Why Your Phone Can't Be a Karaoke Mic

I don't sing. I was at a karaoke room recently, not singing, and had one of those idle thoughts you get when you're not the one performing: why does this need a dedicated mic at all? Everyone in the room already has a phone. Just talk into that.

The Idea

Simple idea. App on your phone, connects to the karaoke speaker, amplifies your voice. I thought I'd found a clean little product idea. Then I actually thought about how sound gets from a phone to a speaker, and the idea started falling apart in a way that was more interesting than the idea itself.

The One Number That Matters: Latency

The whole thing hinges on one number: latency. If there's a gap between when you speak and when the speaker plays it back, you hear yourself twice — once in the room, once delayed through the speaker — and it messes with your ability to talk, let alone sing. Anything under about 30ms and your brain doesn't notice. Past that, it starts to feel like talking through a bad phone line.

None of the Obvious Paths Are Built for This

So I went through the ways a phone could actually get audio to a speaker, and none of them are built for this.

Bluetooth is the obvious first guess, and it's the worst one. BT audio profiles are designed for the phone to send audio to headphones or receive it from a headset — not to stream a live mic to an external speaker as a source. Even where it technically works, the codec encoding and buffering adds 150-300ms. You'd hear yourself with a full-on echo.

AirPlay and Chromecast are worse. They're built for media playback, which means they buffer aggressively to stay in sync with video — 1-2 seconds of delay. Completely dead on arrival for anything live.

Wi-Fi, streamed to a custom receiver app, can get better — 50 to 300ms depending on how it's built. Still noticeable, still off, but at least in the realm of "usable if you don't mind it."

Wired is the only option that actually works well. USB-C out, digital signal, 5-20ms. That's the only path that feels like a real mic. Which means the "just an app" idea quietly turns into "an app, plus a cable, plus figuring out how that cable gets to the karaoke speaker across the room."

The Phone Itself Is Part of the Problem

And even wired, the phone itself is a liability. Capturing audio through the phone's own mic means the signal has to go through the OS audio stack before it goes anywhere — buffering, processing layers, more buffering. iOS keeps this reasonably tight, around 10-15ms. Android is inconsistent, anywhere from 20ms to 100ms+ depending on the device, because most apps don't bother using the low-latency audio paths that exist. So the same app could perform differently well on an iPhone versus an older Android device, and that's not something you can fully fix from the software side.

Why Real Wireless Mics Don't Use Bluetooth or Wi-Fi

This is basically why actual wireless mics don't use Bluetooth or Wi-Fi at all. They use a dedicated point-to-point 2.4GHz radio link — no protocol overhead, no negotiation, just raw audio frames flying straight from transmitter to receiver. Total latency: roughly 4-8ms. It's a solved problem — it's just not solved by anything a phone comes with built in.

Reflection

Which is the part that killed the "just an app" version of the idea for me. To get this to actually feel good, you stop building software and start building hardware: a transmitter, a receiver, a radio link. At that point you haven't built an app that turns your phone into a mic — you've built a wireless mic that happens to let your phone hold the settings menu. The phone becomes a remote control, not a mic. Different product, different price, different pitch entirely.

So the idea survives, just not in the form I first had it in the karaoke room. "Everyone already has a mic in their pocket" is true and also irrelevant, because the thing that makes a mic good was never the capsule — it's the milliseconds between your mouth and the speaker, and a phone was never built to be fast at that specific job.

EGON