When audio and video go out of sync in a pre-recorded YouTube stream, first find the earliest version where the mismatch appears: the source file, a local recording, the encoder preview, or the stream viewers receive. The right fix depends on that point, so changing encoder settings before comparing those versions can hide the evidence rather than solve the problem.
“Pre-recorded stream” can mean a video file sent through OBS or another encoder to YouTube Live, or a regular video uploaded to YouTube. Those are different workflows. The checks below separate them and avoid assuming that YouTube automatically repairs or creates sync problems.
Identify which workflow you are using
Start by naming the path your video takes. In a live-encoder workflow, a file plays in OBS or similar software, is encoded as a live signal, and is sent to a YouTube Live event. You may also save a local recording from the encoder. A regular upload instead goes from a file into YouTube as an uploaded video; it does not pass through your live encoder preview or stream health checks.
That distinction matters because YouTube's live-stream troubleshooting guidance is about live streams. It recommends checking quality in the encoder and looking at the local archive. It does not establish a universal repair process for an ordinary uploaded video. For an upload, compare the original file with the published playback and do not infer that a live-stream setting applies.
Write down what you observe before changing anything. Is the sound ahead of the picture, or does it lag? Is the mismatch already present at the start, or does it become more noticeable as playback continues? A constant offset and a mismatch that grows over time are different observations; neither, by itself, identifies the cause. Record the point in the video, the device or player used, and whether you are checking a local file, encoder preview, live playback, or replay.
If your channel plays a long lecture file repeatedly, keep the case-specific checks in our guide to audio drifting in a 24/7 lecture stream in mind, but still locate where the fault begins. A long-running stream can make an existing source-file issue easier to notice; it does not prove the encoder is responsible.
Compare the source file and local archive
Play the source file in a local media player and check a clear sync point. Choose a moment where a visible action has a corresponding sound, such as a speaker beginning a word, a bell strike, or a handclap. If the file contains music without visible events, use a section with a clear change in the picture and listen for the corresponding sound. Repeat the check near the beginning and later in the file, rather than judging from one short moment.
Next, compare that source with any local recording or archive saved by the encoder. YouTube specifically recommends checking the local archive when troubleshooting a live stream. Use the same scene and player where possible. If the source and archive are both wrong in the same way, the problem is already in the material or its playback path. If the source is correct but the archive is not, focus on how the encoder receives and records the media before investigating delivery to YouTube.
A useful record is a small comparison table. It helps distinguish “the stream is wrong” from “the stream is the first place I checked”:
| Checkpoint | What to compare | What a mismatch points towards |
|---|---|---|
| Source file | Picture and sound at the same identifiable event | The supplied file, its edit, or the player used to inspect it |
| Local archive | The same event in the encoder's saved recording | Local playback, source routing, or encoding/recording path |
| Encoder preview | The corresponding action and sound while the file plays | Input selection, media playback, or local production path |
| Viewer-facing stream or replay | The same event as received on YouTube | Delivery path or an issue that appears after the local output |
These clues narrow the next check; they do not prove a cause on their own. If the local archive is healthy but the stream is not, for example, that is a reason to inspect stream health and the outbound connection, not proof that YouTube is at fault.
For a channel that runs continuously, the computer doing the playback is part of this local path. If you are weighing a small always-on PC against hosted operation, the India electricity cost example for a 20-watt mini PC helps frame the running-cost question, but it does not diagnose sync. First establish whether the local file or archive itself is sound.
Check the encoder preview
With a live-encoder workflow, watch the encoder preview while the file plays. Check that you have selected the intended media and audio sources and that the sound you hear corresponds to the visible action. If audio comes from a separate source, verify that it is the intended track rather than a microphone, desktop mix, or duplicate source. The exact controls differ by encoder, so avoid copying a fixed offset or menu path without checking instructions for your software and setup.
Compare the preview with the local archive rather than treating either as a perfect reference. If both show the same fault, work on the local production path: confirm source selection, playback, routing, and encoder diagnostics. If the preview appears correct but the archive does not, check which output the encoder is recording. Keep the test focused on one identifiable event and repeat it at a later point in the file.
YouTube advises using the latest encoder version and checking encoder errors and CPU load as part of live-stream troubleshooting. Those checks are most useful when the local output shows trouble or the encoder reports a problem. They do not mean a busy processor is automatically the cause of every audio/video mismatch. Note any error messages and the time they appear, then compare them with the mismatch.
A test should resemble the real programme, not just a static logo or a few seconds of silence. YouTube Help says: “Make sure that you test before you start your live stream. Tests should include audio and movement in the video similar to what you'll be doing in the stream.” Use a representative section with movement and sound, and monitor the encoder while it runs. If you are preparing an OBS layout for a continuous station, our article on adding a clock and ticker to an OBS 24/7 stream covers a different production detail, but the same discipline applies: test the actual scene you plan to broadcast.
Check the stream or replay
If the source, local archive, and encoder preview appear aligned, check what viewers receive. During a live event, compare a recognisable moment in the preview with playback of the stream; allow for playback delay when matching the moment, but judge whether the visible and audible event within the received stream agrees. Afterward, check the replay as another observation. Do not compare a live preview and a later replay at different moments and mistake their delivery delay for sync drift.
YouTube Help notes: “If your stream looks and sounds healthy, there may be an issue with your outbound Internet connection.” That is a branch to investigate, not a finding that the platform caused the mismatch. Check stream health and the connection from the encoder to the internet. Google's LiveStream health status documentation describes configuration and quality statuses, including insufficient video delivery for smooth streaming. Treat a health warning as evidence to follow up, not as a complete diagnosis of audio sync.
OBS also notes that increasing dropped frames can indicate an unstable connection or a bitrate the connection cannot sustain. If the encoder preview and local recording remain healthy while connection or stream-health indicators show trouble, test the outbound connection and consider the bitrate in relation to what the connection can sustain. YouTube's guidance suggests contacting your internet provider if testing identifies a connection issue. If the indicators are healthy, continue gathering evidence rather than changing bitrate speculatively.
Keep a short log: the time, the checkpoint, whether the mismatch was fixed or growing, and any encoder or stream-health messages. Repeat the test from a different network only if it is practical and relevant to the setup; do not confuse a viewer's buffering or device playback problem with a fault in the encoded programme. If the stream and replay disagree with each other, preserve both observations and compare the same event before changing the source or encoder.
Fix problems already present in the source
If the source file is out of sync in a local player, changing the live encoder is unlikely to correct the underlying edit. Return to the editing project or obtain a corrected export. Check the video and audio together at the beginning and a later point, then export a short representative test and inspect it locally before replacing the file used for the stream. Keep the original unchanged until the replacement has been checked.
If the source seems aligned in one player but not another, repeat the test in a second player before concluding that the file itself is faulty. That comparison can reveal that the inspection setup is affecting what you hear or see. For a source that gets progressively less aligned, note where the difference becomes clear and test an export across the same span. A growing mismatch is worth separating from an offset that is already present at the start, but the observation alone does not tell you which correction to apply.
Avoid applying an arbitrary audio delay based on a single moment. A fixed adjustment might make one point appear better while making another worse if the mismatch changes over time. The official guidance cited here does not provide a universal sync-offset procedure for prerecorded playback, so use the documentation for your editor or encoder and verify any adjustment against the entire relevant section. If the source is a supplied recording, ask for a corrected file when you cannot confirm its timing or make a reliable edit.
For a loop or playlist, check each file and the transitions separately. One clip may be aligned while the next begins with a different offset, or a black transition may make it harder to identify the matching sound. Our guide to looping short clips without black frames addresses continuity between clips; it is still worth checking the audio and picture at each join in your own test.
Review encoder settings and test again
Once you have located the first checkpoint where sync changes, review settings that apply to that path. YouTube's encoder settings guidance lists AAC or MP3 audio and recommends a sample rate of 44.1 kHz for stereo and 48 kHz for 5.1 surround. These are documented encoder recommendations, not a promise that changing sample rate will fix a particular mismatch. Confirm that the audio format and channel layout match the material and encoder configuration you intend to use.
The same YouTube page recommends a two-second keyframe frequency and says not to exceed four seconds. These are live-encoder recommendations to check against the intended stream format; they are not a blanket remedy for audio timing. Avoid changing unrelated settings all at once. If you change a relevant setting, write down the original value, run the same representative test, and compare the same event in the preview and local recording before checking the received stream.
If encoder diagnostics point to performance trouble, investigate performance. OBS's encoding performance troubleshooting guide describes GPU overload or bottlenecks as possible causes of encoding performance problems and recommends reducing scene or rendering load. YouTube also recommends checking encoder CPU load. These checks are conditional: a rendering or encoding warning gives you a reason to reduce unnecessary scene work and retest, but a sync symptom alone does not establish a CPU or GPU bottleneck.
For connection indicators, OBS's stream connection troubleshooting guide discusses dropped frames and connection stability. Check the outbound connection if the local output is healthy and those indicators show a problem. If a different encoder is available, YouTube suggests trying one as a diagnostic when local output has problems; the result may help narrow the path, but does not itself explain the original fault.
Run another realistic test after any change. Include sound and movement similar to the actual broadcast, inspect a local recording, and monitor stream health and messages. Change one relevant thing at a time so you can tell what the test did and did not show. If the mismatch remains, restore settings that were not implicated and continue at the earliest faulty checkpoint rather than stacking speculative adjustments.
Keep upload and live-stream issues distinct
A regular YouTube upload has no encoder preview or live-stream health status to inspect. Compare the exported source with the published video at the same identifiable events, using more than one playback device or player where possible. If the source is already out of sync, correct or replace that export. If it is aligned locally but the published playback appears different, document the exact segment and playback conditions, then consult current YouTube Help for the upload workflow. Do not borrow live-encoder troubleshooting as though it proves how uploaded-video processing behaves.
A prerecorded file sent through OBS to YouTube Live does have a local production and delivery path. Use the checkpoints in order: source, local archive, preview, then stream or replay. This approach avoids two common dead ends: changing upload settings for a live encoder fault, and changing a live encoder for a file that was already misaligned before broadcast.
If you are deciding how to operate a continuous channel after finding a local-computer problem, first separate the sync issue from the operating choice. StreamNeo can remove the need to leave your own computer playing the file overnight: you upload the video, provide your YouTube stream key, and the broadcast runs with monitoring and automatic restart if it drops. It does not change a sync fault already present in the source, and it is for YouTube only.
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
How do I tell whether the source file or OBS is causing the sync problem?
Play the source locally, then compare the same event in the OBS preview and a local recording. If the source is already wrong, start with the file; if it first appears in the preview or archive, inspect the local production path and encoder diagnostics. If those checks are healthy, move on to stream health and delivery.
What if the audio is aligned at the start but drifts later?
Record where the mismatch first becomes noticeable and compare the same span in the source, archive, preview, and received stream. A growing mismatch is different from a constant offset, but the pattern alone does not identify the cause or dictate an adjustment. Test changes against the full relevant section rather than one moment.
Will changing the audio sample rate fix every mismatch?
No. YouTube lists 44.1 kHz for stereo and 48 kHz for 5.1 surround as encoder recommendations, but those values are not a guaranteed cure. First identify where the fault begins, then check that your format and settings match the workflow.
Does this advice apply to a normal YouTube video upload?
Only the source-file comparison applies directly. A normal upload does not have the live encoder preview and stream-health checks described here, so keep it separate from a prerecorded file sent through OBS to YouTube Live. Consult current official YouTube Help for the upload workflow rather than assuming the live-stream guidance covers it.