DAW plays through your audio interface, but the browser is silent

A DAW playing through a USB audio interface does not guarantee the browser uses the same output. Keep your monitors low, play a short browser reference, then check the operating system's selected output and any app-specific route. Change one setting at a time. The goal is to find which part of the path differs, not to reset everything that makes the DAW work.

Write down the working path

In the DAW, note the selected audio device and output pair. Note where your headphones or monitors are connected on the interface, and whether the sound you hear is DAW playback or direct monitoring of an input. Direct monitoring can let you hear a microphone or instrument even while computer playback is misrouted. Focusrite's direct monitoring explanation describes that distinction for its devices. Other interfaces use different controls, so check the manual for yours.

Open the WaveLocus speaker test on the same computer. Select a comfortable low physical volume and play its ordinary reference. Watch for the on-page playback indicator. If it says the signal is active but you hear nothing, you know only that the webpage started its signal. The indicator cannot read the signal at the interface's headphone jack.

Compare three outputs, one at a time

  1. Keep the browser page fixed. Test the same WaveLocus reference twice. Note whether the operating system's default output names the USB interface or built-in speakers.
  2. Check the browser's app route. On Windows, open the app volume and device preferences. Microsoft's app-audio troubleshooting explains that an individual app can have its own output choice or volume even when system sounds work. Check whether the browser is muted or sent somewhere else.
  3. Change only that route if it differs. Select the intended interface output, return to the same page, and replay the same short reference. Record whether it became audible.
  4. Check the interface's playback channels. If the system and browser target the interface but sound is absent, inspect the interface maker's mixer or output-routing instructions for its exact model. Do not assume the DAW output pair and system playback pair feed the same physical socket.
  5. Compare a second computer app. Play a short familiar system sound or local file through the same interface at low volume. Keep the browser, interface connection, and monitor level unchanged. This separates a browser-only difference from a wider system-playback route.

Interpret the first clear difference

ObservationNext inferenceStill unknown
Browser plays on built-in speakers, DAW on the USB interfaceThe two apps are using different outputs.Whether the interface is faulty.
Browser is muted in app settingsRestore its app volume and retest the same reference.Why that setting changed.
Browser and system apps silent through interface; DAW playsCheck system playback pair and interface mixer routing.Which stage is responsible without another comparison.
Browser silent only when DAW is openCompare with the DAW closed, then review device sharing and driver settings documented for your system.A universal exclusive-mode cause.

Do not confuse input monitoring with playback

If you hear a live microphone through the interface, you may be listening to its hardware monitor path rather than sound returned from the computer. Verify using a recorded DAW track that you know is actually playing back. If that track plays but the browser does not, check the interface's software mixer or routing matrix. Focusrite's routing support note also shows that available routing controls differ by product. A generic instruction to “change loopback channels” is not useful if your model does not expose them.

Browser output selection has limits

Some browsers and pages can request a particular output device through APIs such as AudioContext.setSinkId(), subject to browser support and permissions. Do not assume every page offers that control or that it overrides your operating system's app route. WaveLocus's ordinary speaker reference uses the current browser playback path. Check the system's actual output choice instead of looking for a selector that this test does not provide.

On a managed workstation, audio policies or driver controls may be locked. Do not install unknown drivers or disable system protections just to make one test work. Use the interface vendor's instructions and your administrator's support path when the route is controlled centrally.

Keep a useful support record

Write down the operating system, browser, interface model, DAW output pair, system output, interface monitor socket, and results of the same short browser reference before and after one change. Avoid uploading a private project or audio recording merely to prove the issue. If closing the DAW changes the browser result, record that exact transition and whether the same interface remained selected. Then take the observation into guided diagnosis.

A browser-only failure is not an electrical loopback measurement. The electrical loopback tool requires a suitable physical return path and appropriate level protection. It will not automatically explain why Windows or macOS routed a browser to a different device. Keep routing diagnosis and hardware measurement separate until you have a reason to connect them.

When to stop

If changing an interface routing control produces unexpectedly loud sound, stop playback and restore a safe monitor level before continuing. Do not unplug speakers or insert cables while a test signal plays. If the output still differs after a documented route check, preserve the evidence and contact the device maker or local support with the exact one-factor comparisons. “DAW playback works on output 1–2; the same browser reference stays silent when system output is the interface” is a much stronger report than “the interface is broken.”

Sources and limits

Microsoft documents app-specific audio routes and volume. Focusrite documents direct monitoring and model-dependent routing. MDN documents browser output-device API limits. These sources explain possible differences in a playback chain; they do not diagnose your interface. The method here uses listening and visible settings, not a calibrated electrical measurement.