strange noise

[steve]

I wouldn't rule out the laptop being the culprit as an impossibility - as suggested previously it could be noise from the USB power supply (even if the Zoom is powered independently the USB power is still supplied to the Zoom), but this would still suggest that the Zoom is partly to blame as noise from the USB supply should be rejected by the Zoom and not encoded into the audio data.

Have you tried adjusting the input gain on the Zoom (and/or the output level from your guitar) to see if that makes any difference?

[bgravato]

I'm assuming hp laptop is running windows... On audacity's preferences, under devices tab, what devices have you selected for host, playback and recording?

[Reto]

I recently went through a nightmare with noise over USB 2.0 connection. Here is one site that talks about it: http://archive.evaluationengineering.co ... 602usb.htm

In short, my old laptops running USB 1.1 don't have any noise, and my new Gateway i3 laptop has noise. I tried switching out my Alesis 4 channel USB mixer with another mixer, thinking it was defective. 4 mixers later -- two of the same type, and two different brands, I had the same noise on the Gateway. With nothing connected but the mixer itself (i.e. no inputs coming into the mixer), and noise remaining, I compared with my old laptops. They are clean (and all run USB 1.1 BTW). So what's the deal I thought, is my laptop bad? So today I tried out a slew of other laptops at Best Buy, admittedly a noisy environment to begin with (note I do I have an A/C power adapter noise suppressor which helps to some extent). In any case, every laptop I tried -- Dell, Sony, Asus (sp?), HP -- other Gateways -- all models were very noisy on the USB mixer connection. Very noisy and disappointing. The only thing that was very clean at Best Buy was ... guess what ... a MAC laptop. That was clean, running Audacity on it. No noise on the USB input, or much much less. It was less than -48db with mixer running empty inputs at their standard settings (vs. about -30db to -20db for the other laptops, which is very audible noise).

Solutions since this afternoon: I have raised the gain on my wireless mic system output, that feeds into the mixer, to maximum levels. This helps as it allows me to turn down the mixer gains to lower levels, where the USB noise ends up feeling in. Simply put: the lower the gain levels on the mixer, and the hotter the signal coming into the mixer, the better.

Another step I haven't tried -- don't have the laptop here -- is seeing if I can switch the USB connectors to run version 1.1. That may work well based on the article above. Either USB 2.0 is the problem (but apparently not for MACs, or do they simply communicate using USB 1.1 with the mixer), or many of the newer laptop construction have bad noise issues -- but I haven't seen that much blogging about this subject. Side note: my desktop computer once accidentally downgraded to USB 1.1 and it was a bios setting there - not in the Device Manager. Perhaps the Gateway laptop has a similar bios setting to use when I don't need the USB 2.0 speeds (which aren't required for audio recording), and that will fix the problem. So, you might see if you can downgrade your USB version/connection to 1.1.

Other solution is apparently there are USB -> Ethernet converters, but does Audacity support an ethernet type recording input? This is my current question -- perhaps someone has an answer on this forum. That might be another solution.

P.S. I also tried about 4 different powered USB hubs hoping that might shield the noise coming from the laptop. No dice -- noise was worse in every case with hub.

[bgravato]

That's a very interesting posting by Reto!

I'm happy I have a mac laptop. My Art USB Dual Pre is very silent on my mac. I haven't tried it in any other laptop. I tried it on my desktop PC (Asus motherboard) and from what I remember I think the noise level was on the same level as on the mac.

If the noise is coming from the usb it might be through the 5V power supply from the USB connection (pins 1 and 4). I don't know if there's any usb-usb filter for that on the market, but might be an interesting product to develop... (I better register the patent )

I haven't read the article on the link you posted yet, but probably the higher noise on the 2.0 connection might be because it has higher transfer rates which probably means more noise, specially at high frequencies...

That might suggest that the noise might also be going through the RX/TX connection (pins 2 and 3) which is not being properly filtered on the mixer and passes through to the analog part of the mixer...

[steve]

For comparison information, I use a Behringer UCA 202 sound card on an Acer 5735 laptop.
Noise level is -78.3 dB with the input level set at 100%
This gives a 0.4 dB drop if I link the output directly into the input.
Testing the frequency response - flat (within 0.1 dBfs) from 20Hz to 19kHz (-2.5 dB at 20kHz)

Using USB 2.0 port with Linux Ubuntu 9.04 operating system.
Results identical whether running from mains power or batteries.

[bgravato]

What's the proper way to measure noise level?

In lack of a better way I used Contrast Analysis tool (see screenshot below).

This is for my ART USB Dual Pre connected to my MacBookPro USB Port, powered from the usb only. No mics connected to the preamp inputs. Both channels' gain at max. Phantom power on (with phantom power off the noise level is pretty much the same anyways).

Measured noise level, using contrast analysis tool, was about -77dB.

audacity screenshot

Screenshot - 2010-06-09, 15.21.14.png (153.17 KiB) Viewed 2214 times

[steve]

Usually pre-amp SNR is measured with the input short circuit. This is because high gain amplifiers are often noisier when the input is open circuit, but microphones and other input devices usually have pretty low impedance (low resistance to AC), so short circuit is a better approximation than open circuit. Some pre-amps do not vary very much, particularly low gain pre-amps (Line level inputs are often close to unity gain).

Also, what we are measuring here is the peak level of the noise, whereas it is more usual to measure the RMS level, and the level may also be "weighted" according to frequency content. I'm not sure if it's possible (certainly not easy) to get a true SNR measurement using just Audacity, but reading the peak level will give some indication for comparison purposes (and will always be higher than a proper SNR measurement).

-77dBfs (peak) at full bore is pretty good - my laptop internal sound card reads -34dB using the same method that you describe (which is rubbish )

[bgravato]

Looking at my screenshot looks like the peak is at about -66dB. I'll try later to short circuit the inputs and see if there's any difference.

I also noticed a considerable decrease in noise level when I lower the gain on the preamp from the max (48dB) to 30dB. Below 30dB the difference in the noise level is very subtle.

I'm not at home now, but I'll try to do some more tests and take some more screenshots later, both on the laptop and on the desktop pc (and also for my internal soundcard line-in).

[aviator]

yay =) I've found the mistake... it was windows 7 usb audio driver
now i have original zoom g series audio driver and new noise level is -77dB =P

[steve]

now i have original zoom g series audio driver and new noise level is -77dB =P

Congratulations. That's more like it