If audio stops when OBS moves from one video to the next, first find out where it disappears: the OBS meter, your local recording, your headphones or speakers, or YouTube playback. That comparison narrows the fault without assuming that the transition itself proves a particular OBS, media, encoder or YouTube problem.
Make one controlled test that includes both videos and their transition. Compare each output point, then change one setting at a time; otherwise, a fix that appears to work may only have hidden the symptom.
Find the point where audio disappears
There are four useful checkpoints: the source meter in OBS, a local recording, operator monitoring, and the stream as viewers hear it on YouTube. They represent different parts of the signal path. A moving meter is evidence that OBS is receiving audio at that point, not proof that the source is assigned to the stream output or that viewers can hear it.
Start by noting what you can actually observe. Does the relevant OBS meter keep moving after the second video begins? Does a recording made in OBS contain sound through the transition? Do headphones go quiet while the recording remains audible? Does the YouTube stream go silent even though the local recording sounds normal? Write down the answers before changing a setting.
| What you observe | What it helps you investigate next |
|---|---|
| The source meter stops moving | Whether the media source still plays audio, remains active, and is the source you expect |
| The meter moves, but the local recording is silent | Recording track selection, source track assignment, and the recording’s output configuration |
| The recording has audio, but your headphones do not | Monitoring mode, monitoring device, and whether monitoring follows a scene or source change |
| The recording is healthy, but YouTube is silent | The track sent to the stream, encoder output, and evidence in YouTube Studio |
| Both local recording and YouTube are silent | Source playback and routing are sensible first places to inspect |
These are diagnostic clues, not verdicts. For instance, a moving meter can coexist with a track-routing problem, while a quiet monitoring device can coexist with a healthy stream. If you are unsure which sound source matters, start with a bhajan or lofi track you can recognise and listen through the entire transition, not just the opening seconds.
If the silence is only in the broadcast, resist replacing a microphone or buying an audio interface before you have evidence of a hardware fault. A video file that played normally a minute earlier can still be routed differently from another source, and monitoring can mislead you about what the encoder receives.
Make a short controlled recording
Before changing your scene collection, save a copy of it or note its current settings. Then make a brief local recording that contains the end of the first video, the transition, and enough of the second video to tell whether its audio returns. Use the same scene and media sources as the live setup. A test that omits the transition cannot answer the question.
In OBS, start recording, play the first video, let the second one begin, and stop the recording after the test. Listen to the saved file from before the transition through to its end. Do not judge it only by the preview in OBS, because the preview is not the same evidence as a saved output file. If you use a separate audio track for recording, note which track the recording uses before drawing conclusions from silence.
Keep the test small and repeatable. Avoid changing audio devices, track assignments, source visibility, and playback controls all at once. If the first recording is silent, repeat once after checking that the right source is present and that the recording has not simply been played through a muted media player. The aim is to tell whether the local output reproduces the symptom, not to prove a cause in a single pass.
YouTube’s live-stream troubleshooting guidance advises checking the stream in the encoder and looking at the local archive when investigating audio or video quality. That split is useful here: if the archive shares the silence, concentrate first on OBS playback and routing; if it is healthy, look at the encoded stream and delivery evidence. This is a diagnostic inference, not a guarantee about where the fault lies.
If the test is for a live channel that normally runs while you are away, keep the controlled recording separate from the normal schedule. You can use a private or otherwise appropriate test setup rather than interrupting a public devotional or news loop. Check YouTube’s current guidance for the visibility and testing options available to your channel before going live.
Check the source audio meters
In OBS, identify the meter that belongs to the affected video or audio source. Play each video in turn and watch whether that meter responds. If it goes quiet exactly when the second source begins, inspect that source’s playback and whether the intended source is active. If the meter continues moving, audio is reaching that metering point, but routing and output still need checking.
A scene transition can involve more than one source change. The first media source might be stopped, hidden, replaced or restarted while the next one appears. Inspect the source list and playback controls at the moment of transition. Confirm which source is meant to provide sound, and whether audio is embedded in the video or comes from a separate source such as a music track.
Do not infer that a source is silent merely because you cannot hear it in your headphones. Monitoring is a separate path. OBS provides monitoring choices including “Monitor Off”, “Monitor Only (mute output)” and “Monitoring Enabled”; the wording itself distinguishes monitoring from the output sent to the stream. A monitoring device may also be set differently from the one you are listening to.
For a clean comparison, temporarily test with monitoring off, then compare the source meter, local recording and YouTube output. This avoids treating a monitoring-only silence as proof that viewers hear silence. Restore your preferred monitoring configuration after the test, and check the selected device if monitoring alone was the problem.
There have been reports of monitoring behaving differently across specific OBS versions and Windows setups, but those reports are not evidence that your case has the same cause. One OBS issue report describes a particular older Windows setup where monitoring stopped intermittently after scene transitions even while stream audio continued. A later issue report describes different monitor/output behaviour on a particular newer setup. Treat both as examples of why to test your own output paths, not as proof of a general defect.
Verify track routing
OBS can show a source meter while that source is not assigned to the audio track used for the stream. Open Advanced Audio Properties and find the affected source. Check which tracks are ticked, then confirm which track OBS is using for stream output in your audio settings. A source must be assigned to a track included in that output path.
The OBS interface warns that a source assigned to no audio tracks will not be audible in streams or recordings. Its interface text is a useful reference for that warning and the monitoring labels. The important practical point is to check both ends: the source’s assignment and the output’s selected track. A tick on a track that is not being sent to the stream will not solve the broadcast problem.
If your first and second videos are separate media sources, compare their track assignments side by side. A common configuration mistake is to route one source to the stream track and leave the other assigned elsewhere or to no track. Check any separate music source in the same way. Do not assume that sources in the same scene inherit one another’s audio settings.
Recording and streaming can use different track selections. If the local recording is audible but YouTube is silent, verify the stream track before editing the recording settings. If the recording is silent but the stream sounds normal, inspect the recording track instead. After making one change, repeat the short test and listen to the saved file or stream that was previously silent.
If you have an established setup that sends different language, music or microphone mixes to different tracks, write down the existing assignments before adjusting anything. A quick trial-and-error change can restore one source while removing another from the broadcast. Test the two video sources separately as well as in sequence, so you can tell whether the issue follows a particular source or appears only during a transition.
Check media playback transitions
If the local recording also goes silent, inspect what happens to the media sources as one video gives way to the next. Confirm that the second source is visible and active, that its playback has started, and that its audio is part of the file or supplied by the expected separate source. If a source is being restarted, replaced or hidden, note whether that action also changes its audio state.
Play the second file on its own in the same OBS scene. Then test the transition again. If it is silent both alone and after the first clip, the second source or its routing deserves closer inspection. If it plays alone but not after the transition, review the scene change and media controls around that point. Neither result, by itself, proves an OBS fault; it tells you where to focus the next test.
Check for unintended overlap or gaps. A transition may hide the first source before the second has begun, or leave two sources active when you intended only one. Look at source visibility and playback controls while the test runs, and listen to the recording around the change. If your playlist relies on a separate music bed that should continue underneath both videos, test that bed independently from the video audio.
Where a long playlist is built from prepared clips, the media files and the way they join matter as well as OBS settings. The guide to preparing seamless loop videos covers planning clips for continuous playback; it is relevant when a transition is part of a repeated sequence, though it does not replace checking OBS audio routing. For a playlist assembled from multiple sources, keep a test copy so you can adjust transition behaviour without disrupting the version already scheduled.
Do not assume that the first-to-second-video transition establishes a specific media-player bug. It is simply the point at which your test revealed a change. Compare the two sources, their tracks and their playback state, then repeat the recording after the smallest useful adjustment.
Review encoder and connection status
When the local recording is healthy but the YouTube stream is not, inspect YouTube Studio’s live dashboard for encoder errors or other warnings. Check the stream directly rather than relying only on OBS’s preview. YouTube’s troubleshooting guidance recommends looking for audio and video issues in the local archive and, if the encoder output is healthy, checking the outbound connection.
Also look at OBS’s connection status and logs around the time the audio disappears. OBS’s connection troubleshooting guide explains that dropped frames indicate connection instability or an inability to sustain the configured bitrate. That clue matters if you see connection symptoms, but dropped frames alone do not explain audio-only silence. Record what the dashboard and logs show before changing bitrate or encoder settings.
If the dashboard reports encoder trouble, check the settings and system load against the requirements YouTube currently publishes. Avoid changing resolution, frame rate, bitrate and audio routing together: if the result changes, you will not know which change mattered. A resolution mismatch is a separate issue from a source that has no audio on the selected track; consult the encoder resolution mismatch guide if YouTube reports that specific warning.
If connection instability is present, use the evidence to test the network path and whether the configured stream can be sustained. For a home setup, note whether other network activity coincides with the trouble and whether the dashboard reports dropped frames. Do not treat a network adjustment as an audio fix unless a repeat test shows that it changed the outcome. An audio-only symptom with no encoder or connection warnings should send you back to the track and source checks rather than into speculative network changes.
For an always-on channel, restarting or reconfiguring a live stream can affect viewers and the schedule. Make a test plan before changing a working broadcast. If you are also investigating recovery after an OBS or streaming process exits, the separate guide on restarting a YouTube live stream after FFmpeg exits addresses process recovery; it does not diagnose audio routing in OBS.
Retest the YouTube stream
After you have a candidate fix, repeat the same controlled sequence: first video, transition, second video. Check the source meter, make a fresh local recording, listen to monitoring if it is part of your normal workflow, and inspect the stream on YouTube. Keeping the test sequence constant makes it easier to tell whether the change fixed the output that was actually silent.
Use a small record of the results: the OBS version, the two sources, the track assignments you changed, what happened to the meter, and what you heard in the archive and stream. This is more useful than a note saying “audio failed after video two”. If the problem returns overnight, the record gives you a starting point and makes it easier to distinguish a repeatable source transition from a one-off connection interruption.
Once the stream works, avoid changing unrelated audio settings immediately. Let the normal sequence run long enough for you to verify the transition that previously failed. For a channel that plays licensed or devotional music, keep separate records of the content and operational tests: a clean audio signal does not answer whether the programme is cleared for your intended use. The guide to music sources for a 24/7 stream discusses the separate question of finding music sources; check current rights terms yourself.
If the controlled recording and stream disagree but the basic track and monitoring checks do not explain why, preserve the OBS log and the relevant YouTube dashboard details, then seek help with those specifics. Include what remains audible at each checkpoint and the exact transition sequence. That evidence is far more actionable than changing cables, interfaces or microphones without a confirmed hardware problem.
Before committing, compare the operating options on the pricing page. When the file and channel are ready, start free — 24-hour trial, no card.
FAQ
Does silence after the second video prove that OBS has a media-source bug?
No. The transition tells you when the symptom appears, not which component caused it. Compare source meters, a local recording and YouTube playback, then inspect playback state and routing based on what those tests show.
The OBS meter moves, but viewers hear nothing. What should I check?
A moving meter does not prove that the source is assigned to the track sent to the stream. Check Advanced Audio Properties for the source’s track assignment, then confirm that the stream output uses that track. Verify the result with a fresh recording or YouTube test.
I cannot hear the second video in my headphones. Is the stream silent too?
Not necessarily. Monitoring has its own settings and can differ from the stream output; “Monitor Only (mute output)” specifically describes a monitoring mode that mutes output. Compare a local recording and the YouTube stream before changing the broadcast mix.
Should I replace my microphone or buy an audio interface?
Not without evidence that a hardware component is at fault. If the symptom follows one media source or its track assignment, replacing recording hardware is unlikely to address the diagnosed issue. Test the source, routing and output paths first.