Why USB noise is harmful to HiFi

You cannot hear USB noise so why should audiophiles be concerned with it?

[thefloop]

Love the videos Paul. As always keep em coming!!!

And like i said before, every time this comes up i will respectfully point out that There is zero detectable with ear, nor measurable noise or jitter with my interface used as dac or adc, on either thunderbolt driver or regular usb. This whole home audio debate still confuses me. USB, when properly implemented, has been the superior dac/adc connection for like a decade now. That's a big part of why it's used exclusively in the mixing/mastering world now a days. We used firewire before that because it was also better then any other method of connecting a dac/adc to a computer source, and also allows a much broader customization of I/O parameters. And one thing is for certain, we wouldn't use it in the creation of music itself if noise was a problem with the implementations of used. Modern USB implementations can have a lot more bandwidth available to them as well. You definitely can't cleanly run a PCI express x16 lane with literally hundreds of GB of of throughput per second on toslink with practically zero data transfer errors. This is a cake-walk with current thunderbolt usb.

[forbeginnersandbeyond]

In other words, the digital noise which you cannot hear degrades the resulting audio sound that you hear. Meaning the sound you hear is not the noise component, but a degraded aural sound (or unfaithful reproduction of the source audio).

[ThinkingBetter]

Theoretically, yes, the DAC oscillator circuit might yield more jitter if the power rails are more noisy from an external noise source. It would however be interesting to see actual measurement showing such case. Some DACs might be poorly designed to easily allow USB noise to affect the DAC clock. I would love to see actual measurements or blind tests of this subject. Why is it that these gray-zone audio concerns never get a proper measurement or blind-test? In this case it's validation of "USB noise immunity on DAC clock jitter". People happily swallow someone's opinion of something sounding more resolving, better soundstage etc. when such opinions always are without even a quick A/B switch allowing proper comparison. I believe a lot in Paul aiming at being genuinely honest in every statement made, but then why not take the lead in this industry and professionally prove or debunk every claim? Perhaps PS Audio DACs indeed are sounding better because of it, but what better way would there be than a proper blind-test and some measurements? PS Audio would be a super audiophile company by doing what the snake oil companies never do...prove those gray-zone topics are really audible or not.

[crazydwarfer]

Blind listening tests have a huge effect on all claims about similar issues. Thus, the best way to get rid of any misconceptions is to so a blind test the likes of Harman did. BTW, great review of FR10s from Erin Hardison. Awesome speakers it seems.

[curtchase]

Good points. I was expecting a rant about digital noise leaking into the audio amplifiers which could cause issues if amps are trying to amplify noise above frequencies we hear which could lead to more IM distortion and possible tweeter damage, as well as just wasted power making amps run hotter. Note, I cannot prove this theory myself exempt I have heard digital noise in cases where I had a ground loop issue. Right now, I use a TOSLINK cable from the S/PDIF jack on my computer main board to a budget outboard DAC. I expect that be enough to isolate ground issues.

[The.Original.Mr.X]

I had to use galvanic isolation on the USB connection between PC music server and DAC. I could most definitely hear it.

[tomekichiyamamoto]

What USB noise are we talking about here?

Eventual noise in the data transmission line?
(Which I don’t think it can be. Data are decoded by the usb receiver which transmits to the DAC board, no? )
Are you talking noise from there?
In that case, I’d like to get more detailed informations. 😮

Or noise from the usb power line, 5v???
(In that case, it would be enough to make use of separate audio PSU, no?)

[garyharper]

I like how people say double blind tests don’t work, only when you don’t like the result.

[AlexAparicio]

Everything that Paul talks about it applies to rolls-Royce equipment (speakers, amps…) most of us use Honda to Mercedes type of equipment. We don’t have to worry about this stuff

[sonnyvarioni]

I stream from an iPad to a Luxman DAC via USB. No audible difference between that and my CD source.

[manitoublack]

Ground loops and main's hum are genuine concerns as they occur in the audible range. Especially mains hum that's either 50Hz or 60Hz.

USB and digital noise that's in the MHz/GHz range... If you're gear is so poorly designed that it's interfering with the sound and causing audible artefacts; then I'm surprised you're watching this channel. I'd be tossing that bit of gear as there is likely a short somewhere. Like the Ultrasonic noise on SACD files or the High frequency switching in Class-D amps... there are filters that supress that ultrasonic noise.

But if you want to know it's galvanically isolated:
-Toslink
-Optical HDMI cables
-Bluetooth
-WiFi
-Optical USB cables.

But the reality is; it's a non-issue or every person using USB head phones connected directly to a PC would be crying bloody murder from all the artefacts. But their silence is deafening.

[KevlarCondom]

PCs have horrible ground. PCs are almost as bad as cellphones for pumping out EMF, even wo wifi, BT, etc. Everything on a PCB wants to act like an antenna.
Forget about signal noise. PCs put out so much noise to ground, you should independently ground them from everything else. Those red ac plugs in hospitals all have individual ground for good reasons. All pcs in an audio chain, be it at home or in the studio need there own ground away from equipment ground.

All electronics pick up emf. Why the FCC approval sticker mostly states that a device can handle the RF noise, and not become a radio station by itself. The FCC sticker does not intend to say that RF or EMF will not cause problems with the device. It's a game of limiting interface. In any wire or trace on a pcb, there is noise. In any audio circuit, there is noise. If left over from the big bang, to the sun, to a baby monitor, or a microwave oven; keeping out the noise you can control makes everything else easier.

The data stream is not the problem. It's the wires to ground that make antennae, and all rf and emf in a box is shared and must be dealt with. The difference in a recording signal to noise, and a cpu signal to noise is over 20 trillion to 1. Cpus work with voltage differences of today. Keep the noise out.

[jozno]

Truthfully, Paul, "Noise" could be a 3-hour discussion at an audiophile roundtable. It wasn't until I'd gotten more into high-end, revealing audio that I realized tape hiss, a dirty record or a poorly recorded CD weren't the only types of noise to enter a system. Vibration, radio frequencies, static electricity, etc can all create noise that you wouldn't perceive until you removed it. When you rid yourself of it and hear more of the music, you realize it right away.

[NoEggu]

Jitter is mainly the responsibility of your transport.
CD players have a built-in transport.

Audio produced on your computer uses your sound card's (usually built-in to your motherboard) built-in transport.
Even if your computer feeds your DAC a digital signal, your computer is sending the bits, and since it is the device doing the sending, it is by definition the transport.

Reducing jitter will reveal a sonic breath of fresh air.

I once tested an Aesthetix Romulus Eclipse DAC, being fed by an i5 laptop (not an ideal choice for a transport).

I used a USB cable that was included with my HP printer, and also used a USB cable that was included with my USB hub.
Both of the above USB cables wrecked the sound quality. There was nothing musical left in the sound. Everything was crushed, without bloom, without air, without space. It was dull.

I also tested an Audioquest Diamond USB cable. All of the above issues were gone. The music was back. It was wonderful.

Somehow, that Audioquest cable kept noise from the laptop from entering the DAC. Or, perhaps, the Audioquest USB cable allowed the bits to reach the DAC with better precision; better timing.

Perhaps there are DACs that do the same thing, internally, negating the need for a quality USB cable?

I put the printer's USB cable back, and put the USB hub's cable back. Both the printer and the hub work perfectly (although I never tested connecting the Audioquest cable to the printer or the hub -- I can't imagine an improvement in their specific functions).

Another source of noise could come from your DAC (or your CD player). That noise can seep in to your other components.
Using a quality power cord will make a noticeable improvement in darker sonic backgrounds, and air between instruments. You might not immediately notice the improvement. But after a song or two, you will realize that you are in the zone -- sucked in to the listening experience. When you put back the stock power cord to your CD player (or DAC), you will immediately hear the degradation.

So if you demo a quality power cord from, perhaps Shunyata or Audioquest, be prepared to buy it. Once you use it for your CD player or DAC, there is no going back.

[edwardmichaelkenway]

You are absolutely right...some people think that because everything is digital before the DAC there is nothing to worry about...but that is completely wrong as I fully understand the problems with digital timing and also the distortion between the computer and the DAC due to ground problems

[daveduffy]

Are there any published test results you can point to?

[odizcvw]

Jitter is all about timing accuracy. Music with jitter is like looking at a blurry video where the projector was vibrating. Instead of sharp accurate picture it is kind of out of focus. And if the three colors are moving independently the colors will be all messed up. Reducing jitter in a DAC sharpens the image, improves perceived dynamics and particularly improves bass. When it gets really good, the sound becomes natural and analogue in nature.

[olhoTron]

Even when the signal is on the MHz, sometimes you can hear the packets, also controlers going in and out of sleep modes make noises in the audible range

[rosswarren]

Everything Paul says is true, but my understanding (correct me please if I am wrong) is that you HAVE to use an asynchronous digital stream - like USB - from a streamer in order for the external DAC to use its own clock and reclock everything to a better level without as much jitter. Because the DAC makes the two clocks sync up at the same time, it drives the converter directly, so it doesn’t have to rely on the computer’s unstable clock. Using Toslink Coax or SP/DIF optical would limit you to how good the clock is inside the streamer, such that you are wasting some of your money on a better DAC having a better internal clock. Of course, maybe I2S is also asynchronous and offers an even better solution than USB, but does it also introduce "noise"?

[mrronenza]

Thanks for the video, Why in a Coax cable (Also Digital ) there is less noise ?