How do I set up audio for live streaming? Start with the microphone or other audio device, set its input level there, then confirm the signal in OBS and listen to what OBS is sending. Add routing or processing only when your production needs it.
A USB microphone may be enough for a speaker and a small number of sources. A separate interface or DAW route can help when you need XLR or instrument inputs, independent sources, direct monitoring or loopback. The dependable choice is the simplest path you can test from source to stream.
Map the audio signal path
Before changing settings, write down what viewers should hear and where each sound begins. A spoken programme might need one microphone. A music stream could have a microphone, a DAW mix, playback from another application and alerts. Each source needs a clear route into the final mix; otherwise, you may miss a source or capture it twice.
Think of the path in stages: source, device input, any interface or software routing, OBS source, OBS mixer and the stream output. At each stage, ask two questions: can you see or hear the signal, and can you adjust its level independently? If the audio disappears, those checkpoints help locate where it stopped. If it sounds too loud, they help identify where the level became excessive.
For a simple arrangement, connect a USB microphone to the computer and select it as an audio input in OBS. You do not need to add an interface just because a stream is live. An interface becomes useful when you need the inputs, outputs, monitoring or routing it provides. The right answer depends on the sources in your production, not on a universal ranking of equipment.
| Workflow | When it can fit | What to account for |
|---|---|---|
| USB microphone into the computer | One speaker and few other sources | A straightforward path, but fewer separate hardware inputs and routing choices. |
| XLR microphone into a USB audio interface | You need an XLR or instrument input, interface controls or direct monitoring | You must select the intended interface input and connect the microphone to it. |
| DAW audio routed to OBS | A music mix or other DAW output must join the stream | The route depends on the interface, DAW, operating system and software configuration. |
| Two interfaces in a loopback workaround | A specific setup lacks a suitable loopback route | It adds hardware and cabling; it is not a default requirement. |
Ableton’s guide to setting up an audio interface describes common interface uses such as multiple inputs and outputs and direct monitoring. PreSonus documents particular Studio One routing approaches, including loopback where supported and a two-interface method in its Studio One and OBS guide. Treat those as vendor examples, not steps every streamer must follow.
If you are sending a recorded music programme rather than producing a live mix, distinguish the file’s audio from any live microphone or application sound. The routing problem is different when a continuous playlist is the main source; the guide to streaming classical music with a still image is a relevant example of that format. Whatever the content, keep a written list of intended sources before you configure them.
Set levels at the microphone or device
Begin at the source. OBS’s Audio Mixer Guide advises starting at the device in question. In practice, speak or play at the level you expect during the stream, and use the microphone, interface or device’s input control to establish a usable signal before leaning on a software fader.
Watch the device’s own meter if it has one, and listen through its headphone or monitor output if available. A meter that barely moves can mean the input level is too low; a signal that reaches the limit or sounds rough may be overdriven. Lowering an OBS fader later does not undo distortion that has already happened at the microphone input or device stage.
Then check the same source in OBS. Speak at normal volume, and include the louder moments you expect: a sung phrase, a strong instrument note, a raised announcement or an alert. The OBS meter should respond consistently. Avoid using the meter as the only check; a signal can appear present while a cable, mic position or source selection gives you the wrong sound.
The final digital ceiling is 0 dBFS. OBS’s Audio Mixer Technical Details explains that this is the maximum representable level; audio at or above it in a recording or stream can distort. Keep the final mix below that ceiling, including during the loudest expected passages. Do not target the ceiling as if it were a level to hit.
A single microphone connected through an interface may arrive as a stereo input with audio on one side, depending on the device and its configuration. Focusrite’s OBS setup instructions describe selecting the interface as an Audio Input Capture source and setting a single mic input to mono in Advanced Audio Properties so it is centred. That is a practical example for that case, not a rule that every source should be mono. Check the channel layout and listen to your own signal.
Choose inputs and routing in OBS
In OBS, add or select the source that corresponds to the physical or software input you intend to stream. A USB microphone should appear as its own capture device; an interface microphone should come from the interface input you actually used. Give sources recognisable names, such as “Voice mic” or “DAW return”, rather than leaving several indistinguishable device labels in a busy scene.
For every source, confirm that the correct device is selected and that its meter moves when you make sound. If your interface has more than one input, speaking into one microphone does not prove you chose that input in OBS. Test the intended socket, then mute or disconnect one source at a time if you need to identify which OBS channel is carrying it.
Keep voice and programme audio separate when you need to control them independently. For example, a presenter may want to lower a music bed while speaking without reducing the microphone. If the mic and music have already been combined in another application, OBS may receive one mixed source, leaving fewer independent controls. That can be perfectly workable when the mix is deliberate; it is less convenient when you need to correct one component quickly.
A DAW introduces another route to verify. It may send a stereo output to a supported loopback input, while the microphone remains a separate OBS source. PreSonus’s Studio One example shows one way to arrange that, but available paths vary with software and interface. If you follow a vendor-specific guide, use its exact model, operating-system and application context as a boundary; do not assume the same labels or virtual inputs exist on another system.
Avoid capturing one sound by two routes at once. A microphone might reach OBS directly through an input source and also through desktop or application capture. If both are live, the result may sound louder, hollow or echoing. Mute one route temporarily and listen; restore only the path you mean to use. Keep a simple signal map beside your streaming setup so you can repeat the known-good configuration after a restart or change.
For a channel that runs continuously, the same care applies when audio is part of a longer playback loop: confirm that the audio source is actually present in the scene and that the chosen route remains available after you reopen the production software. The practical considerations in keeping a 24/7 YouTube radio stream alive when OBS updates are useful context for planning changes without leaving a channel unattended mid-session.
Monitor the signal you are sending
Monitoring has two jobs: catch problems close to the source, and confirm what is reaching OBS. If the microphone or interface offers a headphone output, listen there first. It can reveal a loose cable, a muted input, room noise or a poor mic position before you investigate software routing.
Next, use OBS audio monitoring where appropriate to hear the signal after it enters the application. Set this up carefully: monitoring can create a feedback loop if the sound from the computer’s speakers is picked up again by the microphone, or if the same source is monitored through multiple paths. Headphones are usually more practical than speakers while checking a live microphone. If you hear an echo, mute monitoring temporarily and determine whether the doubled sound is in the stream mix or only in your local listening path.
Make a short recording of a normal session and listen back before going live. Include each source viewers will hear: voice, music or DAW output, playback, alerts and any other programme audio. Listen at an ordinary volume, not only while staring at meters. Check that speech is understandable over music, the voice is where you expect in the stereo image, and no sound is missing or doubled.
This matters particularly for devotional, study and ambience channels, where a low-level bed may play for a long time and a brief test announcement is not representative. Check a normal quiet passage as well as the louder moments. If listeners are expected to hear a sustained musical or ambient source, make sure the test recording includes enough of it to reveal dropouts or an unintended second route.
A recording also gives you a reference after a change. If you move a microphone, adjust a device level or change an OBS source, record again and compare what you hear. Do not assume the stream mix stayed the same because the interface still shows activity. The objective is not a particular tone or loudness; it is a mix that remains intelligible and free of obvious clipping, silence or duplication for your intended programme.
Add processing only when needed
Filters can help solve a specific, audible problem, but they cannot repair every source issue. First correct microphone placement, input gain and routing. A compressor, for example, can reduce louder peaks, but applying one to a distorted input will not remove the distortion that was already recorded. Likewise, noise processing cannot compensate for a muted or incorrectly selected device.
OBS documents a compressor as a way to reduce louder peaks and notes that it is typically placed near the beginning of a filter chain. Its sidechain option can duck desktop audio under a microphone. That can suit spoken commentary over a music or application bed, where the bed should recede while someone speaks. It is not automatically suitable for every music mix or devotional stream; listen for whether the change feels appropriate to the programme.
An expander can reduce low-level sound more gradually than a hard gate. OBS’s Expander Filter guide describes its role, but the threshold still needs care. If it is set too aggressively, quiet words or the ends of phrases may be reduced. Test normal and quiet speech, pauses, and the softest intended source before keeping the filter enabled.
Make one adjustment at a time. Record a few representative moments before and after: ordinary speech, a quiet phrase, a louder phrase, and any music transition that matters. If the change solves the problem without making another part of the programme harder to hear, keep it. If you cannot identify a clear benefit, remove it and return to the simpler path.
Test before going live
A reliable rehearsal follows the complete chain, not just the microphone. Confirm the source device, OBS input, channel layout, mixer level and monitoring route. Then record or make a private test in the way your chosen platform and workflow allow. This article does not set destination-specific encoding requirements; check the current official guidance for the platform and any remote-guest tool you use.
Use a repeatable check rather than relying on memory:
- Speak or play into each source and confirm the intended OBS meter responds.
- Listen at the device or interface monitor, then listen to the OBS path.
- Check that mono speech is centred where that is intended and that stereo sources retain their intended placement.
- Test quiet and loud passages, music beneath speech, and any alerts or playback.
- Listen to a recording from beginning to end, noting clipping, silence, echo, dropouts or sources that overpower one another.
Keep notes of the working device names, OBS source selections and any processing you chose. If something changes, such as a USB port, interface input, scene or application output, repeat the test rather than assuming the old route still applies. For streams that play for long periods, make audio checks part of your routine after updates or configuration changes, not only at initial setup.
A one-person stream may need only a microphone and one OBS source. A music production with a live presenter may justify an interface and separate DAW route. The useful distinction is not basic versus professional; it is whether the added routing solves a real need you can monitor and maintain. If you are planning a long-running music channel, the Marathi Vitthal bhajan stream guide offers a format-specific example to consider alongside your own source map.
If your audio is a finished file rather than a live mix, you may not need a computer-based production chain running throughout the broadcast. For that particular pain of leaving a personal computer on to keep a prepared file playing, StreamNeo turns an uploaded video into a YouTube live stream after you provide your channel’s stream key, so the computer can be switched off. It is YouTube-only; it does not replace an OBS routing setup when you need live microphones, a DAW mix or other real-time sources.
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
Do I need an audio interface to stream?
No. A USB microphone connected directly to the computer can suit a speaker with few sources. An interface is useful when you need XLR or instrument inputs, more separate inputs or outputs, direct monitoring, or loopback routing.
Should a single microphone be mono or stereo in OBS?
That depends on the microphone and the way the device presents its input. For one mic on a Focusrite interface, Focusrite’s OBS guide recommends setting the source to mono so it is centred. Check your own source and listen rather than applying that vendor example to every setup.
Why does my voice sound doubled or echoing?
The same microphone may be reaching OBS through two paths, such as a direct input and desktop capture, or you may be hearing both the monitored signal and another local route. Mute one path at a time and record a test to identify whether the doubling is in the stream mix or only your monitoring.
Should I add a compressor or expander to my microphone?
Only if it addresses a problem you can hear and verify. A compressor can reduce loud peaks, while an expander can reduce low-level sound; thresholds that are poorly suited to your voice can make the result less useful. Test normal and quiet speech after each change.