are laptop USB inputs just (low-level) noisy? (for recording audio)

[Sokel]

The input is RCA,. The device has RCA in/out , headphone 1/8" out, and USB out.

Either of its outs can be run into its in, of course. If that helps totest anything.




IIRC it is there with an input signal (my guitar), too. I will double check.

I still don't grok what 'shorting its RCA input' would mean unless it involves one of these



a-ha. I was going to say, how is any of that relevant if the desktop USB noise performance is 1000x better than any laptop USB?



You people...

what would 'run a thick wire from any laptop's shield connection to a solid ground around your house' look like? What is my laptop's 'shield connection'?


You and antcollinet converge here

I could use the laptop AC power source, which has a ground plug, instead of battery power, is that worth a try?

To the last question, the ground plug probably earths the brick and stops there but that may not apply to all laptops.

To test with the thick wire, just connect one end to a solid earth connection (like some bare metal under a radiator, or VERY carefully to the ground connection of a receptacle) and the other end to an unconnected USB shield (the shell) .

For testing it will be enough to see and compare if any of the junk is reduced.

 

[antcollinet]

I still don't grok what 'shorting its RCA input' would mean unless it involves one of these

That'll do it.

IIRC it is there with an input signal (my guitar), too. I will double check.

Guitar is a high impedance source with low level signal requiring a lot of gain - noise can be picked up by the pickup coils themselves, or in cable to input.

 

[babadono]

I still don't grok what 'shorting its RCA input' would mean unless it involves one of these

That'll do it.

I humbly suggest a 1 k ohm resistor from center pin to shield on the RCA shorting plug to more closely simulate the output impedance of the driving device. Unless you are wanting to short unused inputs and leave the shorts intact.

 

[krabapple]

That'll do it.

so, two, then, one for L one for R?

 

[antcollinet]

so, two, then, one for L one for R?

yes. Though I'm going to have to back track through the thread to remember why we advised it

 

[popej]

What about system audio settings? The same for laptop and workstation? A simple mute on input could really reduce the noise

I don't know, what ADC are you using, but total silence on workstation is suspicious.

 

[somebodyelse]

yes. Though I'm going to have to back track through the thread to remember why we advised it

Just to check that it wasn't just picking up interference on a floating input.

 

[krabapple]

yes. Though I'm going to have to back track through the thread to remember why we advised it

FWIW today I tried using a grounded power pack to run a laptop, intead of its own battery. Results were same.

You guys might lose your minds if you saw the waveform variation I see, simply by hitting 'record' multiple times in series on the same laptop. There is certainly a stochastic element to the noise. Granted, again, I'm recording at an absurd level, and then for comparing waveforms, normalizing to 0 dBFS. Like making a fly into a jet engine.

But the fact still remains (confirmed again today): when using the desktop's USB in, the same absurd procedure produces: silence.*

I still don't quite understand why you are all focused on the potential noise-gathering properties of the unused input jacks of the ADC devices (which again, I have varied) when the determining factor seems to be the recording device (laptop/computer). The USB ADCs I used so far were:

The two red ones are brand new, and one of these is not like the other two, but they are all USB-powered from a connected device. I'm loath to buy an externally powered USB audio interface just to chase down the source of these tiny noise gremlins, though I guess that would be one of the next experimental moves.

I will try the shorting plugs simply because it's easy, cheap, and I prefer to record to laptop at home than to desktop at work. If they don't change anything, I'll decide whether the gremlins will haunt me enought to matter.

(Last, I did also try running the headphone out of the UCA202 ADC (the silver thing on the left) to its L/R RCA inputs, just to see what happens. With its monitoring off the result (on a laptop) was the same low level noise as before; with monitoring on, same again, until I cranked the Volume dial to max or near-max, whereupon a bizarre-looking ~6 Hz wave appears. Same happens with the desktop, though the shape is different. So, that is something from the ADC's headphone circuitry)




*OK, there is a tiny bit of noise -- a 'ping' sound when it is magnified grotesquely-- within the first second or so. It's only visible/audible using the ridiculous magnification regime I described. After that: black silence.

 

[Alan O]

FWIW today I tried using a grounded power pack to run a laptop, intead of its own battery. Results were same.

Unless it's actually grounded from the mains all the way to the laptop, I wouldn't have expected it to make any difference.

 

[somebodyelse]

Since I've got a UCA-202 I thought I'd see if I get the same issue. It's an older one with the real TI codec, not the later Coolaudio one in the UCA-222 tested here with worse performance. As I understand it the 202 and 222 have identical hardware, but a different driver which offers a few more options. On my laptop I get the same silence you do with your desktop - the spectrogram is all black. I tried different ports, both with and without PSU which is somewhat leaky. On my HTPC there's just a hint of noise at the lowest frequencies on the spectrogram, nothing like you're seeing on your laptops. It's the same with my desktop, via a bus-powered USB3 hub. All are linux - I don't have Windows anywhere at the moment.

The UCA-202 does have a known crosstalk issue between playback and record, but that doesn't extend to the Scarlett as far as I know. You'll find various descriptions of the 202 crosstalk issue on the web, and a few proposed fixes. Given you're seeing the same noise issue with 2 different hardware designs I'm starting to suspect a software cause on your laptops, not a hardware one. Perhaps some processing/effects happening in the background that you're not aware of? The 'comfort noise' sometimes added in VoIP systems springs to mind. I don't know the Windows audio system well enough to give hints on how to check that though. On linux Audacity uses direct access to the audio device, skipping any effects that might be present.

 

[antcollinet]

FWIW today I tried using a grounded power pack to run a laptop, intead of its own battery. Results were same.

there is a good chance the grounded power pack doesn't ground the laptop - mine certainly doesn't

If you have a multimeter can you measure the resistance from the laptop case (or grounded connector if it is plastic) to the ground pin on the power pack?

 

[rdenney]

Laptops have grounded power bricks, but are DC with two wires to the laptop. The DC ground is likely isolated from the power supply's earth ground. That might not be true in a desktop--but I'm at work and can't measure anything. Is the DC ground to a desktop motherboard (whence comes the ground for the USB connections) at earth-ground potential by the desktop power supply? I suspect it is, because the metal case would have to be earth-grounded for safety, and accessory cards in a desktop are intentionally grounded to the case via their faceplates.

We used to have to run desktop computers with an isolated earth ground (recognizing the danger--but we were a bunch of electrically savvy folks) when using them to log contacts and feed base-band audio and control signals to radio equipment in weak-signal radio contests. All devices were powered by a large generator on site. When the desktops were earth-grounded, we would get noise intrusion into the (extremely) sensitive receivers, which were running signal/noise ratios of 130 dB or better, with extreme RF amplification of incoming signals. Any noise anywhere was deadly. When we switched from desktops to laptops, the problem went away. This is the reverse problem than Krab is experiencing, and the connections we were making were not through USB ports so it may not be relevant. But it suggests to me different grounding schemes in desktops versus laptops. In Krab's case, the connection of the USB ground to earth ground (if the desktop's power supply works that way--I'm asking) is draining ground loop noise. The laptop isn't, and ground-loop noise may well be leaking into the analog audio circuits of the ADC through its USB power source.

Grounding the inputs of the ADC may well eliminate that noise present the analog audio circuits even if they are fed from the laptop via the USB ground. That's why it's a test worth conducting, it seems to me. Even if the laptop were strictly battery-powered, the batteries aren't at the 5 volts used by USB and so there must be some switching going on to create the various voltages used within the computer.

Rick "whose main ADC is a Benchmark ADC-USB that does not use the USB connection for power" Denney

 

[somebodyelse]

Laptops have grounded power bricks, but are DC with two wires to the laptop. The DC ground is likely isolated from the power supply's earth ground. That might not be true in a desktop--but I'm at work and can't measure anything.

In my experience it's a small minority of laptops that have continuity between mains ground and output ground. Conversely most desktops have continuity between mains ground and chassis. It became memoriable 20+ years ago when some of our corporate laptops got recalled because their ground could exceed 50V from mains ground. When we were using them for measurements they were on battery power, and 'ground' was somewhat variable around 25kV overhead lines on the railway.

 

[DVDdoug]

OK, there is a tiny bit of noise -- a 'ping' sound when it is magnified grotesquely

There is always SOME noise in active* analog electronics. That would the analog sides of ADCs and DACs. The mic input on an interface is usually the worst because the high-gain preamp amplifies noise.

Ignoring ground loops or interference**, it's usually "hiss" and some electronics are quieter than others.

Once you are in the digital domain there is no (additional) noise. A digital audio file full of zeros is dead silent at minus infinity dB. (Once there is some signal/data you quantization noise on top of the signal, and you can hear it with 8-bit files.)



* Technically, passive electronics (like resistors) generate thermal noise but it's VERY low-level. It's worse at higher temperatures and/or higher resistance.

** I would classify power supply noise as interference.

 

[antcollinet]

* Technically, passive electronics (like resistors) generate thermal noise but it's VERY low-level.

It is - but it does tend towards being the limiting factor when it comes to designing the output stages of 24bit (and higher) dacs. (-144dB is also very low level )

***
1kOhm resistor - 20C, 20kHz - thermal noise is about 0.57uV

1LSB of 24bit audio with 2V output is 0.12uV - a factor of 4 lower.

It is no coincidence that even the best DACs measured here are struggling to achieve 22bits of dynamic range.

 

[Sokel]

FWIW today I tried using a grounded power pack to run a laptop, intead of its own battery. Results were same.

You guys might lose your minds if you saw the waveform variation I see, simply by hitting 'record' multiple times in series on the same laptop. There is certainly a stochastic element to the noise. Granted, again, I'm recording at an absurd level, and then for comparing waveforms, normalizing to 0 dBFS. Like making a fly into a jet engine.

But the fact still remains (confirmed again today): when using the desktop's USB in, the same absurd procedure produces: silence.*

I still don't quite understand why you are all focused on the potential noise-gathering properties of the unused input jacks of the ADC devices (which again, I have varied) when the determining factor seems to be the recording device (laptop/computer). The USB ADCs I used so far were:

View attachment 521923 View attachment 521924 View attachment 521925


The two red ones are brand new, and one of these is not like the other two, but they are all USB-powered from a connected device. I'm loath to buy an externally powered USB audio interface just to chase down the source of these tiny noise gremlins, though I guess that would be one of the next experimental moves.

I will try the shorting plugs simply because it's easy, cheap, and I prefer to record to laptop at home than to desktop at work. If they don't change anything, I'll decide whether the gremlins will haunt me enought to matter.

(Last, I did also try running the headphone out of the UCA202 ADC (the silver thing on the left) to its L/R RCA inputs, just to see what happens. With its monitoring off the result (on a laptop) was the same low level noise as before; with monitoring on, same again, until I cranked the Volume dial to max or near-max, whereupon a bizarre-looking ~6 Hz wave appears. Same happens with the desktop, though the shape is different. So, that is something from the ADC's headphone circuitry)




*OK, there is a tiny bit of noise -- a 'ping' sound when it is magnified grotesquely-- within the first second or so. It's only visible/audible using the ridiculous magnification regime I described. After that: black silence.

Grounded PSU or not, laptops are rarely grounded through them, if at all.
If you don't actually ground the machine itself , testing with 2-prong or 3-prong PSUs won't tell you if there might be a difference.

 

[krabapple]

Since I've got a UCA-202 I thought I'd see if I get the same issue. It's an older one with the real TI codec, not the later Coolaudio one in the UCA-222 tested here with worse performance. As I understand it the 202 and 222 have identical hardware, but a different driver which offers a few more options. On my laptop I get the same silence you do with your desktop - the spectrogram is all black. I tried different ports, both with and without PSU which is somewhat leaky. On my HTPC there's just a hint of noise at the lowest frequencies on the spectrogram, nothing like you're seeing on your laptops. It's the same with my desktop, via a bus-powered USB3 hub. All are linux - I don't have Windows anywhere at the moment.

To replicate what I see/did you have to set your input recording level to maximum or close to it. (Which, again, is not what I have ever done before, in the dozen times I have recorded using the device). The waveform of such recording is still essentially flat. Use spectral view to see the noise. Or, if you normalize the waveform to some silly value like 0dBFS, you can get the highly magnified view of it in the time domain.

Did you do that?

[My UCA202 was manufactured in July 2011 and purchased around then. My UCA222 was bought new last month, no date code on it.]


Maybe I should emphasize: At normal recording levels with these devices, the spectral view is essentially black, and noise is not audible at my listening levels. I only discovered it in the first place by 'fooling around' in curious-cat mode.*

Anyway, to update this adventure:
Tried my new shorting plugs on inputs, same noise result as before for both Behringer ADCs. (I returned the Fosi days ago)

Also tried eliminating external ADC interfaces entirely and using the laptops' internal ADC instead (via 3.5 mm microphone/headphone TRRS plug, or via USB -- I bought very cheap microphones that have these different termini, to use as testing devices). Noise results (again with Audacity as recording software) are...curious, but I'm still messing around with it. (For one of the laptops, just getting it to STOP using its built-in screen mic and use the 3.5 mm jack mic instead, was an absurd hours-long time sink. The other laptop 'got it' right away. Both run Windows, one is 10 one 11) I'm awaiting delivery of guitar cables that are quarter inch at the guitar end and USB or 3.5 mm TRRS at the other, as part of these experiments.



*FWIW, I found a 2014 web page by Jim Lesurf that measured ADC performance of three cheap USB-powered interfaces, including the UCA202 . Relevant part:

The UCA202 will play and capture 16bit and 8bit samples in stereo or mono at 32k, 44.1k, and 48k sample rates. Hence it is limited to a maximum of 48k / 16bit. It will also capture at various lower sample rates down to 8k samples/sec. Overall, it therefore behaves pretty much like an old fashioned non-audiophile ‘sound card’. However when capturing 48k/16bit stereo it showed a THD of only 0.008% for signals at the -1dBFS level. The maximum input level the device would accept without clipping was about 1.25Vrms, and the noise background was at -81dBFS. This is hardly state-of-the-art, but better than many internal soundcards, and it can be bought for around 25 pounds. So may be useful for some undemanding source material. Its main advantage is that it is, small, convenient and works with no fuss.

 

[krabapple]

Grounded PSU or not, laptops are rarely grounded through them, if at all.

That seems at odds with multiple sources I read on the web, e.g. this, but...it's the internet.

If you don't actually ground the machine itself , testing with 2-prong or 3-prong PSUs won't tell you if there might be a difference.

FWIW these laptop tests were done in my work office, which has lab-grade three-prong outlets and I presume this institutional building's system is properly grounded. But who knows?

My desktop (which has superior noise performance in my tests) is plugged in to these same outlets.

 

[Sokel]

My desktop (which has superior noise performance in my tests) is plugged in to these same outlets.

That's a good index about buildings solid ground and electrical installation.

But laptops can not take advantage of it as they have no way to connect to it unless their power input is a 3-prong one (not the PSU, the laptop itself)
The only way is to do it yourself in order to test if it works, just a wire from laptops USB's bare metal to a receptacle's ground (carefully so not to put it at live or neutral)

That would be sufficient for testing.

 

[Alan O]

FWIW these laptop tests were done in my work office, which has lab-grade three-prong outlets and I presume this institutional building's system is properly grounded.

That'll make absolutely no difference if the ground is not fed all the way back to your laptop. The power brick for my MacBook, for example, has a three-pin (UK) mains plug with what looks like a conductive earth pin... but the plug itself is interchangeable (to use different plugs in different markets) and if I detach it I can see only two connections actually make it into the power supply box itself. No laptop I've personally ever used has been grounded all the way from laptop to mains plug.

Here's a photo of the detached (UK) plug...