People assume broadcast sound comes from expensive microphones. It mostly comes from three things that cost nothing: putting the microphone in the right place, closing it when nobody is talking, and controlling the level. Get those right with a cheap microphone and you will beat an expensive one used badly.
Spend your money in this order
- A basic lavalier or a USB dynamic microphone. This is the only thing worth buying at first.
- A stand or clip that puts it close to the mouth.
- A cable long enough that nobody trips on it.
That is the hardware. Everything after this is free.
Placement beats everything
Every doubling of distance takes a large bite out of the signal and brings the room up with it. A microphone at fifteen centimetres sounds professional. The same microphone at a metre sounds like a conference call.
A lavalier belongs a hand-width below the chin, clipped to something that will not rustle. A desk microphone belongs a fist away from the mouth, slightly off-axis so breath does not hit it directly.
If you find yourself pushing gain very high, do not. Move the microphone closer instead. High gain amplifies the room as much as the voice.
Close the microphone when nobody is speaking
An open microphone is a live feed of your room. Multiply that by the number of microphones and the room becomes louder than the person talking.
A noise gate closes each channel below a threshold and opens it the instant someone speaks. Set the threshold just above the room noise and set the release slow enough that it does not chop the ends of words.
On a panel, automatic microphone mixing goes further — it gives the gain to whoever is actually talking and closes the rest as it goes. On six microphones this is the difference between usable and unusable.
Control the level so the viewer does not have to
Three tools, in order. A compressor evens out the moment-to-moment swing when someone leans in and out. A leveller catches the sudden spike when somebody shouts or laughs into the microphone. Automatic gain control holds the overall level over minutes rather than syllables.
Set the compressor gently — a ratio around 3:1 and only a few decibels of reduction on the loudest moments. Heavy compression makes the room noise louder along with everything else.
Fix the room before you fix the sound
A hard room puts the same voice into the microphone several times. No processing removes reverb convincingly once it is recorded.
Anything soft helps and costs nothing: curtains, a rug, a bookshelf, people. Record a test in the room and listen on headphones — you will hear the room far more clearly than you do standing in it.
The one check nobody does
Listen back on phone speakers. That is what most of your audience will use, and it is unforgiving. Anything muddy, boomy or quiet shows up instantly there and nowhere else.
Where the software does the work for you
All of the above is technique you can apply anywhere. What removes the effort is having the processing already on every channel rather than assembled from plugins.
Skynat Live puts the full chain — gate, voice isolation, four-band EQ over a live spectrum, compressor, leveller and automatic gain — on every input with defaults chosen for speech in a room. Voice isolation in particular will strip a hum or a buzzing cable that no EQ setting removes cleanly, and Auto FX will thicken a thin singing voice.
It is the same result the technique above gets you, without needing to know any of it first.