Signal Chain

The Real Cost of “Good Enough” Audio

Most audio gets approved on laptop speakers. Somebody plays back the recording, the words are all there, and it gets a thumbs up. Then it goes out into the world, where people listen in the car, on earbuds at the gym, or through a ceiling speaker in a conference room. That’s where “good enough” stops being good enough.

The problem with bad audio isn’t that people complain about it. Most of them never will. They just tune out, trust the speaker a little less, and don’t come back.

Bad sound makes you less believable

This one surprises people. In 2018, researchers Eryn Newman and Norbert Schwarz played the same conference talks and the same NPR interviews to listeners, some in good audio quality and some in poor. The content was identical. Only the sound changed. The people who heard the worse audio judged both the research and the researchers less favorably.

The explanation is processing effort. When something is hard to listen to, the brain quietly turns that effort into a judgment about the content. Nobody thinks, “The mic was too far away.” They think, “I’m not sure about this guy.”

That should worry anyone whose recording is supposed to sell something, teach something, or ask for trust: a CEO’s town hall, a pastor’s sermon podcast, a product demo, a fundraising video. The words can be perfect and still lose the room.

What “good enough” actually introduces

Bad audio usually isn’t one big failure. It’s a handful of small problems that stack up.

  • Room echo. A laptop mic three feet away in a drywall office records the room as much as the voice. Consonants smear, and the listener works harder to understand every sentence. This is the most common problem I hear, and the easiest to prevent.
  • Noise, and the cleanup artifacts that replace it. HVAC hum, a fridge, traffic. Noise reduction can remove it, but pushed hard it leaves a watery, underwater sound that’s often more distracting than the noise was.
  • Levels that wander. One person is loud, the other is quiet, and the listener rides the volume knob the whole time. In a car, they give up.
  • Clipping. If the input was too hot and the peaks distorted, that part of the recording is gone. You can soften it later. You can’t un-break it.
  • Processing on top of processing. A meeting app applies its own noise suppression and compression. Then the editor adds more. Then the streaming platform encodes it again. Each step is reasonable on its own. Together they leave a voice that sounds thin, pumped, and tired.
  • Listening fatigue. None of the above has to be bad enough to notice consciously. It just has to make the listening a little harder, for long enough that people stop before the end.

Why audio gets less forgiveness than video

People will sit through a soft picture, a shaky camera, or a webcam with bad lighting. They leave when they can’t follow the words. In most presentations, podcasts, and services, the audio is the content. The video is there to support it.

That’s also why audio is often the better place to spend a limited budget. A modest improvement in sound does more for the audience than a modest improvement in picture.

The fixes, cheapest first

You don’t need a studio to get out of “good enough.” Most of the improvement comes from a few habits, and the first ones are free.

  • Get closer to the mic. Distance is the biggest variable you control. Halving the distance between mouth and mic does more than doubling what you spent on the mic.
  • Fix the room before buying gear. Record in the smallest, softest room you have: carpet, curtains, a couch, a closet full of clothes. Avoid glass, bare walls, and kitchens.
  • Use a decent dynamic mic. A solid dynamic mic, worked close, rejects a lot of room sound. It’s more forgiving than a condenser in an untreated space.
  • Leave headroom. Set levels so normal speech sits well below the top of the meter and the loud laugh still has room. Quiet and clean can be fixed. Clipped can’t.
  • Record each voice on its own track. If two people share one mic or one mixed track, every problem with one of them becomes a problem with both.
  • Listen on headphones, then on the worst speaker you own. Headphones show the noise and the edits. A phone speaker or a car shows whether the voice still holds up.
  • Don’t lean on “fix it in post.” Modern cleanup tools are impressive, and they still work best on recordings that didn’t need much fixing.

When to bring in a pro

If the recording is a one-off internal update, the list above will get you most of the way. Bring in an audio engineer when the stakes are higher: a show you publish every week, a message you want people to trust, a brand you want to sound the same everywhere it appears, or anything going in front of an audience that didn’t have to choose you. That’s when the cost of sounding “fine” is higher than the cost of doing it right.

The short version

Bad audio doesn’t sound like a technical problem to your audience. It sounds like you. Get close to the mic, pick a better room, leave some headroom, and check it on something other than your laptop.

Source: Eryn J. Newman and Norbert Schwarz, “Good Sound, Good Research: How Audio Quality Influences Perceptions of the Research and Researcher,” Science Communication 40, no. 2 (2018): 246–257.