USB Mic Problems

Audacity 3 Help Forum Windows
1

Hey, folks, I’ve been Googling about my issue for a couple days and the most useful hits so far have come from here, so I thought I’d register and ask about the specific problem.

Basically, my brother and I have a YouTube gaming channel we put on hold for a couple months sorting out tech issues. We recently started it back up, using a webcam (Logitech C525) mic to more easily record both of us. Turns out it has a constant “rushing water” noise that basically makes it unusable. I thought it was a problem with the mic (we had been using an AT USB condenser mic, and it was defective), but it doesn’t have that problem on my laptop. That’s really the reason I’m here; conventional opinion seems to be that all USB mics have some degree of noise, but it’s imperceptible on one and overwhelming on the other.

Here’s the sample. The first part is from my desktop, starting up with 5 seconds of silence than half the Pulp Fiction quote, then the second half and another 5 seconds of silence from my laptop (I spliced them together in Audacity and rendered in Sony Vegas). On both everything was disabled except the webcam mic and the default playback (Cirrus on the laptop (it’s a Dell Inspiron 3520), Realtek on the desktop), with the mic at 100% and the speakers at 1%.

My desktop was a DIY, I think the relevant datum are:

  • MSI 970A-G43 motherboard
  • AMD FX 8320 processor, 16GB RAM
  • Windows 7

I tried a dynamic mic (with the 1/4 to 1/8" converter) on both machines and had basically the same result (exact same noise as on the USB mic), except that it was really quiet, even at maxed volume. I’m also relatively sure it’s a hardware problem on the new desktop. The previous one (almost the same hardware except a weaker CPU and an ASRock motherboard; same case, drives, etc.) was what I recorded the Dark Souls videos on the channel with, and the audio is from the aforementioned USB condenser mic. Quality wasn’t amazing, but it doesn’t have that background noise.

I used Plot Spectrum on 5 seconds of silence on each machine with drastically different results. Audacity is definitely “seeing” the noise, but I don’t have the expertise to make heads or tails of what it means. They’re attached below.

Thanks in advance for any help, this is the first time I’ve ever posted asking for tech support (and I’ve been no stranger to this stuff for over 15 years, heh). Right now my brother and I are looking into recording on my laptop and adding it to our streams in real-time with Skype or something, but fixing this would definitely be preferable! :laughing:
desksample_spectrum.jpg
lapsample_spectrum.jpg

2

If somebody asked me cold what that was, I would have said you were talking into the wrong end of the microphone, but that’s not likely with a webcam…

Most of the stuff behind you that sounds like a working machine shop is the pickup from the room, echoes, computer fans, traffic noises, continental drift, etc, etc. I can’t get Logitech to admit it, but I’m betting that’s an omnidirectional microphone—receives from all sides.

You can play all sorts of games with a camera, zooming in to a tiny spec in the distance, but the sound lady then has to get in her car and go to that spec to record the sound. They’re very different. Just because you can fill the camera frame with your face is no guarantee that the sound is going to work.

Unless we specifically tell you, never submit processed sound. Once you mess with it, we have no way of knowing what went wrong. This page was written for audiobooks, but it’s not a dreadful format for submitting problems.

http://www.kozco.com/tech/audacity/TestClip/Record_A_Clip.html

The rain in the trees (shshshshshshs) is actually microphone noise, or, more accurately, microphone preamplifier noise. In the case of a USB microphone, they’re all in the same package.

Pure USB microphones can work well, but it’s not likely. The top three considerations of someone shopping for a microphone are price, cost and money. That puts the sound quality fourth at best and it’s usually even further down than that, just below colour of the paint.

That means there’s no money devoted to making sure the digital trash doesn’t get into the sound which is “The Yeti Curse.” Convenience is high on the list and that means low volume. Not directly, but having no knobs means recordings have to be made “safe.” You can fix low volume in post production but overloaded sound is garbage. So most of them deliver low volume right out of the shrink-wrap.

I still can’t shake the impression that you weren’t recording from the webcam. Do the scratch test. Start recording and scratch the webcam and then scratch the laptop built-in microphone or any other microphone that the system can see.

http://kozco.com/tech/audacity/clips/ScratchTest2.wav

Koz

3

From that YouTube submission, I’m guessing you’re not using Audacity 2.1.0. You can get that from here.

http://audacityteam.org/download/

If we start telling you about adjustments and settings, they’re going to be for that version.

Koz

4

It’s probably noise in the USB power. Noise is an analog issue, so somehow the noise is getting into the microphone’s analog electronics. It can potentially pick-up electrical noise from the environment, or from the power supply, and the analog* electronics itself always generates some noise. And of course, a microphone picks-up acoustic noise from the environment. :wink:

Some USB mics may be more immune to power supply noise than others. But, I’ve read about people having similar issues with the Blue Yeti mic so a better mic is not a sure-fire cure.

You can try a powered USB hub (a hub with it’s own wall-wart power supply). But, most people will advise you not to use a hub with USB audio, and it may not help. So… it’s a bit of a gamble.

Probably the most reliable solution would be a USB audio interface with it’s own power supply. But, these only work with studio/stage mics (analog mics with a low-impedance, balanced, XLR connection).



\

  • There is noise in digital circuits too… Quite a bit actually… But as long as it’s not bad enough to change a 1 to a zero or vice-versa it is of no consequence, and that virtually never happens… If digital data was not reliable the Internet would be useless and your bank account balance could be off, etc. If digital audio (or video) data gets fouled-up it’s usually fouled-up very badly.
5

Thanks for the quick replies, guys. :slight_smile: Oh, and I do have 2.1.0 on my desktop and by the time this posts on my laptop too (it was 2.0.6 since I hadn’t updated since we bought the old mic).

To an extent, yeah. :confused: I can sit the mic about a foot away from the case exhaust fan facing away from it and it still quite clearly picks it up (facing it is just slightly louder, but with a noticeable whine). I can record some omni/cardioid tests for next time if you’d like, just getting a handle on how big a WAV file I can fit under 1mb, heh.

Right, but 1) I didn’t want to track down a free place to host the audio, and 2) I didn’t think it’d matter (you can still clearly hear the difference between the first 5 and last 5 seconds). I did keep the raws, of course, and they’re attached below (trimmed to the 4/5 seconds of “silence”).

Gotcha. I had heard that from my Googling, but I wasn’t sure why the problem was so much more acute on the desktop than the laptop. Maxing out my speakers on the laptop sample and I can hear it a bit too, but it doesn’t bleed into the vocals like it does on the desktop version. And the mic was in the same place (about 18" from my head sitting on top of the open laptop case) both tests, I just moved the cable. So any environmental sounds should’ve been in both, you know?

Come now, I may have dropped out of DeVry, but I’m still at least technically an IT/telecom guy. :stuck_out_tongue: But fair enough… that’s the 3rd attachment.

Did you record that one yourself, Koz? Dig your voice if so. :]

That’s what I was afraid of. I did try the mic with a USB hub (cheapo Radio Shack thing we got for our PS3 because the slim models only came with 2 USB slots…), but it wasn’t independently powered, so it had the same problem. I’d also heard the problem might be case noise vibrating the internal sound card and causing noise that way (and it is constant), but then why would it affect a USB mic?

I tried hooking up an external USB sound card and it’s OK, but it doesn’t have a USB port itself (so the webcam mic wouldn’t work), and it didn’t have any boost settings (so our dynamic mic was completely inaudible). It may have solved the noise problem, but if so the cure isn’t any better than the disease!

Is that getting into phantom power and stuff? I did a very cursory examination of higher-end mics when we got started last October, but settled for the AT2500 USB condenser mic (and to be fair until it crapped out on us it worked great, heh).

I’m willing to throw some more money at this because my brother wants to facecam it up at some point so we’d be using the webcam anyway, but I’d definitely want to know it would actually fix the problem first.


6

how big a WAV file I can fit under 1mb, heh.

Just over 10 minutes [correction Seconds] 44100, 16-bit, mono WAV (Microsoft). That’s why the sound clip instructions are written like that. We did do some of the leg work.

You know the old microphone died because it was crushed under a Semi/Lorry? Microphones, even the delicate ones are pretty robust. Dog get to it?

Did you record that one yourself, Koz? Dig your voice if so. :]

DeVry college of elocution and articulation.

It will take a bit to crank through that posting. Please note that if the hiss is not “gentle rain in the trees,” then it may very well be something else, and just to cover it, this is approximately what the Yeti Curse sounds like.

http://kozco.com/tech/audacity/clips/FryingMosquitoes3.wav

Once you hear it, it’s pretty unmistakable no matter which data stream is making it. I think it was flynwill who experienced the noise and could identify specific, individual portions of the USB data stream, leaking into the sound. There is no practical fix. The choices are get a different microphone, or take apart and redesign the one you have.

As we go.

Koz

7

somehow the noise is getting into the microphone’s analog electronics.

Not even somehow. Cheap, inexpensive microphones that don’t cost very much are forced to use the USB power system directly to run the very sensitive microphone electronics. Even with stunningly good power supply rejection ratio, who-ha, who-ha, any trash on the USB Power wire is going to have about the same volume as a very, very quiet performance.

Ratty computers can make this problem much worse by “saving money” on the USB system. Are you really going to notice if your mouse doesn’t track exactly perfectly every time? Are you a good enough typist to notice that every 100th keystroke is wrong? The five volt USB system isn’t exactly at five volts. That’s not a problem, is it?

My Shure X2U USB microphone amplifier was not particularly cheap and it just doesn’t have the Yeti Noise problem (it has other problems). I don’t know that anybody has ever complained about whine on a USB mixer or one of the USB Microphone preamplifiers. They’re all steps up in quality and cost.

The ACX recording class recommends a Rode NT1-A microphone and the MBox Mini-2 Mic-Pre and digitizer. I don’t think the digitizer is made any more, but you get the idea. If you throw a little money, it’s possible to engineer out a lot of problems.

A note, Frying Mosquitoes will kill you if you’re trying to read an audiobook. It’s possible to pass ACX Noise Conformance and still hear the whine in the background. Kid screaming on a jet.

Koz

8

Oh. I bet the laptop is running Windows Enhanced Services.That’s why it’s so quiet. I’m hearing odd noises in there which don’t sound natural.

http://manual.audacityteam.org/o/man/faq_recording_troubleshooting.html#enhancements

Do you hear that “blowing into a milk bottle” sound here and there? See if windows noise reduction is selected or if you’re running one of the other conferencing services.

Koz

9

Is that getting into phantom power and stuff?

Phantom Power is a little magic. I use the phrase “48 volt phantom power” to split it off from some other, less popular, oddball methods. It’s applied to XLR, 3-wire microphone systems.

The sound comes down from the microphone exclusively on pins 2 and 3. Pin 1 is the protective shield, earth, ground, or whatever you want to call it. It’s not directly associated with the sound. You can damage pin1 and the show will certainly get noisier, but will not fail.

Phantom Power works between pins 2 and 3 as a group and pin 1. 48 volt juice (technical term) runs up both pins 2 and 3 and returns on pin 1. Properly used, they don’t interfere with each other and each service is a phantom to the other.

Koz

10

Gotcha, forgot to check the settings to drop it to 16. I’ll start doing that for next time [if there is one, heh!].

I live in the Midwest US, definitely semis here. :smiley: Well, I think the locals prefer “18-wheeler,” but yeah. No, I don’t know what happened to it. I’ll post later when I get back home (I’m at Dad’s house doing more troubleshooting, see below), but basically it has a constant noise similar but not identical to the sample I posted on the webcam mic, and it’s uniform in volume and everything regardless of the mic gain, up until I lower it below 20% or so (when it starts fading and eventually is inaudible, and you can just barely hear vocals). It did that on the laptop too, though.

:laughing: I like that. The “water” sound (kinda like rain, like you said) in the background is the same, but my sample doesn’t have the, uh, reedy skeety harmony.

I took a look, but there’s no Enhancements tab or anything like that. Unless I’m missing something? I don’t know 8 that well and Google didn’t turn up anything else about Enhanced Services (at least vis a vis audio).

Anyway, the test I talked about earlier was taking the USB mic to my dad’s house to use his computer (a Dell Inspiron 570), I attached the WAV and analysis below. Unfortunately I couldn’t turn off the TV in the living room even for a 5-second test :stuck_out_tongue: , so that’s what the background “talking head” noise is; but it definitely doesn’t have the noise problem my desktop at home does. So I think that pretty decisively narrows it down to the USB power theory (darn it :confused: ).

Does seem to be the case. What’s weird then is that my current motherboard (the MSI) is rated so much higher than the ASRock I had last time (which didn’t have this problem, though it certainly had other ones, heh). The mic is also the only USB device that seems to have any problems; I have a PS2 controller on a USB connector, the mouse, Elgato capture device and sometimes one other thing all plugged in, and none of those have had problems [I could detect, anyway].

Thanks again for your help. :slight_smile:

11

D’oh, forgot the attachments. I hit Preview, had to log back in, panicked a bit when I thought I lost that post, heh. Here they are:
dadpc_sample.jpg

12

By he way. That’s a ten second clip, not ten minutes. I need somebody proofreading after me.

Koz

13

It’s a magnitude thing. Microphone level is reeeely tiny. We’re talking very small fractions of a volt. When designers make a microphone amplifier, they have to choose parts that do the basic electronic job, and then consider how noisy they are. Molecular level noise. Microphone signals are the butterfly wings of the sound world.

It’s certainly possible to not have this problem by throwing money. My ratty sound test was made on $19 USB analog headset adapter. My $100 Shure X2U USB Microphone adapter doesn’t do that. The Shure is an expensive pro adapter with filtering and processing to get rid of USB noise problems. (It does have low volume and I don’t recommend it).

The UCA202 has the problem, too, but since that adapter doesn’t manage microphone levels, the whine is so tiny that it took flynwill lab instruments to find it.

Koz

14

Heh, OK, the 10 minute thing made me raise an eyebrow, but I was certainly willing to believe dropping by the quality by 2/3rds would drop the filesize by 90%. :laughing:

Fair enough. :] In that case, what would you recommend?

Oh, and going back to the power discussion; I pulled up HWMonitor* and took a look at the voltage section on the motherboard. Everything else seems to be in line, but the +5V is only reading 3.027, while everything else is about right (12V is reading 13 and change, for example).

  • Downloaded that when I started having overheating problems. Turns out my problem was putting the intake fan on top. Did you know hot air rises? :astonished: I swear I used to know this stuff, heh…
15

the +5V is only reading 3.027

That’s not good news. I don’t remember what the lower limit on logic is, but you’re getting close to it.

In that case, what would you recommend?

I have done paid recordings with my analog microphone, simple sound mixer and a Mac with Stereo-In (older design). It has good flexibility, good volume, low noise, and high quality. It’s not convenient and you do need to know what you’re doing. It’s very not plug and play and it needs a wall plug. I have a smaller FP24 Shure mixer that’s actually too small. It starts to have USB microphone problems and it doesn’t have good volume adjustments.

1000×968 81.5 KB

I know you were hoping for a quiet USB microphone that works.

ACX has a recommendation in their videos. They use an analog microphone with a USB microphone amplifier and digitizer.

The ACX recording class recommends a Rode NT1-A microphone and the MBox Mini-2 Mic-Pre and digitizer. If you’re not depressed enough yet, you can watch their video, remembering that they’re playing to the broadcast and audiobook market, not casual recording.

http://www.acx.com/help/video-lessons-resources/200672590\

Play the first one: Setting up a home studio.

Koz

16

I’m guessing HW Monitor is reporting this for the 3.3V power. If it would be for 5V, it would be far below the 4.85V limit. Most 5V logic will still work at 4.75V, except some audio gear because it needs to transform the 5V to a higher voltage.

Supposing the webcam has an internal low drop regulator and it works on 3.3 V internally (otherwise it shouldn’t work at all), this would explain the excessive white noise on your audio.

I think you need to check your computer’s PSU. The easiest way is by cutting a USB cable and measuring the red and black wires if you happen to have a multimeter. If the voltage on the USB port is around 5V, HW Monitor is wrong. If it’s only 3V, there is something wrong with your computer’s hardware, I’m afraid.

17

Thanks again for the help, Koz, checking that video now. :]

Edit: OK, I don’t have that much money to throw at the problem. :laughing: Bookmarked for when/if we want to take this to the next level, though.

It’s not that I have a problem with traditional mics, it’s just that between the webcam and the condenser (which I did manage to get a proof-of-purchase for and am sending back soon) I have a lot of USB equipment already.

We definitely need something convenient, and preferably mobile (for some recordings I plan to take my desktop to a friend’s house and record there for a couple hours, that edit that down a couple episodes later), which is why as you can imagine the webcam seemed like an especially good idea.

I heard the limit was a pretty narrow +/-10% or so, yeah. All I know is what it says. :confused: (Attached below.)

That would explain a lot!

Argh, no such luck. I have a lot of electricians in the family, but they’re all about 600 miles away, bit far to bum some equipment. :laughing: I’ll keep that in mind, though! I already sent a tech support ticket in to MSI about the problem, might as well get the ball rolling.
hwvolt.jpg

18

We definitely need something convenient, and preferably mobile (for some recordings I plan to take my desktop to a friend’s house and record there for a couple hours, that edit that down a couple episodes later)

Post back when you get something you like. This isn’t a help desk (although it functions like one). It’s a forum, users helping users. You’re a user so you should be helping other users.

We have found that tiny home systems don’t expand. There are vast legions of people desperate to connect two and three USB microphones to their computer. You can strong-arm force two USB microphones, but three isn’t happening. I don’t know of a USB microphone that doesn’t have USB microphone problems. Convenience and Cost tend to bump sound quality.

You can send a microphone back if it’s not convenient enough and you just wouldn’t buy it if it costs too much, but if the quality is a little ratty, well, that may be OK. Let’s see just how ratty we’re talking here.

Koz