Half a second. That is all it takes to lose the room. Your guest is talking, the audience hears it late, chat erupts, and there is nothing you can do because you are already live. That moment costs you more than a technical mark on a run sheet; it costs you credibility, your client's brand, and every referral that person might have sent your way.
Audio latency is invisible until it isn't. And by the time it isn't, you are already in trouble.
Let me walk you through how I actually diagnose and fix this, the way I have done it across productions ranging from NBA broadcasts to Emmy-night live events. This isn't theory. This is what works in the room, under pressure, when the red light is on.
Understand What Latency Actually Is
Latency is delay. Specifically, it is the time it takes for an audio signal to travel from its source through your chain and out the other side. Every piece of gear and software in that chain adds some amount of delay. Most of it is tiny. Tiny adds up fast.
You have analog-to-digital conversion, your audio interface, your DAW or mixing software, your encoding software, your streaming platform's own processing, and finally the viewer's playback buffer. Each one of those is a handoff. Each handoff costs time.
The problem most people have isn't that they don't know latency exists. It's that they don't know where in the chain their specific problem lives. So they start guessing, pulling random settings, restarting things, and hoping for the best. That is not a workflow; that is panic.
Step One: Isolate the Chain
Before you touch a single setting, you need to know which part of your signal chain is actually misbehaving. Here is how I do it.
- Plug a microphone directly into your audio interface and monitor the input through headphones with direct monitoring turned on. If what you hear is delayed, the problem starts at the interface or driver level.
- If direct monitoring sounds clean, open your software and monitor through it. If that is where the delay appears, your buffer size or software settings are the culprit.
- If software monitoring is fine but your stream sounds delayed to viewers, the issue lives downstream: your encoder, your platform, or the viewer's buffer.
That three-step isolation process sounds simple because it is simple. But you would be surprised how many experienced operators skip it and start randomly adjusting things before they even know where the fire is.
Step Two: Dial In Your Buffer Size
Your audio interface and its drivers communicate with your computer using a buffer. A larger buffer means more stability but more latency. A smaller buffer means lower latency but more risk of dropouts and glitches if your CPU cannot keep up.
For live streaming, you want the lowest buffer size your machine can handle without artifacts. Start at 256 samples and work your way down. On a well-configured machine, you can often reach 128 or even 64 without issues. Watch your CPU usage the whole time. If it starts spiking above 70 or 80 percent, you have gone too low for that session.
One thing that has saved me on more than one high-stakes production: close every application you do not need before you stream. Every browser tab, every chat app, every background process is competing for resources your audio chain needs. You would not run a race with ankle weights; do not run your production that way either.
Step Three: Check Your Encoder Settings
Software like OBS, Wirecast, or vMix all have audio delay controls built in. They also have audio sync offset settings that can quietly get bumped or set incorrectly during setup. Go into your audio settings and verify that no unintentional offset has been applied to any of your sources.
Also check your sample rate. Your interface, your software, and your encoder all need to agree on the same number. If your interface is running at 48kHz and your software is set to 44.1kHz, your system is doing real-time resampling on the fly. That burns processing power and introduces subtle timing drift that is genuinely maddening to track down because it does not always show up as an obvious glitch.
Step Four: Test Before You Are Live
I know this feels obvious. I will say it anyway, because I have seen it skipped by people who absolutely know better. Do a full end-to-end test stream before every production. Not a local monitor check; an actual stream to a private channel that you watch on a separate device.
Watch it with your eyes and ears together. Clap once in front of the camera, loudly and deliberately. Watch the playback. If the visual and audio clap do not land at the same moment, you have a sync issue. That test takes under two minutes and it tells you everything you need to know.
When the Problem Is the Platform Itself
Sometimes you have done everything right and the latency is still there because the platform introduces it on their end. YouTube, Twitch, and LinkedIn all add buffering by default. Most of them offer low-latency modes inside their stream settings; find those and turn them on. It does not eliminate delay entirely, but it cuts it significantly for interactive productions where viewer engagement matters.
For events where real-time interaction is critical, like a live Q&A or a ticketed virtual performance, the platform choice becomes part of the technical decision itself. Some platforms are simply not built for that use case, and no amount of encoder tweaking will overcome a platform that buffers aggressively by design.
The Real Cost of Getting This Wrong
I have seen latency problems end client relationships. I have seen them tank product launches that brands spent six figures building toward. Audiences will forgive a lot of things; bad audio sync is not one of them. It breaks immersion instantly, and once it is broken mid-show, you are not getting it back.
The fix almost never requires new gear. It requires knowing your chain, trusting your process, and testing with enough lead time to actually address what you find. That discipline is what separates a smooth production from a stressful one, every single time.
If you are producing live events or virtual experiences and want a technical workflow that holds up under real pressure, see what we offer through our production services. We work with teams at every level to tighten up infrastructure before it becomes a problem on air.