ART USB Dual Pre digital artifact

I’m using a preamp to record voice, the ART USB Dual Pre. It is producing some strange noises every 5 seconds or so. The intensity depends on the position of the device. It is louder or softer when I move it around or up and down. Sometimes the noise disappears altogether if the unit is held in just the right position. I can’t establish any pattern. The noise is present no matter what combination of cables or microphone I have tried. I have tried turning off all surrounding electronic devices, to no avail.

Here’s a sample (seconds 3, 8, and 13).

Does anyone have an explanation?

It’s possible, but I doubt the Dual Pre is doing this by itself. What’s the microphone?

What around you is wireless? It sounds to me like maybe your wireless Baby Monitor or House Lighting Device is “checking in.” Where’s your cellphone? Put that in the kitchen for a while as a test.


The above sample is recorded using a Coles 4104 ribbon mic. Here’s another without any mic. I’ve gone so far as to unplug the wifi router. No cellphones here. And I live in the country, the nearest house is about 500 ft away.

I wanted to be sure, so I shut off power to the whole house from the main breaker, removed the batteries from any device, and that didn’t help. I can’t see how this is anything other than internal noise, but on the other hand, why would it vary in intensity when I move the box around?

SOLVED, I think. I went outside with the Dual Pre and a Zoom H2n, and did some direction finding. Didn’t take long to narrow it down to these things (see attachment). When I stand under them, the meter pegs at 0db. So what are they, and what can I do about them?

That sure looks like the cable TV system. Good luck getting them to service it. Either Comcast or Time-Warner is the the most hated corporation in America. I think the other is second.

They’re not supposed to radiate. Companies are under strict rules as to how much radio trash they can produce. That’s the Part 15 rules you might find listed in the documents for your computer.

I wouldn’t be shocked if one of the screw-on connectors has weather damage and the protective shields are leaking.


Look like cable TV distribution amplifiers. You might be hearing a “keep alive” signal, that’s there to signal cable breaking.

The strange thing is, the ART shouldn’t be affected by this. Are you in an extremely dry area?

What about with no cable or microphone? Does the ART still pick up the interference?
If so, is it substantially less than with a mic connected?

Do you have the “wrong” side of the ART turned down? If you don’t plug anything in, a microphone preamplifier is a little radio receiver.


It’s dry in winter here because it’s very cold, so yes.

Not sure what you mean by having the wrong side turned down. At one point I even heard very faint broadcast radio, so yes it sure is susceptible to RFI. I’m wondering if the “faulty” cable equipment on the telephone pole is actually within spec and the ART is just not well shielded. In any case, I have contacted the local cable provider, but I’m not optimistic that I’ll actually get a response.

The noise is a bit louder without a mic, or at least it seems that way because there’s less white noise.

While awaiting a theoretical response from the cable guys, I’m wondering if anyone could offer some advice on shielding the Dual Pre from RFI.

Not sure what you mean by having the wrong side turned down.

It’s a dual preamplifier and you’re only using one side, so there is a wrong or dead side. That should be turned down.

if anyone could offer some advice on shielding the Dual Pre from RFI.

I don’t think I’ve ever seen equipment that sensitive to radio interference.

Do you have red and white radio towers in your neighborhood?

If you have an AM radio transmitter as a neighbor, your chances of being able to get a clean recording are practically zero. Actual radio stations have trouble doing that.


Very surprisingly, the cable guy came, and he was very courteous, but could not find any problem with their equipment. So I walked down the street listening to the ART, and the noise varied in intensity but never disappeared. He suggested that a nearby farmer might have equipment sparking on the AC line, which would explain why the noise increases as I approach the electric line. I can’t imagine what kind of equipment would draw electricity every 5 seconds. I guess the next step would be to drive around while listening.

He was right. Electric fences are a thing. You can run three strands of barbed wire around your fields, or one thin strand of insulated electric fence and call it done. The stock touches it once and that’s all it takes.

But it is electric and it can have leakage and sparking problems. But that doesn’t account for your ability to listen to local radio.


And just because I haven’t gone down the magic path for a while, in the case of AM radio, any bad electrical connection can be made to receive radio programs. It’s actually not that unusual. In extreme cases, people have been able to faintly hear radio shows on their teeth from acid saliva and metal fillings. If you’re close enough to the transmitter, you can get CBS Radio News on your coffee maker.

FM is much less likely to do that for two reasons. FM Transmit Antennas are about four feet across and they’re placed at the 600 foot level of the tower. So the tower doesn’t actually radiate. Just the little thing on the top. If you do get FM by accident, it doesn’t make music, just noise.

You never said if you knew of any red and white towers. I think the colors are fairly common across countries.

People tend not to see them. They get tuned out until something goes wrong.

“Ever have any trouble with that 5,000 watt transmitter and tower at the end of your block?”
“The what?”

That tower in the illustration is 5,000 watt KABC. One of the movie people saw all those empty, cheap buildings at the tower’s base and decided to shoot a video show there. Fine, as long as you don’t, like, need sound or anything.


Sorry, I forgot to answer your question. There’s a nearby mountain with lots of towers on it, but I don’t think there’s any MW towers, and in any case the only time I heard broadcast radio, it was so faint that it was negligible. I got a kick out of the idea of listening to the radio with your fillings :smiley:

While I like the ART preamp, I think I’ll have to use another one instead. I have a Rolls MP13 which has more hiss, but at least I can filter it out in Audacity. Which brings me to another question about the Noise Removal tool. Would someone please explain the difference between the “Noise Reduction (dB)” and “Sensitivity (dB)” options? Any tips on fine-tuning the settings to minimize distortion?
Screenshot at 2017-01-30 17-35-58.png


  • Start with finding a convenient spot in the house where the RFI is minimal.

  • Get a well shielded USB cable. You might already have one. If you have one with a ferrite bead on it, try it. A ferrite bead is a knobby thing clamped around the cable, close to one of the USB connectors. Usually black. If you have two of these cables, you can detach one of the beads and place it on the other end of the first cable, so you have a cable with two beads.

  • Get well shielded XLR cables/connectors. I suppose you might have those too, since you use a very good ribbon mic. Neutrik has special XLR’s with RFI protection.

  • Check if the ART has the required 12 or 22 pf capacitors between pen 2 and pen 1 and between pen 3 and pen 1. If these are not present, the device will be much more susceptible to RFI. The capacitors need to be as close as possible to the XLR. Especially with a ribbon mic, you could also insert small ferrite beads into the signal path, but that needs to be done by a tech, because you need to cut traces. Unless you are familiar with soldering and circuits, of course.

  • Ground the ART and/or the laptop. The ART is simple, as it’s a metal case. Unscrew one of the screws that goes into the metal case. Attach a wire and connect the other end to mains ground, the water system, or the drain and see what works best.

  • Build a Faraday cage. This is much harder. It’s usually done by covering the entire room in copper matting. See what you can do. Alternatively, you can wrap the ART in fairly thick copper foil and ground that.

  • Get a mic with a much higher output, like the CAD M100, or the Rode NT1. You could also try an interface that is less RFI sensitive. Sound Devices and RME spring to mind, but these are very expensive. From 750$ upwards. And some handheld recorders are also much less sensitive to RFI. Olympus, fi. Maybe you can make a deal with a vendor to test a number of devices?

  • You could also try a transformer between mic and preamp. But that’s not as easy and it might even make matters worse. So I’d take this as last resort.

And I’m a bit skeptical about cattle fences. Even when it’s a once every five seconds pulse, which is plausibel, these are usually not HF. But it is still possible.

While I wrote this, you posted:

There’s a nearby mountain with lots of towers on it, but I don’t think there’s any MW towers, and in any case the only time I heard broadcast radio, it was so faint that it was negligible.

Any chance it is military? Radar could show a 5 sec period and is very hard to suppress. Last time I had a studio with that kind of problem, they moved in the end, after spending a lot of money.

Any chance it is military?

That may be even more remote than getting Doctor Laura on your fillings.
Much more likely that’s FM and television.
That’s what our collection looks like. That’s the top of Mount Wilson. The optical Wilson Observatory is in the back. Sucky picture, but it is some 30 miles away—oh, and a mile up.

Noise Removal.

That’s not a thing any more. Audacity 2.1.2 has Noise Reduction and in my opinion works better. But even with all that, you’ll never reduce the noise beyond about 12dB before voices start to sound a little funny. It’s worth a shot. Use the forum generic format for recording a test clip. A mono clip can’t be much over about 20 seconds for forum posting, but for personal experimenting, you can go as long as you want. It’s good to have that silent stretch in there. It’s rough to analyze noise without it.

“Cleaning It Up In Post” is not something you go into before you even get the show. It should be reserved for Disaster Recovery on a show which is normally well behaved. If you blow your wad before you even get the first show, what are you going to do when you create trash by accident instead of an import show?


I agree, but you do what you can with your equipment. I’d have to drop several hundred for a professional preamp.

So I upgraded to 2.1.2 and probably broke my distribution while doing it. Linux is great, until dependency hell.

See attachment as per instructions. This was recorded with a Coles 4104 mic through the Rolls MP13 preamp to a Zoom H2n recorder. I’m wondering how well you Audacity gods can clean this up.

Ah. Comes the dawn. A Coles 4104 Lip Microphone is designed to be screamed into by someone in a very high noise environment.


Right. So disconnect the microphone and preamp and put those in a drawer somewhere. Try straight recording into the H2n. Follow the instructions for one of the stereo options or, if they admit to it, a mono recording. I think you can get there with the Mid-Side settings reduced to minimum separation.

I have an H4 and it will produce a respectable recording by itself with far less hiss that your posting.

Post a new clip.


This is a test recording I did with an Olympus 823 Personal Recorder (no longer being sold, unfortunately).

That’s in its performance setup. In Voice Recording setup, it will record conversations anywhere in a reasonably quiet room. I haven’t tried it in larger rooms or outside.

That clip will meet AudioBook sound standards.