Test your webcam and mic before an interview

A ten-minute check of camera, lighting, microphone level, background noise and headphones before a video interview, with the numbers that tell you it is fine.

On this page

The first minute of a video interview sets the tone, and too many of them are spent on "can you hear me?". Almost every problem in that minute can be found the evening before: a camera another app is holding, a microphone that is not the one you think, a window behind you, a laptop fan, headphones plugged in the wrong way. This is a ten-minute routine with our webcam and microphone tester, with the numbers that tell you each part is fine. Do it in the same room, at the same time of day, with the same headset you will use.

1. Permissions and the right devices

Open the tester, press Start camera and allow access. If the browser says access was blocked, there are two switches, not one: the site permission (the camera icon in the address bar) and, on macOS and Windows, the system privacy setting that can block the whole browser. If it says the device is busy, another app has it open: close the call apps, not just their windows.

Once access is granted, the device list shows real names. Laptops with a docking station often list two or three cameras and microphones; pick the one you will use and remember its name, because the interview platform will ask you again.

2. What the camera really sends

The panel shows the live resolution and frame rate. The tool asks for 1280 × 720 to begin with, which is what most call apps send anyway. Then press Test resolutions: it asks for 360p, 480p, 720p, 1080p, 1440p and 4K one by one. A typical built-in laptop camera answers something like this:

SizeResult
360p (640 × 360)Delivered exactly or refused, depending on the camera
480p (640 × 480)Delivered exactly
720p (1280 × 720)Delivered exactly
1080p (1920 × 1080)Delivered exactly on newer cameras
1440p, 4KRefused

Do not worry about the refusals: video platforms adapt the size to the connection and rarely send more than 720p or 1080p. What matters is that 720p works and that the measured frame rate is close to the set one. If the camera is set to 30 fps but delivers 15, the usual cause is dim light: the camera lengthens each exposure. More light fixes both the picture and the frame rate.

3. Light and colour

The lighting check averages the frame once a second. Aim for no warnings at all. The common ones and their fixes:

  • Too dark (average brightness under about 27%): put a lamp or the window in front of you, not behind.
  • Washed out or bright areas blown out: you are facing direct sun, or a bright window is in the frame. Close a curtain or turn a little.
  • Warm or cool cast: mixed light, for example a yellow lamp on one side and daylight on the other. Use one kind of light if you can.

The check cannot tell a colour cast from a coloured wall, so a strong colour behind you will trigger it. Use the snapshot button to save a frame and look at it at full size: that is what the interviewer sees, minus compression. Keep the camera at eye level, and remember that the mirror switch only changes your preview; the other side always sees you unmirrored.

4. Microphone level and background noise

Switch to the Microphone tab, choose the microphone and press Start microphone. Stay silent for five seconds, then talk as you would in the interview. Three numbers matter:

  • Peak: your speech should peak somewhere between −20 and −6 dBFS. Peaks at 0 dBFS distort; peaks below about −30 dBFS are quiet and the platform's automatic gain will raise the noise along with your voice.
  • Noise floor: the level of the quiet moments. Around −60 dBFS or lower is a quiet room; around −45 dBFS means a fan, traffic or hum that the other side will hear.
  • Speech above noise: the difference. Take the example in the tool's own tests: quiet moments between −70 and −60 dBFS and speech reaching −18 dBFS. The floor comes out at −68 dBFS and speech sits 50 dB above it, which sounds clean. Under about 30 dB, find a quieter spot or move the microphone closer.

Try the processing switch both ways. Most call apps apply echo cancellation, noise suppression and automatic gain; turning them off shows what the microphone really picks up. If the raw sound is full of fan noise, the suppression will be working hard and your voice may sound watery: better to remove the noise.

5. Hear yourself

Press Record 5 s and play back with headphones on. Listen for three things: the room's echo (hard walls make you sound distant: a sofa, curtains or a bookshelf help), plosives (pops on "p" and "b" mean the microphone is in front of your mouth: move it to the side), and the keyboard. For a longer rehearsal, the voice recorder records several minutes.

6. Headphones and speakers

In the Speakers tab, play the left tone, then the right. Each must come only from its side. A tone in both ears when you chose one side means the headset is in a mono or "hands-free" mode, which on Bluetooth often also means low-quality audio while the microphone is in use. Bluetooth headsets switch to that mode whenever their own microphone is active; a wired headset or a separate microphone avoids it.

Common mistakes

  • Testing in a different room or at a different time: the light and noise will not match.
  • Leaving another call app running, which keeps the camera busy or grabs the microphone.
  • Trusting the camera's box: "1080p" on the box does not mean the platform will send 1080p.
  • A window behind you: the camera exposes for the window and your face goes dark.
  • Using laptop speakers with the laptop microphone: the other side hears an echo of themselves. Headphones fix it.
  • Forgetting the system settings: an input volume at 10% in the operating system makes every app quiet.

Ten minutes the evening before is cheaper than the first five minutes of the interview.

Tools mentioned in this article

Back to the blog