[Audio Theory] What's quieter USB or Audio Interface

Theoretical Question:

What setup is going to be quieter (have less of the “background electrical hum”)

1.) A traditional microphone (with an XLR jack) plugged into an Audio Interface, with a USB hookup to a PC.


2.) A USB Microphone–plugged directly into a PC.

In other words, will running a traditional mic through an Audio Interface be quieter than buying a USB mic? Or is there no difference.

I’m a newb.

Background noise is a big problem for me. So I’m just wondering if there is any advantage to getting a good Audio Interface, and moving it away from the PC.


The “safest” thing is a USB interface with it’s own separate power supply (not USB powered).

It varies… Noise can get-in through the USB power. It’s usually not 50/60Hz power line hum. It’s usually higher-frequency electrical noise from the switching power supplies inside the computer, or from other high-frequency signals/switching.

Some interfaces/USB microphones are more noise-immune than others, and some computers are noisier than others.

And, all preamps generate some noise (typically hiss). The preamp is either built-into the USB mic or the interface. But usually, preamp noise (noise that’s not getting-in through the USB power) is much lower than the acoustic noise in the room so it’s not a major concern…

Are you reading a theoretical audiobook or a theoretical podcast?

There are some high points.

You can make a USB microphone cheap and inexpensive by leaving out the filtering and processing for the power coming up the USB cable. You can make a computer cheap and inexpensive by leaving out the filtering and processing for the USB power. Put those two together and you can get The Yeti Curse.


It’s rough to eliminate in post production although we produced a custom filter for it. The filter, unfortunately, deletes some musical tones from your show. The only sure cure is change the microphone or the computer.

USB microphones have other problems as well. They come almost universally with soft volume. These microphones are designed for casual users. If the performer is too loud, distortion can kill the performance permanently. Full Stop. But if they’re too soft, the show can be boosted in Audacity and many times used. So, no contest, right? They all come out of the tin too quiet and usually there isn’t anything you can do about it.

USB Interfaces (I use the Behringer UM2) usually do OK, but there, too, soft volume can be a problem. Most times I use the interface adjusted all the way up. If your interface has the ability to use condenser microphones, then chances are terrific it has 48 volt phantom power and it has the filtering and processing not to have The Yeti Curse.

Mine is on the left.

The other thing good about interfaces is the ability to separate the noisy computer from you. The limit for a USB microphone is about six feet (2M). The limit for an XLR cable can be 100 feet or more.

When someone is writing me checks, I use a small analog mixer although a small USB mixer would work. This gives me the ability to adjust the sound perfectly. The only down side I can think of is the need to be there for the whole show.

This was a radio broadcast I shot. The picture is two shoots, so there’s many more wires than there needs to be.

There is another option. One of our posters gave up on all the computer nonsense and is using a Zoom H2n stand-alone recorder. The daddy to that recorder (H2) is on the stand in this recording.


I have a Zoom H4.
No computer anywhere near, it has headphone monitoring and it’s portable.


What’s the job?

Please note the dark blue furniture moving pads in each picture, including the conference room. The conference room is totally soundproofed, but that pad is there to get rid of wooden table slap and honk audio damage.

As DVDdoug above, internal microphone noise is not what normally kills you. It’s the metrobus outside or the wooden floor echoes.


That’s like a blog post

I’m doing audiobooks.

So it sounds like the consensus is that audiointerfaces have less hum…

Not necessarily.
You can get good or bad internal sound cards, good or bad USB interfaces, good or bad USB microphones.

“On board” sound cards (built into the mother board) are usually pretty bad.

USB microphones and USB interfaces can be prone to whistles and whines from interference from the USB power. If they don’t have a built-in analog gain control, the recording level may be very low, leading to excessive hiss. You are usually limited to using only one USB microphone at a time.

Internal sound cards can be prone to picking up interference from other components inside the the computer case. Connections are usually via unbalanced mini-jack connectors, which are incompatible with most high quality microphones.

the consensus is that audiointerfaces have less hum…

None of them have hum. If you have hum, you’re doing something wrong or something is broken. I had hum for several months no matter what microphone and it drove me nuts before I discovered a speaker system in the room with me was broken and low volume humming whether I turned it off or not. The specification for audiobook background noise is -60dB which, in English, means the background noise has to be 1000 times quieter than your voice.

One of my small sound mixers has a power supply brick that hums if it gets too close to the microphone—no matter which microphone it is.

Did you have a bad experience with hum or are you following someone that did? Or are you calling everything hum? This is hum and buzz from a broken microphone cable.


Hum is way down the list of common home recording problems.

Separate interfaces are not get-out-of-jail free cards. I have a Shure X2U USB interface I stopped using because it had low voice volume and high internal noise (fffffff). But it doesn’t hum.

We publish audiobook mastering and diagnostic tools and have a pretty good track record for getting people published.

Do Not read a whole book before you find out you are recording in permanent sound damage. Record short tests first. You can post them here or submit tests to ACX, but that takes longer.