START HERE
Record a visible clap and inspect the saved file. If sound is early, add positive Sync Offset to the relevant audio source. If picture is early, delay that video source. Repeat later in the recording: a growing gap needs a stability investigation, not a larger fixed offset.
Judge a saved recording first
Use a local recording with one camera and one microphone. Wear wired headphones when reviewing it, or compare more than one playback device if you suspect the headphones. A live platform preview adds delivery delay, while your OBS monitoring headphones follow another path. Neither is a reliable substitute for inspecting the file you are trying to fix.
Stand where you normally present and clap where the camera can clearly see your hands. Repeat three times with gaps between claps. Keep the microphone and camera positions unchanged through the comparison. Start with existing offsets noted, so you can restore them, and remove accidental corrections applied in several places at once.
Measure the direction and approximate size
Open the clip in an editor that displays the audio waveform and lets you step through video frames. Identify the first frame showing hand contact and the nearby clap peak. If the peak comes before contact, audio is early. If contact comes before the peak, video is early. Use several claps rather than trusting one blurred frame.
For example, two frames at 30 frames per second represent about 67 milliseconds: 2 / 30 × 1000. That is a practical alignment estimate at the recording’s frame resolution, not a calibrated end-to-end latency measurement. Compare the same event in each test and avoid claiming precision finer than you can actually inspect.
Delay the signal that arrives first
When audio is early
Open Edit > Advanced Audio Properties and find the correct audio source. Add the measured positive delay in Sync Offset (ms), then record another clap test. The Plugable capture-card procedure documents this control; its example delay is device-specific and should not be copied to every webcam or microphone.
When video is early
Open the video source’s Filters and consider Render Delay, which delays the picture. The OBS Render Delay guide describes that purpose. Add only the delay your comparison calls for and verify the new file. Adding positive delay to already-late audio moves the two signals farther apart.
Distinguish a fixed offset from drift
After the short test looks aligned, record for several minutes and repeat the claps near the end. A similar gap at both ends suggests a fixed correction may be appropriate. A gap that grows, jumps, or changes after reconnecting a device points toward a different problem.
- Repeat with the simplest source path and temporarily bypass extra audio processing.
- Check supported sample-rate settings in OBS and the involved device controls; avoid an accidental mismatch where a consistent supported rate is available.
- Compare a direct device connection with a dock or hub if timing changes coincide with reconnects.
- Reproduce the problem before collecting an OBS log, and note whether it occurs during recording alone or only while streaming.
Device timestamp options are capture-specific; the OBS source reference describes the Windows control. Do not toggle several timing options together: you would lose the ability to identify which change mattered.
Keep three kinds of delay separate
| Symptom | What to compare |
|---|---|
| Voice and lips disagree in the saved file | Camera and microphone timing inside the recording. |
| The recording is aligned but your headphones feel late | The local monitoring and headphone path. |
| Viewers see and hear a correctly aligned event several seconds later | Encoder, service, network, and player delivery delay. |
The browser sync check can help compare listening routes, but it does not automatically measure your camera-to-microphone offset. Use the actual camera and mic recording for that decision.
Verify every path you use on stream
Check the final recording after a scene transition, a device reconnect, and a longer run. If the webcam and capture card have different timing, assess them separately rather than applying one number everywhere. Keep the measured offset, source names, frame rate, and connection layout in your setup notes.
Will changing bitrate fix lip sync? A fixed capture offset should be measured at its source. Changing bitrate adds another variable without showing which signal is early. Investigate performance separately when the delay changes under load or the recording also stutters.
Sources & version notes
Documentation checked November 18, 2026. Use the platform and version notes in this guide when comparing settings. The troubleshooting checks are intended for your own setup; this review date does not represent hardware testing.