Waveform gradually behaves more and more erratically during recording

Audacity 3 Help Forum
, ,
1

First time posting (please have mercy!)

As stated, sometimes when I’m recording, the waveform becomes more and more erratic over and under the 0 line until it is beyond reasonable. Usually, this hadn’t happened often, but it had occasionally cropped up and ground my recordings to a screeching halt, or at least a crawl, while I restarted/rebooted and waited for it to resolve. Each time, I scoured the internet for even a mention of the problem, (let alone a solution) but found nothing, and then it would stop for a while. Lately, though, it has continued resurfacing every ten minutes or so - and with several deadlines approaching, this is really stressing me out. Can someone please help me figure this out? Thanks!

Equipment details: I record using a shock-mounted Røde NT-1 hooked up to a Focusrite Scarlett Solo. My computer is a stock Dell (gag-inducing, I know, but budgetarily necessary at the moment) running Windows 11. All software & drivers remain up to date. I have the ACX check, special copy, and special paste plug-ins installed, but the problem predates them.

Recording environment: hand-built isolation booth with double-layered walls & ceiling equipped with insulation, 1/8" MLV, Z-channels (properly installed), and sound proof drywall. Outlets and seams are sealed with acoustic putty & acoustic caulking; solid-core door treated with more MLV on both sides and frame is treated for a snug seal. During recording, quiet appliances in adjacent rooms are turned off; louder appliances like A/C and the dryer are off in the rest of the house. The room has no air flow when closed, and the computer tower is outside, with only necessary peripherals inside, cord paths sealed. Interior walls are treated with 2" open-cell foam pads. All that to say: there is no extraneous noise; it is quiet.

Solutions attempted:
-disengage/reengage mic’s phantom power
-restarting Audacity
-power cycle PC (both via restart & via shut down, variable duration pauses, and reboot)
-disconnect/reconnect interface, microphone, XLR from interface, even headphones (each separately in an attempt to isolate the problem)
-use a different XLR cable (and again, even headphones, though that shouldn’t matter)

Images:

20250815_103123
20250815_1031231920×1440 248 KB

Current setup. Different configurations, such as with microphone isolated (no monitor) and in the center, have made no discernible difference either to recording or the problem.

[Sorry, too new to fit more images]

I suspect it’s equipment-related rather than software, but I am really at a loss here. Please help!

2

20250815_103443
20250815_1034431920×1440 177 KB

Here is what the problem looks like just before I abandoned recording.

3

You can’t tell a lot from the waveform because it depends on the sound you’re recording…

It sort-of looks like Dropouts but they aren’t “clean” dropouts.

Your operating system is ALWAYS multi-tasking and interrupting, even if you are only running one application, but you should try to minimize the number of applications running and the background operations. Some people have to (temporarily) turn off their Wi-Fi or anti-virus.

You can try increasing the buffer length. That will allow more time for the other stuff to finish-up before the buffer overflows.

There is a FREE online book about optimizing Windows for audio called Glitch Free.

My computer is a stock Dell (gag-inducing, I know, but budgetarily necessary at the moment) running Windows 11.

Although a faster computer can help because it can finish-up whatever is interrupting faster, computers have been fast enough for basic mono or stereo recording for decades…

P.S.
Another thing can cause all kinds of weird problems is Windows “enhancements”. Make sure they are turned OFF.

4

The waveform on the first image looks like it is being displayed linear, while the waveform in the second image looks logarithmic. Did you change the setting between pictures?

1 Like
5

Plan-B: Audacity’s free competitor OCENaudio, (if possible with ASIO4ALL).
If the problem persists then it’s not Audacity’s fault.

6

Or, I believe Focusrite supplies REAL ASIO drivers (probably already installed).

ieremias77,
FIY - Audacity doesn’t support real ASIO or ASIO4ALL. ASIO4ALL adapts hardware that doesn’t have true ASIO drivers to work with ASIO applications but it doesn’t work the other way around.

7

Messages being posted here lately would indicate that Audacity might be having trouble updating the screen in real time while recording. Reducing the screen size or window size while recording might help, if that is indeed the problem.

8

No, I wish it were that easy. It is the same file; over a period of several minutes it gradually progressed from the first image to the second.

9

Thanks! I’ll give it a look.

10

Dang, it’s not them. I’ve run into problems with them before - randomly fading out audio to near silence in that case - so I always keep them off. Still, I checked them, hoping that maybe it was a simple as that, but unfortunately they’re still off.

11

Another thought.

Sometimes a nearby cell phone will cause interference and somebody recently discovered that a newly-installed Wi-Fi router in the same room was causing periodic interference.

1 Like
12

Update: so I’ve tried a few more things.

-I swapped XLR cable again; unless they have they same issue (different manufacturers), it probably isn’t that.

-I turned my microphone back to right-side-up, in case the positioning was somehow messing with the way the diaphragm was picking up my voice. Issue continued.

Whatever it is, I agree that it probably isn’t Audacity’s fault; it seems to be an issue with the analog signal rather than anything digital.

Things I think may be potential candidates:

-Temperature: my body heat relatively quickly warms my small, well-insulated room during the summer heat. This may be affecting the mic somehow; I don’t know. Points in favor of this: as I feel hotter, it often gets worse, and blowing a fan into the room and waiting seems to reset it. Points against: heat isn’t always corollary, and I also cut phantom power when the fan blows in, so it may be something to do with that. Also I’ve never heard of anything like that.

-Room size: my recording booth is relatively small, so I wonder about phasing effects. Points against that are that I’d expect a regular frequency rather than the erratic one I’m getting, and that the room is well-dampened.

-Something something phantom power: cutting phantom power (and getting quiet) seems to reset it for a time. Not sure what that would mean, but there it is.

All told, I still don’t know what is going off about it, but I’m still open to any thoughts.

13

CPUs can be affected by heat. When it get’s hot they slow-down to protect themselves. (I had that happen once a long-long time ago and the CPU fan had to be replaced.) Maybe try set the computer on an ice pack if it’s a laptop, or maybe put the ice pack near the fan intake.

That’s unlikely, but at this point I wouldn’t rule ANYTHING out… “You never know what the problem is until it’s fixed”.

1 Like
14

If it’s just cosmetic, a high pass filter will get rid of it …

hipass 20hz 48db

1 Like
15

GORGEOUS.

I tried it; it worked. This doesn’t “solve the problem,” but it solves the problem. It really helps, because the wild waveform is bad for editing, and anyway, I often apply a high pass filter during mastering to correct for minor noise floor issues anyway.

Thank you!!!

(And thanks to everyone for jumping in to help!)

If I ever figure out what’s causing the issue in the first place, I’ll try and find a way to connect it to this post for the next poor souls pulling their hair out over this. In the mean time, future lost souls - just use the filter.

16

Also to add, for future posterity: experiment with the frequency you’re cutting (this goes for the settings of any effect you’re experimenting with of course). I tried between 1 hz and 80; you’re just balancing between how effective it is at eliminating the problem and how concerned you are with keeping certain frequencies. Honestly, I use 80 on my own voice all the time - I am doing audiobook narration with a natural baritone range - and my ears can’t detect that the filter has been applied, let alone a drop in quality.

I’m only a little experienced with audacity & audio effects, however, with just a few hundred hours of very specific use. Someone with more experience, please feel free to add more, correct, or clarify.

Thanks again, everyone!

Edited to add: I use a relatively gentle roll off of 6db in my HP filter, not the 48db in the gif above. Experimentation with that obviously affects results as well.

17

It’s outside the room, so it’s not getting hot with everything else. But that makes me wonder about the interface…

Might be worth my putting up some blankets and doing a test recording in a larger room just to rule it out.

18

I’ll turn those off too; I don’t need internet while recording. Bluetooth as well, just in case.

19

Also, where did you get this gif, if you don’t mind me asking? It might help me find the underlying cause.

20

I made it: in Audcaity I added infrasound-noise to speech from an audiobook.

Capacitors are the components which block that type of noise.
But they typically take decades to go bad.

1 Like