS audio source

Your OBS scene can capture audio from either the physical Wave:3 hardware device or a Wave Link virtual mixer output. The important part is source matching: measure the hardware Wave:3 path when OBS uses "Elgato Wave:3", and measure the Wave Link path when OBS uses "Wave Link Stream" or another Wave Link output. Using the wrong source's offset in OBS produces a correction for the wrong audio pipeline.

Decide which path OBS records before you measure, then keep it fixed. Switching between the hardware Wave:3 and a Wave Link output after calibration changes the audio side of the offset, so the value you entered no longer matches. When the OBS source and the browser dropdown name the same device end to end, the correction stays accurate from measurement through to the live stream.

Applying the result in OBS

Enter the measured value in OBS Advanced Audio Properties > Sync Offset (ms) for the Wave:3 or Wave Link audio source.⁴ The value should match the source name you selected during the browser test, not just the device brand. Double-check that the OBS source name and the browser dropdown entry describe the same physical or virtual device, because applying a Wave:3 hardware offset to a Wave Link Stream source corrects the wrong audio pipeline and can leave the actual recording out of sync.

Verify the sync by recording a short clap test clip in OBS and watching the output. If the clap sound arrives before the visible hand impact, increase the Sync Offset value. If the clap sound arrives after the visible impact, decrease the value. Re-record after each adjustment until the sound and motion line up in the exported clip. If your Facecam source has OBS Video Capture Device buffering enabled and you see persistent device delay, OBS notes that disabling buffering can help with delay on capture devices.⁵

Wave Link state before the final OBS value

Wave Link can change the audio source OBS records, so the browser dropdown must match the OBS audio source exactly. Measure with the hardware Wave:3 path when OBS records the physical microphone, or with Wave Link Stream when OBS records the virtual mix. A clean note beside the value prevents future confusion when the same Elgato devices appear as multiple Windows audio endpoints.

Keeping the Elgato software state fixed

Do not change Camera Hub, Wave Link, or OBS source settings after you measure the offset. If Camera Hub changes the Facecam mode or Wave Link changes the audio source, the measured value no longer matches the production path. Keep the same camera resolution, same audio source, and same software state while measuring and applying the OBS Sync Offset. A quick checklist saved beside your streaming profile, listing the active Camera Hub preset and the Wave Link output name, helps you restore the exact state before each session.

When to use this

Use this guide when setting up an Elgato Facecam and Wave:3 combination for the first time, when adding Wave Link software to an existing Elgato setup, or when changing the OBS audio source between Wave:3 hardware and Wave Link virtual output.

Notes

Wave Link can expose virtual Windows audio devices such as Wave Link Stream and Wave Link Headset alongside the hardware Wave:3 microphone. Measure with the same audio source your OBS scene uses, then enter the resulting value in OBS Sync Offset for that source.

Examples

Measure before stream

Measure before stream

Try in the tool