Jump to content

Problem with computer noise/static/hiss on laptop with USB audio interface


Mike Bond

Recommended Posts

Hi all, this seems to be a common problem discussed across the internet, but I have yet to find a truly similar situation to mine, so I thought I'd reach out to you lovely people.

I'm experiencing some "noise" in Cakewalk and I'm trying to diagnose the source so I can hopefully eliminate it. First up, here is what I am using:

  • HP ZBook 15 64 bit Laptop running Windows 10 with Cakewalk
  • Behringer U-Phoria UMC22 USB audio interface  
  • Behringer USB WDM AUDIO 2.8.40 drivers (as ASIO4ALL is rubbish)

What I'm getting is a "static" sounding hiss that steadily increases across the frequency range, hitting -72 db max. That's the loudest it gets and that's at around 24 kHz. Somewhere along that "static" is a hum at around the 630 Hz mark (all this noise can be seen in the image of my graphic equalizer below).

image.png.39fe91d033a202ae101f6a4270a27603.png 

 

More info...

  • I have NO inputs connected to the audio interface (no instruments or microphones)
  • I have NO speakers connected to the output
  • Nothing is connected to the mains (the laptop is running on battery) so now Ground Loop issues
  • NO wireless devices are connected
  • There are NO other electrical or lighting related items switched on in the room.
  • All other speaker or audio devices on the laptop have been disabled.

What I have tried...

  • I tried using a different laptop (same problem)
  • I switched to using the laptops own internal mic and speakers (there was some other lower noise going on, but could still see the same noise pattern from 1 kHz onwards)  

My brain has now run out of ideas. I can only assume that it's computer related, but it seem strange that I get the same thing on another laptop.

The reason that this is an issue for me, is that I am recording distorted guitars so the gain on this noise is also significantly increased. Yes, I am using a noise gate, but you still get slight trail of the "static" just before it gate kicks in, which just sounds weird. Making the gate react as fast as possible doesn't quite cut it and isn't the sound I'm after.  

Any help or ideas would be very much appreciated at this point!

 

Link to comment
Share on other sites

3 minutes ago, bdickens said:

WDM is the absolute worst driver mode there is for Windows. Absent an interface that has proper ASIO drivers, try switching to WASAPI.

It is possible too that it is the interface itself is just noisy.

I believe you may be thinking of MME

When a real ASIO driver is not available

  • WASAPI is the preferred driver for Win10
  • WDM is the preferred driver for older OSes

MME usually works with anything, but it is the poorest performer.

Link to comment
Share on other sites

1 hour ago, bdickens said:

WDM is the absolute worst driver mode there is for Windows. Absent an interface that has proper ASIO drivers, try switching to WASAPI.

It is possible too that it is the interface itself is just noisy.

 

1 hour ago, scook said:

I believe you may be thinking of MME

When a real ASIO driver is not available

  • WASAPI is the preferred driver for Win10
  • WDM is the preferred driver for older OSes

MME usually works with anything, but it is the poorest performer.

Thank you both!

Sadly the MME didn't work, with my interface and I couldn't get any audio from it.

The WASAPI did work, but had pretty much the same noise profile as using WDM.

That info is useful though. I think I might need to look at a different USB Audio Interface that has specific drivers for it. If it doesn't work, I can always return it and I'm no worse off.

I will report back what I find, thanks again

Link to comment
Share on other sites

31 minutes ago, tecknot said:

Hi Mike and welcome to the Cakewalk forums.

I am going to take a stab and ask, is the bit depth and sample rate matching between CbB, your U-Phoria and windows?

Kind regards,

tecknot

Hi and thanks! :)

Yes, I have been making sure that everything is all the same in each case. The U-Phoria only likes 16 bit, the rest I've just kept at 44100. Thanks for asking.

Link to comment
Share on other sites

So are you hearing a crackle (static electricity does not produce noise until it is discharged, but a crackling noise is often so described) or a hiss? The noise is probably coming from the signal chain that includes the electronics of the input devices, the computer and the output devices. The software (Cakewalk, drivers etc.) should not produce noise, although it will faithfully propagate, measure, and record it. Software can be responsible for distortion of a signal, (clipping, dropouts etc.), but if you are not feeding it a signal, then the signal must be coming from somewhere else. The electronics themselves generate noise at a low level. Even the best equipment will necessarily add to the noise floor unless it is turned off and kept at absolute zero--there is thermal noise in every active circuit. The circuits used in audio are also frequently subject to picking up electromagnetic radiation from the environment. The fact that an EMR source is not readily apparent in your work room is hardly dispositive. If your computer is not shielded so well that a nearby radio will not pick up anything , then it can still pick up things like neon signs from blocks away.

A 72 dB difference between signal (what you want to hear in this case) and noise (the stuff you do not want to hear) means that the digital audio signal at clipping (0 dB) is about 4000 times the noise you are measuring. If you turn up the gain enough on any equipment, you can probably probably produce a hiss that will be audible against a "silent" signal. So to prevent annoying noise, you can reduce the noise level or decrease the gain required to hear the signal. Likely you are not recording anything that peaks 70 dB higher than the silence between songs. Also likely, you are not going to be able to do much to actually decrease the noise in your system. So a better option might be to record at high enough levels so that anyone listening to the recording will turn his volume knob down to the point where the noise will no longer be audible. 

https://hub.yamaha.com/audio/music/what-is-dynamic-range-and-why-does-it-matter/

  • Like 1
Link to comment
Share on other sites

16 hours ago, Base 57 said:

If you have not already, try using USB cables that have ferrite chokes on the ends. Especially the one that connects the interface to the laptop.

Thanks very much for the for the advice, I do have the ferrite chokes on either end of my USB cable already unfortunately, so that's not it.

Link to comment
Share on other sites

15 hours ago, slartabartfast said:

So are you hearing a crackle (static electricity does not produce noise until it is discharged, but a crackling noise is often so described) or a hiss? The noise is probably coming from the signal chain that includes the electronics of the input devices, the computer and the output devices. The software (Cakewalk, drivers etc.) should not produce noise, although it will faithfully propagate, measure, and record it. Software can be responsible for distortion of a signal, (clipping, dropouts etc.), but if you are not feeding it a signal, then the signal must be coming from somewhere else. The electronics themselves generate noise at a low level. Even the best equipment will necessarily add to the noise floor unless it is turned off and kept at absolute zero--there is thermal noise in every active circuit. The circuits used in audio are also frequently subject to picking up electromagnetic radiation from the environment. The fact that an EMR source is not readily apparent in your work room is hardly dispositive. If your computer is not shielded so well that a nearby radio will not pick up anything , then it can still pick up things like neon signs from blocks away.

A 72 dB difference between signal (what you want to hear in this case) and noise (the stuff you do not want to hear) means that the digital audio signal at clipping (0 dB) is about 4000 times the noise you are measuring. If you turn up the gain enough on any equipment, you can probably probably produce a hiss that will be audible against a "silent" signal. So to prevent annoying noise, you can reduce the noise level or decrease the gain required to hear the signal. Likely you are not recording anything that peaks 70 dB higher than the silence between songs. Also likely, you are not going to be able to do much to actually decrease the noise in your system. So a better option might be to record at high enough levels so that anyone listening to the recording will turn his volume knob down to the point where the noise will no longer be audible. 

https://hub.yamaha.com/audio/music/what-is-dynamic-range-and-why-does-it-matter/

Thanks for taking the trouble to provide all that detail and for the link, I will give it a read.

It is a steady hiss really, sorry if the term static caused confusion (I'm not especially technical in this area). Once I have ruled out the audio interface and/or the non-specific ASIO drivers, I will have a look at the dynamic range, perhaps there is something I an do there to work around the problem if I can't eliminate it. 

Link to comment
Share on other sites

Was your Gain knob all the way left or not? In other words, is pre-amp working?

The pattern (continuous increase) is normal. A bit more at 50/60Hz is also fine. 660Hz can be your setup specific (or interface specific... I don't have this model).

In any case, if your Gain knob is all the way right (max possible amplification), with disconnected input -72dB is better then usual for middle-range interfaces (sure, that is also max gain value dependent). If you get it with Gain all the way left, that is too high noise level for middle-range interfaces.

But (a) this interface is entry level, (b) -72dB is not bad for home studio (c) guitar produce way more noise (especially entry level guitars).

Link to comment
Share on other sites

1 hour ago, chris.r said:

Hi Mike, try with unplugging the charger from laptop for a moment and check if the noise is gone. If it's gone then it's ground loop, if not, then something else like the Uphoria is having issue or something else in the chain.

Thanks for this, but I do get the exact same issue with the laptop unplugged from the mains (running on battery)

Link to comment
Share on other sites

31 minutes ago, azslow3 said:

Was your Gain knob all the way left or not? In other words, is pre-amp working?

The pattern (continuous increase) is normal. A bit more at 50/60Hz is also fine. 660Hz can be your setup specific (or interface specific... I don't have this model).

In any case, if your Gain knob is all the way right (max possible amplification), with disconnected input -72dB is better then usual for middle-range interfaces (sure, that is also max gain value dependent). If you get it with Gain all the way left, that is too high noise level for middle-range interfaces.

But (a) this interface is entry level, (b) -72dB is not bad for home studio (c) guitar produce way more noise (especially entry level guitars).

Thanks for the info. Unfortunately, I get the exact same signal profile with the input gain knobs on either mic/line in or instrument inputs both all the way down (left) or up (right).

I appreciate the comment about the quality of the interface, I thought I'd see what I could get away with for a low cost. I have spend a reasonable amount on decent, balanced cables throughout, as well as effort on shielding/grounding the guitar, but If it does eventually prove to be the interface/drivers (or lack of) then upgrading to better quality kit is fine... it worth a shot I guess and hey... I'm learning stuff, right!? 😅

Link to comment
Share on other sites

2 hours ago, Mike Bond said:

If it does eventually prove to be the interface/drivers (or lack of) then upgrading to better quality kit is fine... it worth a shot I guess and hey... I'm learning stuff, right!? 😅

 

2 hours ago, azslow3 said:

(a) this interface is entry level, (b) -72dB is not bad for home studio (c) guitar produce way more noise (especially entry level guitars).

So, it's not a bad interface, but not very good, either. You can certainly learn and create with it, but you might consider something better in the future.
Good luck!

Link to comment
Share on other sites

Are you portable? You might try to record your mic somewhere else and see if you get the same amount of noise. It could be that your mic cable is working like an antenna.

Depending on where you are, it might record less noise. This would at least be a clue to the issue. If that is the issue, a shorter, shielded mic cable should catch less noise.

I think this tends to be more of a thing with guitar cables, but still, you could check.

I'm guessing you just have a noisy mic on a noise pre amp and you are like, hey! what's with the noise. One way to reduce the relative amount of noise is to play louder when you record. 🙂

Edited by Gswitz
Link to comment
Share on other sites

17 hours ago, Mike Bond said:

Thanks for the info. Unfortunately, I get the exact same signal profile with the input gain knobs on either mic/line in or instrument inputs both all the way down (left) or up (right).

I appreciate the comment about the quality of the interface, I thought I'd see what I could get away with for a low cost. I have spend a reasonable amount on decent, balanced cables throughout, as well as effort on shielding/grounding the guitar, but If it does eventually prove to be the interface/drivers (or lack of) then upgrading to better quality kit is fine... it worth a shot I guess and hey... I'm learning stuff, right!? 😅

With gain all the way up, that noise level is in fact great (even for middle range interfaces of the same age). I have just re-measured one of my interfaces, with the same plug-in it hit -48dB at 20kHz with full gain (well... not sure gain ranges are comparable).

What I have not realized before in that UMC22 is 16bit interface. For 16bit signal the profile you see (except 660Hz pike) is almost the best you can get. If you look at real samples, you will see something like:

bitnoise.png.dd0e0cf5edff5e0d35557981a25d781f.png

So, least significant bit is flipping. When "converted" to frequency domain, you get the profile you see.

When you record in 24bit, least significant bit has smaller amplitude. So overall profile is lower. But switching to 16bit (on the same interface) produce the same level as yours.

Fazit: change interface to 24bit capable one if you want lower noise profile. Note that is not a claim you really need lower profile.

Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...