• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

All that "USB interface" audio noise. And better sounding USB via?

vvcv

New Member
Joined
Mar 19, 2020
Messages
3
Likes
1
Greetings from a new member,

Glad I found this site and this forum, as i've already learned a lot. Not sure I should have posted here, or in Tech. Newbie or perhaps even in the DACs forums.

My question is in regard to everything i've read regarding the USB noise that seems to be 'inherent' to the USB interface. I've read that this USB chain noise can come from the wall, switching power supplies, EMI from the motherboard, PC fan motors and even the USB cable itself. With all this 'stuff' potentially causing noise, can it be anything in the usb chain picking up this noise, amplifying it, and/or causing its own noise? Given all this, shouldn't I be hearing some noise? A little more: If I do not hear noise with out music playing and volume up on all devices, and I do not hear noise during a silent passage on a good (no mic hiss) recording and do not hear noise during very soft piano passages, for example, could it be the DAC filtering the noise out? Could this noise simply be less prevalent than it seems to be? I mean, I see people going so far as to buy fanless PCs to have the system noiseless, not from the acoustic fan noise, rather the fan mother EMI. I mean, heck, with my noisy PC and a USB cable that probably came from an old 1999 portable DVD drive I would think the noise it's all making is probably being picked up by NASA. My source is an everyday laptop grabbing files from local, portable and network drives (WiFi and ethernet) going into one of four DACs (1 tube) into one of four amps (3 tube) into speakers with 96db sensitivity being the most efficient of them. And no, I don't have any of those wonky gadgets that claim to remove noise then magically add something like pushing the violin section further left of the sound stage. No! Some cleaner and quiet rests? Yes, okay. I get the sense that if these things worked, there wouldn't be PC based music noise problems seen so often. Oh, I only mention tube to state that even with that gear there is 0 noise.

I'm not a lucky person, could it just be luck that I do not hear USB noise? And if USB noise is as common as so many reviews say it is, then where did my noise go? Geez, I've never worried so much about a problem I don't have. I feel mad asking, but as far as this kind of technical stuff goes, I have 0 education in it. Aside from seeing people having USB noise problems (from studios to the end game home audio system) which are real problems, I feel that it may only be marketing teams who have noticed, and are now trying to convince us all that we have it and how glorious everything will be without it. So, with these questions in mind, I figured I'd ask publicly, as opposed to asking a few dealers or audio techs and this seems like the perfect place to ask such a question. If any of you guys and gals can either give some 'simple' whys and/or why nots, or perhaps point me to some 'simple' papers/articles, I would be grateful and entertained.

As fast as PC based systems have grown, from music servers to upsampling music players, and USB playing a big roll, I'm curious if anyone has found anything that can add some sound quality within the USB portion of their systems, even if it's not everyone's cup of tea? Like a kit that includes everything to solder my ssd drives to my motherboard then solder my mb to my DAC? :D

Anyway, thanks for reading, my brain hurts from all the noise, so good night
robert
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,201
Likes
16,982
Location
Riverview FL

PaulD

Senior Member
Joined
Jul 22, 2018
Messages
453
Likes
1,340
Location
Other

BDWoody

Chief Cat Herder
Moderator
Forum Donor
Joined
Jan 9, 2019
Messages
6,948
Likes
22,625
Location
Mid-Atlantic, USA. (Maryland)
I mean, heck, with my noisy PC and a USB cable that probably came from an old 1999 portable DVD drive I would think the noise it's all making is probably being picked up by NASA.

Well, as long as you don't hear from them, you're probably fine!

I feel that it may only be marketing teams who have noticed, and are now trying to convince us all that we have it and how glorious everything will be without it.

We have a winner!

If you can't find a problem in your system, no matter how hard you try to, I'd say you can relax and just enjoy your music.

Some have issues, but they aren't automatic, and it sounds like you are fine.
 
OP
V

vvcv

New Member
Joined
Mar 19, 2020
Messages
3
Likes
1


Thanks a lot Paul. The first link (review of the Kimber Cable, a company i've been told, numerous times, should have a great USB cable) has a quote (below) with two important areas to look into. Especially the one I was thinking could be the key to managing potential USB nosie -the DAC,- and now it's been stated by someone I feel I can trust and can look into further. I'll be able to pass this information to a good friend who has become a budding audio enthusiast after he heard some cans of mine. Strange how one certain piece of gear can change a persons perspective. This alone should soak up a good amount of my free time. Thanks again, and quote below, just to be a bit more complete with my reference (again, from your first posted link).

"Discussion and Conclusions
No, bits are not bits. The PC connection can cause noise to travel into a PC so there is some analog aspect to digital audio communication. It is the job of the DAC to be immune to that noise, jitter, etc. And that is certainly the case with vast majority of audio DACs."
 
OP
V

vvcv

New Member
Joined
Mar 19, 2020
Messages
3
Likes
1
Some have issues, but they aren't automatic, and it sounds like you are fine.


Good to know. That gives me some piece of mind. I've really only begun to worry about noise due to moving into a new house we started building about a month ago. I realized that things have a potential to change, even if only minor. But I've always tried to keep things as simple as possible.. The only thing that will be changing, aside from all the power in the new house area, is that I'll be running a dedicated line along with a 220 outlet as well -just in case I want a quiet washer dryer in my office/music room. Will it help, don't know for sure, but not costing much more at all
("if anything") is the quick quote I got from a contractor working in the area. But, that can change pretty quick.
 

pozz

Слава Україні
Forum Donor
Editor
Joined
May 21, 2019
Messages
4,036
Likes
6,827
Last edited:

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
3,334
Likes
3,278
Location
.de
While the possibility for audible interference in such a setup is quite real, it has precious little to do with USB itself - that wouldn't make it any better, but the underlying cause would be a plain ol' ground loop. Quite possibly, either all of the OP's amplifiers are constructed sensibly enough, and/or the laptop power supply design may be doing its best not to unduly provoke such issues.
 
  • Like
Reactions: PSO

Asylum Seeker

Senior Member
Joined
Feb 12, 2020
Messages
414
Likes
295
Location
Guatemala
I don't have any USB noise issues. Why? Because I use mobile, battery-powered devices exclusively. My philosophical approach to this problem is from Matthew - If the computer causes you to noise, cut off the computer.

Oh yeah, no driver issues either.

KISS.
 
Last edited:

Foxenfurter

Active Member
Joined
Mar 5, 2020
Messages
129
Likes
157
Location
London
The only bad effects I have actually heard on USB are clicks and ticks on low powered (by that I mean low bus and cpu speed) devices probably caused by clock synchronisation, and that was 10 years ago!

In terms of USB devices, yes I have "experimented".

I have an ifi USB 2.0 with the weird USB cable that I bought in a moment of stupidity. I have a setup that allows me to A/B stuff easily and I can honestly say I have not heard any difference with this in the chain, other than it makes the setup look really messy. The PSU didn't last that long either!

I also have an intona 2 which I have measured a difference with on a very old USB interface. But I have not been able to hear a difference with when using on my main system. Added to this it causes problems with certain USB devices being recognised. I have also experienced weird volume changes when plugged in. i.e. going massively louder and distorting badly.

So keep it simple is cheaper; neater and less hassle.

I also agree with Fluffy's advice above.
 

Justin Ayers

Member
Joined
Oct 5, 2020
Messages
60
Likes
47
If you have a player that is streaming data on the fly I would replace it with one that puts the entire file into RAM memory prior to playback. That will make all cable transmission issues (other than outright failure to transmit the file) completely impossible. Either the file will get there, completely unchanged, or it will fail to arrive.

The originator of the file (the device it's stored on first) will provide a checksum (a digital signature) that will be matched by the device the file is copied to. If the checksum doesn't match the device the file is copied to will know it and should tell you that there was a malfunction. It should not be mysterious at all.

Given the low cost of RAM memory and the small storage needs of audio files, there is zero reason to design a digital audio system to stream data through cables.

(The only exception is for live instrument performance, via connections like MIDI. In that case you have lag to deal with and USB can be laggier than some other connection types. There is a lot of data involved in that issue, though.)

--------- more detailed information ---------

This is exactly how downloading files from the Internet to your computer works. The file will be exactly as it was on the server you downloaded it from. You can check the checksum and it will be unchanged, as long as the checksum information hasn't been hacked somehow by malicious parties. The only time a checksum won't match is if some third party ("man in the middle") substitutes a different file that seems the same for the one you intended to get. This could also happen if the server you attempt to download the file from surreptitiously switches files on you. Those sorts of things are not normal behavior at all. They are spycraft. So, for playing your audio files you don't need to consider them. The only other way the file would be changed is if your computer is unstable (such as from overclocking or overheating). In that case, you have the extremely serious problem of random data corruption.

Digital files are never supposed to change without manual user intervention. They are not like analog (recorded to tape, for instance — where the tape's physical deterioration will change the data gradually over time: print-through, drop-outs, reduced magnetic intensity, etc.). When a digital file is stored somewhere it's supposed to be unchanged permanently. Only defective functionality can change them (generally catastrophically), like an SSD/hard disk that fails (prompting data reconstruction software rescue that can mangle files), RAM memory that's unstable, bit flips from gamma rays interacting with circuits, etc. Unless some part of the computerized device that stores your audio or the device that plays it is malfunctioning seriously you will find that your files remain static (unchanged).

Because of the generally catastrophic consequences of changing even one bit in a digital file modern computer systems use journaling and other redundancy techniques so they can reconstruct and/or replace damaged files. The filesystem your devices use to store your audio files will be journaled if the device is at all modern. The old FAT32 format is not journaled but it's fine for secondary storage. Be sure to have at least one copy of your files on a journaled filesystem and don't make it exFAT. I have found that one to be unreliable. Unfortunately, Linux, Microsoft (Windows), and Apple (macOS) can't agree on a better cross-platform filesystem. All the deep metadata tracking mechanisms in modern filesystems, desired by intelligence folk, probably contribute to the problems.

Also, although I contrasted digital file storage with analog, the media used to store digital files can physically degrade. One problem that is growing in intensity is the increase in both the smallness of the electronics used for flash (the process node) and the increase in voltage states (originally SLC — single layer flash, then MLC, then TLC, then 3D MLC or TLC, then 3D QLC). (Hard disks are also becoming finer, in terms of the amount of magnetic material data is stored on.) With each extra flash layer the voltage states increase (which is not good from anything but a density standpoint — how much data can be inexpensively packed into a certain amount of area). This increases the problem of voltage drift which can lead to slowdowns and even data loss. So, although the greater size of most flash volumes these days helps, as does 3D flash rather than planar (this helped TLC immensely but QLC has too many states to be very good even with it) — the process nodes keep going down and QLC has a ton of voltage states. Point is that if you want to back your files up (and you should!) it's best to use an enterprise-grade hard disk that hasn't been used much. Really important files should be kept on one of those in an anti-static bag in a fireproof/heatproof safe. Bit flips from gamma rays may be an issue and hard disks can fail at any time. It's best to have a second backup drive, preferably using a different filesystem than the first one. If the files are priceless then have a second secure offsite backup for the other hard disk (a different model but also enterprise-grade). I would keep one in FAT32 and one in a modern journaled filesystem other than exFAT. The FAT32 makes it very simple to move the files on the drive to any platform and to add new ones to it. For files you can live without but want to keep you can put them onto a USB portable hard drive and onto a USB flash drive. Do both. The hard drives in particular like to act up a lot and flash media should be plugged in more often to keep the drift issue under control.

Degradation of files due to the deterioration of storage media will not add noise to digital audio. The files probably won't open at all. If they do they may play for a few seconds then fail. They may be silent throughout. It's not like analog. Digital files do not age gracefully. However, if they are kept backed up well they never age at all.
 
Last edited:

Katji

Major Contributor
Joined
Sep 26, 2017
Messages
2,990
Likes
2,273
Laptop is also portable battery-powered. (Just more complicated if you keep it connected to the PS all the time.)
Mobiles/phones motherboards are evenmore dense, probably more difficult to isolate and avoid noise spreading.
Powerbanks/batteries also cause noise (rectifier circuits or something.)
 

Katji

Major Contributor
Joined
Sep 26, 2017
Messages
2,990
Likes
2,273
The only way I can see noise being an issue due to lag is if the DAC (or similar) is compensating for fluctuations in the speed with which the information is arriving and somehow interpolating the data (adding its own information on the fly based on the data it has) to prevent drop-outs. This would be similar to how audio CDs compensate with error correction when they encounter dust, scratches, and such on discs.
Lag is a different matter from [EMI, etc.] noise. ...And DAC is not doing any such thing.

But, I suppose a poor cable might be marginal in terms of providing the electricity strongly enough to maintain a solid signal. Normally that would mean simply dropping the playback. But, I suppose a designer could try to imitate something like CD audio by using some kind of data reconstruction on the fly. It would only compensate for quite marginal issues with a cable, though, I'd imagine. I suppose it may also be possible for one or both of the USB ports to be marginal in terms of providing adequate power, due to corrosion or poor design. If the electrical signals are just enough to usually make it through but prone to sometimes dropping too low maybe some kind of error correction could come into play.
No...just no.
USB cable must be according to spec.
The problem is noise on the USB output.
You do what you can to get a motherboard / USB board that does not get noise on the USB ouput. And a low-noise PSU.

Personally, if I were designing a device for playing audio files back I would make it so that all of the data would be placed into a substantial RAM space. That would provide a total buffer against drop-outs. It would completely cover for lag fluctuations from cables and any other marginals. But, I'm no engineer and this is all speculative.
[RAM costs money. What type of device for which price point / production cost[?]] You design according to product design specifications, what the boss says, or you get removed from the team or fired. So yes, speculative indeed.
Lag is not caused by USB cables if according to spec. and not faulty.

A checksum is a signature of sorts for digital files. If unmolested it won't change. This applies to downloading files from the Internet, too. If a person checks the checksum it can expose man-in-the-middle tampering, as long as the original checksum you have is the correct one.
USB Audio is not the same as file transfer/copy.
XMOS - Fundamentals-of-USB-Audio(1.0).pdf
 

Justin Ayers

Member
Joined
Oct 5, 2020
Messages
60
Likes
47
The problem is noise on the USB output.
A problem that doesn't exist in an adequate design.

An adequate design stores the files in RAM and plays them back when the file copy has completed.

You design according to product design specifications, what the boss says, or you get removed from the team or fired.
Then the customer can fire the boss for mandating a substandard design.
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,424
Likes
7,940
Location
Brussels, Belgium
It’s all about how your DAC is designed. Lowering the noise on the output in the computer can help if you have a shit USB input on the DAC.

There are many reviews here that show you the performance of the DAC’s USB input in regular day to day use (Plugged in into a computer). What Amir measures with sensitive equipment is the analog output, he never uses fancy low noise input boards for the input signal.
 

Katji

Major Contributor
Joined
Sep 26, 2017
Messages
2,990
Likes
2,273
A problem that doesn't exist in an adequate design.

An adequate design stores the files in RAM and plays them back when the file copy has completed.


Then the customer can fire the boss for mandating a substandard design.

:facepalm:
The problem exists in reality. You can go wonder - or speculate - about motherboard designs and computer designs.

"An adequate design stores the files in RAM and plays them back when the file copy has completed."
smh, wtaf.

"Then the customer can fire the boss for mandating a substandard design."
Quite far from reality. (Like the rest.)
 
Last edited:

dfuller

Major Contributor
Joined
Apr 26, 2020
Messages
3,337
Likes
5,053
Most of the noise issues I have only occur after I unbalance a signal (say, if I'm reamping a guitar DI), and even then I'm about 90% sure it's not from the USB itself (it's the same frequency as the noise my magnetic hard drive's motor throws off if I hold a guitar near it).

So, the lessons here? Balanced interconnects are way less likely to end up with noise in them than unbalanced, and the noise isn't necessarily from the USB connection itself.
 

Justin Ayers

Member
Joined
Oct 5, 2020
Messages
60
Likes
47
Most of the noise issues I have only occur after I unbalance a signal (say, if I'm reamping a guitar DI), and even then I'm about 90% sure it's not from the USB itself (it's the same frequency as the noise my magnetic hard drive's motor throws off if I hold a guitar near it).

So, the lessons here? Balanced interconnects are way less likely to end up with noise in them than unbalanced, and the noise isn't necessarily from the USB connection itself.
Balanced connections are analog, correct? Analog is inherently noisy.*

Digital is not. Digital transmission is 100% not noisy if implemented with a minimum level of competence.

There is zero speculation involved in stating that digital audio playback can very easily (affordably), with 100% guarantee, have no noise added via digital data transmission via cables, interconnects, etc. The file is transmitted, unaltered. If it is altered in its digital state that's a design defect unless it's intentional processing (like adding a digital reverb to the file or compressing the sound to fit a lower dynamic range). Intentional processing (changing a digital audio file's audible characteristics) is a separate optional functionality paradigm. It is not required and may not even be desirable.

The worst that can happen (with a minimally-competent digital data transmission design) is for the use of partial buffers (instead of 100% file completion transmission prior to playback) to cause drop-outs. In that case, if the data stream is interrupted (such as if a cable comes loose or the transmission ability was so marginal that some sort of fluctuation in power caused it to drop below the ability to transmit successfully) — it causes sudden loss of audio, at least once the buffer runs out. You can see how that works by playing a video on Youtube and turning off your Internet on the computer while it is playing. Most videos will only be buffered partially and playback will suddenly stop when the data transmission has been halted and the buffer's contents are exhausted. A sophisticated player would be smart enough to fade out the music when it sees that the buffer isn't being filled properly anymore, too.

Buffering like that makes sense given the very limited RAM available to a web browser for one webpage, connection bandwidth (which can be a very thin data pipe for people in rural areas in particular), and the use of video data (which is a pig compared to audio when comparable in quality). For enthusiast audio in the local environment (rather than via the Internet) it generally doesn't. File storage is cheap and so is RAM. Portable extremely compact reliable file storage is cheap and easy (USB flash drives). Digital cabling that will transfer files successfully is cheap.

The jarring extremely unaesthetic impression of sudden dropped audio is a good reason alone to avoid using a partial streaming buffered solution in favor of transmitting the entire file into RAM storage prior to playback, unless a fade-out is implemented (which is still inferior to simply playing back the music file the listener is demanding to hear). Some degree of preloading/buffering can then be used for "scrubbing" and very quick track changing, if desired — as the amount of RAM space dictates how much of the playlist/library can be stored at a specific time. Extremely long tracks, especially if they have been upsampled into order to have complex EQ/treatments applied, may mandate some degree of buffering (due to file size bloat) but the buffer should be considerable so the drop-out potential is minimized as much as the customer is willing to pay for. (Things like upsampling are part of the separate optional functionality paradigm, so it is a false dilemma to use them to argue that the move-to-RAM solution isn't cost-effective.

(The cost for enough RAM to store even long classical symphonic tracks in Redbook, CD audio's 16-bit 44K format, with lossless compression, is minuscule in the big picture of enthusiast audio, including very affordable enthusiast audio. Even skipping the compression/decompression—which requires some CPU speed, and therefore a fast-enough CPU to be added, and can add delay without enough preloading buffering—and using uncompressed files, the RAM cost is insignificant. Playing really bloated—from upsampling—files with huge time durations directly from RAM without buffering is an edge case, in terms of this discussion, and speaks to the higher-end equipment niche—for people with deep-enough pockets where the RAM cost is also negligible in the big picture.)

And, to top things off, it's very inexpensive to get a USB flash drive and have the files on that, obviating the need to transfer them via cables while listening in the first place! (Higher-end equipment would have an NVME drive that is user-upgradable, ideally with a second drive for automatic backing-up of the data for greater reliability. In the big picture of midrange and above enthusiast audio, NVME drives are... wait for it... inexpensive.)

*This is why I think it's silly that so much analog connection is used in enthusiast audio. Optical has been affordable for a long time. It also makes a lot more sense these days to localize file storage, via cheap removable USB flash, so cabling isn't needed much, if at all. Worrying about balanced vs. unbalanced is akin to worrying about which tubes sound better. There are better solutions these days than analog cabling, for many paradigms.
 
Last edited:

dfuller

Major Contributor
Joined
Apr 26, 2020
Messages
3,337
Likes
5,053
Balanced connections are analog, correct? Analog is inherently noisy.*

Digital is not. Digital transmission is 100% not noisy if implemented with a minimum level of competence.
USB has both power and data conductors in the same cable. That's where the issue comes from, not from the digital transmission itself.
 
Top Bottom