First find out where the sync error starts. If audio and video are already out of step in a local recording, investigate the phone, audio input and capture path; if the local result is aligned but YouTube’s Live Control Room preview is not, check the outgoing stream and its delivery path.
There is no single Larix setting that reliably fixes every sync problem. Make a controlled test, note whether audio leads or trails and whether the gap stays steady or grows, then change one relevant thing at a time.
First identify which way the error goes
A sync complaint can describe two different things: sound arriving before the matching picture, or picture appearing before its sound. Those observations do not, by themselves, tell you which component is at fault. They do give you a consistent symptom to check at each stage.
Use an obvious cue, such as clapping once where the camera can see your hands. In a recording, compare the moment the hands meet with the sound of the clap. For speech, watch a word with a clear consonant, but a clap is easier to judge because it gives you one visible and audible event rather than a sequence of syllables.
Write down what you observe in plain terms: “audio leads” or “audio trails”. Also note whether the gap appears roughly steady or becomes more noticeable as the test continues. You do not need to estimate milliseconds. A rough impression such as “the clap matches at the start, then drifts apart” is more useful than a made-up offset value.
Repeat the cue once or twice during the same test. A single moment can be misleading if you are watching a delayed preview or if the cue itself is hard to see. Keep the phone, input, connection and Larix profile consistent between tests; otherwise it becomes difficult to know which change affected the result.
Check a local recording or preview
Before changing stream settings, establish what the phone itself is producing. Make a brief local recording using the same camera, microphone or external input, and any processing you normally use. If Larix provides a local preview for your workflow, observe that too, but do not treat a preview and a saved recording as interchangeable evidence: they may involve different steps.
Play the recording back on the same device first, then, if practical, on another device. A playback device or application can make an aligned file seem wrong, so compare the recording with care. The key question is whether the clap or spoken cue is already out of step in the saved result.
If local audio and picture are misaligned, concentrate on the source side before investigating YouTube. Check which input is selected in Larix and whether the phone is using its built-in microphone, a wired or wireless microphone, a USB audio interface, or another route. Confirm that the video and audio are both coming from the devices you intend. An external input can add a separate path to check, but its presence alone does not prove it is the cause.
If the local recording is aligned, keep it as a baseline. Do not immediately change the audio offset, frame rate, protocol and bitrate together. A clean local result narrows the investigation towards the broadcast path, but it does not prove that every later stage is aligned.
For a channel that combines speech and recorded material, keep a known-good test clip alongside your usual setup. This gives you a repeatable way to compare the result after a change. The practical value is similar to keeping a stable programme source for a longer show, as discussed in this guide to keeping a podcast stream running continuously.
Test capture, audio input and device encoding
When the local result is wrong, simplify the source path. If you are using an external microphone or audio interface, make one test with the phone’s built-in camera and audio input. Temporarily remove optional audio processing or adapters, where your setup allows it, and repeat the same cue. This is a diagnostic comparison, not a claim that the built-in microphone is inherently better.
If the simple setup is in sync, add your usual components back one at a time. Test the external microphone first, then any adapter or processing step, so the point where the mismatch returns is visible. If it is still out of sync with the simplest setup, restore your normal configuration and note the phone model, operating system and Larix version. That evidence is more helpful than trying random settings.
Check the device’s capture and encoding load as well. Larix uses the device’s system encoder, and available frame rates can vary by device and platform. If the phone is also overheating, recording other media or running demanding background tasks, repeat the test with fewer competing tasks. This is a way to test whether device conditions matter, not a diagnosis based on phone model alone.
Softvelum explains that Larix’s interleaving buffer gathers audio and video frames for receiver-side processing, while also describing latency as a chain that includes capture, encoding, buffering, transport and platform delivery. That buffering can help organise frames; it does not guarantee that every downstream step will preserve sync. See the Larix FAQ for its explanation of buffering, frame rates and delivery behaviour.
Avoid purchasing a microphone or adapter just because a stream is out of sync. Consider new hardware only if tests isolate the issue to the existing audio-input path and you have checked compatibility with the phone and app. If removing an external device makes no difference, buying a replacement is not supported by the evidence from your test.
Compare the outgoing stream with YouTube’s preview
If the local result looks right, run a short private or unlisted test and compare it with the preview in YouTube Live Control Room. Use the same phone, Larix profile, audio input and network connection as the affected broadcast. YouTube’s encoder setup instructions describe entering the server URL and stream key and checking the Live Control Room preview before going live.
Watch for your cue in the preview. Note whether the same mismatch is present there, whether it is reversed, or whether the preview looks aligned. The aim is not to decide that YouTube is “the cause” from one observation; it is to identify whether the first visible failure appears before or after the local result.
If the preview is aligned but someone watching the public stream reports a problem, check the viewer’s playback conditions before changing the source. Ask whether the issue occurs on another device or connection, and whether the viewer is describing live playback or a replay. Different playback paths can present a discrepancy that is not present in the encoder preview. Confirm the symptom with a controlled test where possible.
Do not use a viewer’s report alone to apply a fixed delay to the whole channel. A correction that suits one playback condition can make the stream worse elsewhere. Keep the original local recording and a note of what the preview showed so you can compare future tests rather than relying on memory.
Inspect the ingest path when local output is aligned
If the local recording is in sync but the Live Control Room preview is not, verify the broadcast setup before altering timing values. Confirm that the stream URL and stream key currently shown in YouTube are the ones entered in Larix, and check that the encoder is sending the intended audio and video. YouTube treats the server URL and stream key as distinct parts of encoder setup; check both rather than assuming a successful connection means every field is correct.
Next, review the connection and delivery conditions. Note whether you are on Wi-Fi or mobile data, whether the connection is fluctuating, and whether the mismatch is steady or gets worse over time. Avoid changing network type, protocol and bitrate all at once. A brief test on a more stable connection can help test a network hypothesis, provided you keep the other parts of the setup the same.
Larix documents different buffering behaviour across connection protocols and notes that network delivery problems can increase delay. It also discusses adaptive bitrate as a response to delivery conditions, not as a universal correction for audio and video offsets. A bitrate or adaptive-bitrate test may be reasonable when the evidence points to changing network delivery, especially if sync worsens over time. It is not a general sync switch, and a fixed local mismatch calls for looking elsewhere.
For a long-running radio or music channel, keep the audio source and its processing consistent during tests. If YouTube loudness or downstream processing is already part of your investigation, separate that from timing: changes intended to make sound levels consistent are not automatically fixes for sync. The distinction matters when troubleshooting a playlist-based channel, as in this article on consistent radio-stream sound when YouTube changes volume.
Make one change at a time and retest
A useful test changes one relevant variable and repeats the same cue. Keep a short log with the date, phone and OS, Larix version, input selected, connection, any setting changed, and what you saw locally and in YouTube’s preview. Include whether audio led or trailed and whether the difference stayed steady or grew. This gives you a comparison instead of a vague recollection that the stream “seemed better”.
| What you observe | First area to check | Next controlled test |
|---|---|---|
| The local recording is out of sync | Capture, selected audio input or device encoding | Simplify to the phone’s built-in camera and audio, then add components back one at a time |
| Local recording is aligned, but YouTube preview is not | Encoder setup, outgoing media or ingest path | Confirm the current URL and key, then test connection conditions without changing unrelated settings |
| Both local recording and preview are aligned, but a viewer reports a mismatch | Playback path or viewer conditions | Compare on another device or connection before changing the broadcast |
| The gap appears to grow over time | Timing or delivery conditions that change during the stream | Compare repeated cues and note network variation and device load |
The table is a way to choose the next investigation, not a decision tree that guarantees a diagnosis. If a change makes no clear difference, revert it before trying another one. If two changes are made together, you will not know which one helped or whether the apparent improvement was coincidence.
After a targeted change, make another short test with the same setup and compare the cue at the same points. Check both the local output and the Live Control Room preview again, even if you expect the change to affect only one stage. YouTube’s workflow asks creators to inspect the preview before starting the broadcast, so do not skip that check when returning to a public schedule.
For a channel that needs to restart after an interruption, sync is only one part of a dependable overnight workflow. Keep the source file and restart plan separate from this diagnosis; a stream that resumes after a power loss still needs a fresh preview check when you change its media or input path. The guide to resuming a YouTube playlist after a power outage covers that continuity problem.
If the source is a fixed video loop rather than a live camera and microphone, check that the audio track in the file itself matches its picture before uploading or encoding it. A broadcast workflow can repeat a mismatch that is already baked into the source. The article on why many YouTube live cams are actually loops gives context for that kind of always-on format, but the local file remains the place to verify its timing.
When the cause remains unclear
Do not guess at a millisecond offset if the evidence does not support one. The sources do not provide a universal Larix-to-YouTube sync recipe or an offset value that applies to every phone, input and delivery path. A setting that appears to help one combination may be wrong for another, particularly if the error grows rather than staying steady.
Collect the details needed for a useful support request: phone model, operating system, Larix version, selected audio source, connection type and protocol, and whether audio leads or trails. Include whether the error is present in a local recording, first appears in the YouTube preview, or is only reported by a viewer. Say whether it is constant or changes over time, and describe the single changes you tested and their results.
For an unresolved Larix-side question, use Softvelum’s official documentation and support route rather than relying on a setting copied from an unrelated device. For YouTube encoder setup, consult YouTube Help. Check the current official pages when troubleshooting because app and platform workflows can change. Neither a successful connection nor a clean preview guarantees that every viewer’s playback will match.
When a repeated video channel is difficult to keep running because the source computer has to stay on, StreamNeo can remove that specific computer-running burden: you upload the video, provide the YouTube stream key, and the stream can continue with your computer off. That does not diagnose or correct an audio/video mismatch in a Larix broadcast, so finish the sync test on the actual file and channel you plan to use.
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
Should I change a Larix audio offset setting first?
No. First find out whether the mismatch is already present in a local recording or only appears in YouTube’s preview. Without that comparison, an offset change may hide one symptom while leaving the source or delivery issue unexplained.
What if audio is in sync at first but drifts later?
Repeat the same visible and audible cue at more than one point and note whether the local recording drifts too. If only the live preview changes, test network and delivery conditions one at a time; adaptive bitrate is not a guaranteed correction for drift.
Is a new microphone the quickest fix?
Not unless a controlled test points to the existing input path. Try the built-in camera and audio input as a comparison, then add external devices back individually; buy compatible hardware only if the evidence isolates the problem there.
What should I send to Larix support?
Include the phone model, operating system, Larix version, audio source, connection and protocol, and whether audio leads or trails. Also say where the mismatch first appears, whether it is steady or grows, and what you changed during repeat tests.