Interesting question – USB microphones and USB switches
I was looking at USB switches (not for a USB microphone, but to switch inputs into a DAC), and came across this curious bit on the Amazon Q&A section for the Plugable brand unit here:
https://www.amazon.com/Plugable-One-Button-Swapping-Between-Computers/dp/B006Z0Q2SI
(Note to Google: This is NOT an affiliate link, nobody here is trying to make a dime, you really don’t have to hide this post from the entire universe to protect them.)
Here’s the Q&A:
_________________________________________
Question:
Is there any static or extra noise when used with a microphone? i have a snowball ice.
Answer:
MANUFACTURER RESPONSE: Thanks for asking! This USB switch is bus-powered which means it draws its power from the USB port on your computer. This can definitely add noise to audio signal depending on many factors. We do not recommend our switches for pro audio/video gear for this reason. Thank you – Plugable Technologies
_________________________________________
My first reaction was that this was absolute nonsense.
First the disclaimers: I like Plugable brand products, I use their hubs, and they’ve gone out of their way to make some compatible with the Raspberry Pi in ways that others do not, and to make that information available. I think they’re being super, super cautious here, which is okay. I don’t think I’m going with them on this purchase, but the reasons have nothing to do with this.
My first thought was that, as they say, “bits is bits”. If there was noise enough in the box to interfere with discriminating between ones and zeroes than it wouldn’t work for any data. It didn’t seem to make any sense.
On reflection, though… maybe. Just maybe.
Again, any low-level noise in the box is irrelevant to the digital signal coming from the USB microphone, that can’t be an issue, but that’s not all that’s going on in this scenario. The answer makes reference to the USB switch being powered by the USB port on the computer, and that’s true, but the relevant part is that the microphone is being powered from that same source. If it were a dynamic microphone requiring no external power I think there’d be zero chance of distortion in this scenario, but the Blue Snowball Ice is a condenser microphone (or possibly a sort-of hybrid electret that is boosted by external power). That means that the 5 volts of USB power is being boosted to some higher voltage inside the microphone and applied to the capsule – the diaphragm assembly that is the condenser/capacitor.
So what happens if you add noise to the 5 volt power circuit?
Well, it’s never going to be a good thing. Unless it’s filtered, it’s going to get amplified by the voltage boost circuit supplying “phantom power” to the condeser.
Okay, I’d guess that the 5 volt USB power is typically pretty noisy, coming from who-knows-what computer or hub. It’s probably very unusual for low noise levels on the 5v line to be much problem for anything plugged into it, virtually all USB devices are either all digital or some novelty item, so I doubt that the designers of the computer or whatever worried much about that, and it also seems likely that the manufacturers/designers of any quality microphones must have anticipated the problem and handled it in their devices, but… clearly, given the universe of all possible USB condenser mics, good and bad, it’s not impossible for noise on the 5 volt power side of the USB connection to add noise to the analog signal inside the mic. If noise on the 5 volt wires is allowed to make it all the way to the microphone capsule it will add some noise to the analog audio, whether the ear can detect it or not.
In other words, noise on the incoming 5v power wires on the USB could be amplified by the voltage booster, applied to the microphone capsule/condenser, converted from an analog signal to an encoded digital signal, then sent back to the computer though that same USB cable to a computer or digital recorder to be stored in a file.
So… I still very much doubt that it’s really a practical problem often, but yes, it’s not impossible.
Part of my initial failure to recognize this is purely my own internal bias- I tend to think of connections as carrying either analog signals, digital signals, or power, with very limited and controlled interaction between them. I even have my bins of cables organized in exactly that way- analog signal, digital signal, and power. Usually that thought pattern works fine, but the truth is that USB and POE and other current-standard cables mix power and digital signal, and when they meet in an analog device (or in a partially analog device) there is always the potential for an exception.
I happen to have a Blue Snowball Ice microphone myself, which I bought for a project that fell through shortly after I bought it, and I’ve never used it much. Since I started thinking seriously about podcasting I’ve gone exclusively with XLR microphones, mostly for the versatility of getting them working with other equipment. So far the ability to mix multiple USB microphone signals seems very limited and/or expensive, the problems of time synchronization and re-encoding are significant… maybe huge. This is not the first time I’ve been glad I made that decision to go XLR analog. Encoding all the concurrent analog signals at the same time in the same device whenever possible seems to be the way to go, at least so far.
– Robert the Wombat