Audacity Forum

Hello, I recorded a bunch of tracks this weekend, and in some of the tracks there is an unwanted static crackling whenever the actor speaks. It seems similar to clipping, but I was recording safely low (-12db at the very maximum).

Some tracks are effected, and some are not. I did not change anything between takes, so I think it's a computer problem, not a recording environment problem. It seems to be happening entirely randomly.

The unwanted effect happens all the way through the track. If it's affecting a track, it goes the entire length of the track.

Here is a very short sample:

http://www.buffalo.edu/~breslin/badstatic.wav

Hopefully I can learn what caused this, so I can avoid it happening again.

Oh, that's good. Data errors only when the sound level goes over a certain amount. And it's not clipping or constant noise. Cool!

OK, so now we need you to finish the sentence:

<<<Hello, I recorded a bunch of tracks this weekend>>> ...using my [model number] microphone plugged into my [model number] mixer and then on to my [model number] computer by way of the [type/analog/digital] connector.

You can't give us too much information. Mac? PC? Recording depth and bit rate?

Has this particular configuration ever worked?

Koz

Do an analysis / spectrum on the voice. I've never seen that before. You get the expected humps and lumps which is the actor speaking, but the hash produces a completely flat display. Sort of the actor's voice poking up over a field of wheat.

Koz

Thanks for your interest. I will explain what I assume are the relevant parts of my setup, but please let me know if more details would be useful.

It's a new RE20 mic, XLR plugged into a M-audio Omni-Studio, which itself plugs into a Delta 66 sound card. Then there's the computer... well, I could list the whole specs, but for overview, it's a dual core 3.0 processor, not overclocked. 4gig ram, Windows XP. Pretty standard stuff.

Like many home studios, I have a small hole in the wall through which I'm running a large number of cords/cables. From mouse and keyboard USBs to monitor and mixer cords, etc. But It sounds like a software problem to me, not a physical cord-shielding problem.

Sorry I did not understand the post about analysis / spectrum. Did you mean that you did one yourself, on my short sample, or you are recommending I do one?

I think the most important diagnostic info is the fact that the bug happened only on certain tracks, and when it appeared, it corrupted the entire track. Pretty clearly a software problem, right?

You did miss some important numbers. 44100, 16-bit, Stereo? Project, recording, everything all running at the same rate?

<<<RE-20 Microphone>>>

A real RE-20? Not that souped up thing that doesn't sound as good? Rubber spider mount?

So let's say you're singing all three parts of the Andrews Sisters in front of the RE-20. You lay down Maxine and play that back into headphones while you lay down Laverne and then Patty. The tracks are one above the other on the timeline and inexplicably, one of them has that distortion in it all the way through -- beginning to end and the other two tracks are completely clean.

That's what I mean by detail. I have to build your studio and operation in my head.

Does the distortion always sound like that?

Is there a cutoff? No matter what, by the third (or fifth, or tenth) track, there will guaranteed be at least one damaged track?

Has it ever worked? This doesn't feel like a system you bought on Thursday and started using on Sunday.

I think the damage is too regular and perfect to be an analog problem, but not likely to happen to an audio bitstream. That leaves the digitizer, the Delta-66.

I'm going down two alleys. If I wanted to create this error, how would I do it, and how would I make it worse? Can you make it worse?

The Omni Studio plugs exclusively into the Delta-66, right? No power supply in the wall?

I need to stop here until you fill in some of those holes.

Koz

Thanks for asking the questions!

--
Mic

I didn't know there's a knockoff RE20! Well I sure hope this one is the real one:
http://www.zzounds.com/item--ELVRE20

(I got the one without the shock-mount, partly because the stand that comes with it isn't rated for a mic that heavy. Also, I already built a shock-mounted stand.)

--
Recording

I'm recording vanilla voiceover work, different scenes on different tracks. Same actor each track, and same setup. No headhones because he's not syncing with anything.

44100, 16-bit, Stereo.

Everything is running at the same rate. I didn't know you could have different rates. (Unless you work at it, I guess.) The output file sample rate & format were specified by the client, so I went to Edit->Preferences, and put the numbers in there. This automatically adjusted the project rate.

--
Distortion

The distortion always sounds the same. Would a longer sample help? If so, should I make a WAV, or can you get enough info from an OGG?

--
Patterns

There's no known pattern to when it appears or not. We laid down about 10 tracks yesterday, and none of them exhibited the problem. Last weekend we laid down about 18 and the bug happened on tracks 5, 6, and 12.

I don't know how to make it happen, or how to vary the problem if it does happen. I haven't been monitoring the signal during recording. So I would only have heard the problem during playback.

The only pattern I know is as stated: if it's happening, the entire track is corrupted. If the first piece of a track is clean, the whole track is clean. (Unless you combine tracks, but you know what I mean.)

This led me to believe that Audacity is generating the error. I am "assuming" that the Delta-66 and the OmniStudio are sending the signal from the mic all the time, and Audacity simply picks it up. I "assume" Audacity does not have to message the Delta-66 or OmniStudio and say "hey, send me the signal, I'm starting to record a new track".

Now, if it *does* require some messaging like this, getting a programmatic "handle" or so, then the Delta-66 might cough when it gets the request. (Then it would have to re-set after the track is ended, accept the request fine for a few tracks, then cough again.) For all I know, it could be this; I don't know how the signals get routed.

--
The System

I've had most of this stuff for a while. The mic is new. Never had a problem like this before. This might suggest that the mic is at fault, but I don't know if this particular problem can be caused by a mic.

The OmniStudio is powered by a 9V AC.

--

Digitizer -- is that the OmniStudio or the Delta-66? (The Delta-66 is an internal sound card that has a jack for a heavy-duty cable that runs from the OmniStudio. That cord looks more-or-less like a monitor cord. Different number of pins of course, but same basic shape.) So I don't know if the signal getting digitized in the OmniStudio or the Delta-66.

If it's probably the OmniStudio or the Delta-66, then I should take it up with M-Audio. They're probably equipped to diagnose the problem. The main thing is to figure out for sure what the problem is, so I don't lose some valuable work.

--

<<<Well I sure hope this one is the real one:>>>

http://www0.shopping.com/xPO-ElectroVoi ... namic-Card

At one time, if you went shopping for an RE-20, they would shuttle you quickly off to this microphone. More Output! Computer Friendly! Gets your Clothes Whiter and Brighter!!

But it doesn't sound as good... They got the extra output voltage (perfectly true -- almost double) by sacrificing quality. Flee! Flee!

I need to leave for a while. I'll read your response carefully when I get back.

Get your screwdrivers out, take apart the case (clean while you're in there) pull out the sound card and put it back in...twice. Screw everything back together. In an unconditionally healthy system, this action will make no difference at all. (If you can't do any of these, just say so.)

While you're in there, look for dust bunnies in the little fans, etc. Inspect your brains out with a powerful flashlight.

If you can figure out a way to warm up the system, you might try that. I'm serious. Park the machine in a closed closet (or equivalent) and carefully monitor the temperature in there. When it gets into the 80's (F) for a time, try a simple capture again. Part of troubleshooting is sometimes to force failure. Unstable systems tend to be ratty over a limited range of conditions. Upset the conditions and see what happens.

I used to do this for very large companies and there are scary things we can do when there's lot of money riding on it.

Koz

I know that the relevance to this is only partial, but in using a Mac QuickSilver system I have a consistent problem with digitized audio. The input was initially through an M-Audio Fastrack (USB) which I could monitor from both the M-Audio and the sound card output and both were OK. However, when played back, the snap, crackle, pops and wadded up paper noise made the input unusable.

Thinking that it might be the Mac soundcard, I bought an M-Audio Audiophile 2496 and installed it into the Mac. The same problem with the addition of intermittent Kernal Panics and the Blue Screen of Death (yes, Macs do this too, regardless of what they say on their T.V. ads).

So on further investigation, just pushing an .mp3 file over to the Mac and burning a CD also has this noise induced.

The QuickSilver is obviously not the latest technology but should have worked for what I was asking of it. So, don't rule out the CPU config just yet.

M-Audio is famous for using ASIO drivers for its computer/sound management. Audacity is famous for not supporting ASIO.

Koz

<<<Mac QuickSilver>>>

Does that machine have USB2? USB1 was notoriously slow and sloppy. Also the earlier machines wouldn't supply enough battery down the USB port to run any serious equipment. You may find a lot of the snapping goes away if you run the USB equipment through a Wall-Powered USB Hub. By itself. You are using the hub for its ability to supply significant battery voltage divorced from the computer.

Sometimes USB2 systems have this problem.

We had one system here that could tell the color shirt I was wearing. If I wore the blue shirt, the microphone would start clicking and snapping. Maybe it was the humidity. Putting an inexpensive, wall-powered hub in cleared it.

Koz