USB vs Coax - my (interesting) impressions

[doomigee2]

Hi all

 

Have had a very interesting couple of days and felt it necessary to share my impressions to help others answer the same questions I have been pondering the last little while. 

 

So, here is the situation. I recently changed from my Bluesound Node 2 to an Auralic Aries LE. The Bluesound has only toslink or coax output, of which the coax is vastly superior IME. The Aries, however, has the added option of USB, which tempted my curiosity and formed the basis of this thread. 

 

Before going any further, its important to list my gear for context:

 

Source: Auralic Aries LE with Geiseler LPSU

DAC: Chord 2qute with Geisler LPSU

Pre: Marantz SR7008 AVR (I know, I know)

AMP: Anthem MCA20

Speakers: Focal Scala Utopia

Cables: Speaker - Analysis Plus Black Mesh Oval 9

             Interconnects - Analysis Plus Solo Crystal

             Coax - Analysis Plus Digital Crystal

             Power - Isotek Premier

             USB - bog standard Type A to Type B cable

 

 

OK, so off I go setting up the Aries, using coax initially, and my initial impression is that there is an improvement over the Node 2. Now this isnt the reason for this thread, but worth pointing out that yes, I recognise the difference between Node 2 (with coax) and Aries (with coax). I listen to it for about a week or so using only coax to get used to the sound. Its nice

 

Then on Friday, I plug in the standard (crappy) USB cable that was supplied with the Chord DAC. Its about 2m long, super flexible, cheap looking, cheap feeling and resembles what you would typically see connecting a PC to a printer. Needless to say, I wasn't expecting much. 

 

But I was wrong, after installing the USB and switching over, I was pretty shocked at what I heard. It was nice, actually it was really nice. I could instantly hear a difference, at this point it was only a difference not so much an improvement, until further listening. 

 

Here is what I noticed:

 

Bass - deeper and more pronounced with USB, but a little better controlled with the Coax

Stage - much wider with the USB, height same with either, depth a touch better with Coax but overall USB is the winner

Transparency - better with USB, certainly more "holographic" 

Image - good with both, but the Coax is a little more focussed here

Highs - MUCH better with the USB - no sibilance (Coax has a touch of sibilance on female vocals) and a much greater sense of "air" 

Decay - better with USB - in particular cymbal crashes seem to just decay longer and with more sparkle. 

Attack - better with Coax, in comparison the USB does sound a little flat. The drum stick strike is much better on Coax

 

Now there are lots of other things missing, but the above are the biggest differences that I noticed. Overall, the USB is much more engaging and enjoyable, although I do miss the attack and better bass control of the Coax. 

 

So the summary of the above - I am absolutely shocked that I am even able to compare a $2 USB cable with a $500 Coax cable. Please note that I am definitely a cable believer and am not insinuating that there is no difference in cables, simply that the interface of USB is so superior that even a standard cable is able to compete (and best) a top end Coax cable. This is a big surprise for me and makes me really excited for what I will hear when I receive my Curious Cable USB......

[MattyW]

Hi all
 
Have had a very interesting couple of days and felt it necessary to share my impressions to help others answer the same questions I have been pondering the last little while. 
 
So, here is the situation. I recently changed from my Bluesound Node 2 to an Auralic Aries LE. The Bluesound has only toslink or coax output, of which the coax is vastly superior IME. The Aries, however, has the added option of USB, which tempted my curiosity and formed the basis of this thread. 
 
Before going any further, its important to list my gear for context:
 
Source: Auralic Aries LE with Geiseler LPSU
DAC: Chord 2qute with Geisler LPSU
Pre: Marantz SR7008 AVR (I know, I know)
AMP: Anthem MCA20
Speakers: Focal Scala Utopia
Cables: Speaker - Analysis Plus Black Mesh Oval 9
             Interconnects - Analysis Plus Solo Crystal
             Coax - Analysis Plus Digital Crystal
             Power - Isotek Premier
             USB - bog standard Type A to Type B cable
 
 
OK, so off I go setting up the Aries, using coax initially, and my initial impression is that there is an improvement over the Node 2. Now this isnt the reason for this thread, but worth pointing out that yes, I recognise the difference between Node 2 (with coax) and Aries (with coax). I listen to it for about a week or so using only coax to get used to the sound. Its nice
 
Then on Friday, I plug in the standard (crappy) USB cable that was supplied with the Chord DAC. Its about 2m long, super flexible, cheap looking, cheap feeling and resembles what you would typically see connecting a PC to a printer. Needless to say, I wasn't expecting much. 
 
But I was wrong, after installing the USB and switching over, I was pretty shocked at what I heard. It was nice, actually it was really nice. I could instantly hear a difference, at this point it was only a difference not so much an improvement, until further listening. 
 
Here is what I noticed:
 
Bass - deeper and more pronounced with USB, but a little better controlled with the Coax
Stage - much wider with the USB, height same with either, depth a touch better with Coax but overall USB is the winner
Transparency - better with USB, certainly more "holographic" 
Image - good with both, but the Coax is a little more focussed here
Highs - MUCH better with the USB - no sibilance (Coax has a touch of sibilance on female vocals) and a much greater sense of "air" 
Decay - better with USB - in particular cymbal crashes seem to just decay longer and with more sparkle. 
Attack - better with Coax, in comparison the USB does sound a little flat. The drum stick strike is much better on Coax
 
Now there are lots of other things missing, but the above are the biggest differences that I noticed. Overall, the USB is much more engaging and enjoyable, although I do miss the attack and better bass control of the Coax. 
 
So the summary of the above - I am absolutely shocked that I am even able to compare a $2 USB cable with a $500 Coax cable. Please note that I am definitely a cable believer and am not insinuating that there is no difference in cables, simply that the interface of USB is so superior that even a standard cable is able to compete (and best) a top end Coax cable. This is a big surprise for me and makes me really excited for what I will hear when I receive my Curious Cable USB......

Not really a surprise for me. It all comes down to the USB implementation. I find that on my media PC with JRiver the best chain is USB port > Gustard U10 interface > XLR cable to Gustard DAC-X12. Provides fantastic sound.

I also prefer using a separate interface as the sound seems cleaner and more relaxed, likewise I can switch DAC's without needing to change anything on the PC.

Anyways, USB is certainly the best interface that I've come across from a PC. For my desktop computer both at home and work I run a little S.M.S.L M3 DAC off USB to good effect.

It really comes down to the individual piece of gear as to whether USB, Coax or Toslink is better. Not all implementations are created equal.

Sent from my Redmi Pro using Tapatalk

[powerav]

Great topic, I have been shot down on here from experienced members that for some reason want to diss USB. 

[Qwertyqaz]

Probably worth a read, Iam USB in both my Rigs many because of flexibility and I like DSD.

 

[Neo]

Galvanic isolation on the chord 2qute on USB implementation and the crystal clock at the source on Le as the output

A match made in heaven 

Neo

[Neo]

23 hours ago, powerav said:

Great topic, I have been shot down on here from experienced members that for some reason want to diss USB. 

Properly implemented it's a great way to transmit audio, and yes different USB cables do make a difference 

[Dcesilo80]

I have never noticed these things before.  return man 3 unblocked Thanks for telling me these.

Edited September 6, 2017 by Dcesilo80

[soundfan]

I'm another fan of USB. Use it from my MacBook running Audirvana 3 and Tidal premium into my Denafrips Ares DAC, which is directly connected to my 2 mono blocks. Sounds excellent, and betters a coaxial connection from my CD player straight to Dac.

I use a iFi Gemini USB cable. Tried a USB Regen and LPS in the chain, but they made no audible difference. Sold them shortly after.

I guess it may come down to how good the USB implementation in DAC is?

[kdoot]

The #1 difference between USB and coax is that USB makes it possible for the DAC to be the master clock, via the asynchronous USB Audio protocol. If everything else is done well enough, that one fact means USB will win just about any comparison with SPDIF.

 

The problem with SPDIF is that there's no explicit clock signal. The DAC has to infer the clock from the edges of the data signal – and that makes jitter really hard to control. 

 

There are challenges with USB too, especially with electrical noise from the source. But that can be overcome. 

[Bilbo]

I'm waiting to hear your impressions when you plug in the Curios cable (after about 50hrs run in at least).

If it is like many others you are in for a very pleasant surprise.

[sloper]

Goodaye all

 

Shouldnt be a great difference in the cables as its all 0's and 1's and there's a clock signal embeddd in all.

USB and coax can pick up some intererence sometimes.

Optical doesnt.

 

regards Bruce

[Guest Eggcup The Daft]

http://www.realhd-audio.com/?p=5971

This link is to a blog entry from Mark Waldrep entitled "Digital errors don't affect fidelity". In turn, it links to a piece by Gordon Rankin on DAR which has, shall we say, a different point of view.

 

I'd like to suggest a subtly different point that refers to the comparison given here. It's this - for cables or interfaces, of whatever digital type, to give a consistent timbral, soundstaging or whatever audio difference, through digital errors - widely differing musical content, and therefore widely differing patterns of 0s and 1s, have to be affected in the exact - and rather gross - manner to produce that result. I'm not aware that that can happen.

 

Of course, there are other ways in which a digital cable could produce consistent results as in the first post here. But to get those described and consistent results, I presume that something would have to affect the analogue side of the DAC or another part of the post-DAC chain.

 

Jitter is a bit of a mystery to me, but I understand that if it becomes audible, it doesn't manifest as a change to a "deeper soundstage" or similar.

 

Does anyone have evidence to prove this suggestion to be wrong? If it is about affecting the analogue stages, are there engineering solutions, and are some DACs less affected by cable changes than others? And can a cable effectively mimic what would have to be a fairly complex digital filter to get these consistent changes?

 

I do know that obscene amounts of noise on its Ethernet input can affect sound from my Oppo 105, from having a switch fail.

[sloper]

Goodaye all

 

Your probaly on the right track, the cable affecting the analouge output of the DAC, theres ground on USB and Coax and a 5v source on the USB.

The ground will be common on both devices and the rest of your house, you may have a real small voltage difference.

And thats a avenue for electrical interference.

 

As for digital error's its not as easy as you believe, try doing a ping test and see how much data you loose, it will proably be zero contacting a Servor halfway across the world.

Error correction could be employed on a DAC if people were convinced there is a problem, but then were have now introduced software another layer of complexity.

 

Thats why l prefer optical inputs.

 

regards Bruce

Edited May 8, 2017 by sloper

[kdoot]

Guys, I have been having this same conversation for years now, which means I'm gonna make this post and let you find your own way after that. I'll answer questions but not have a debate... I'm too tired of debating 

 

The bits are not the problem. That's not why different digital cables/interfaces/etc sound better or worse. It's all about noise and jitter carried in those connections or generated by the processing of the signal as it's received at the DAC end. Noise and jitter, if present in the actual DAC component during D-to-A conversion, degrades the clarity and integrity of the sound. It's not frequency-dependent, it's an across-the-board not-as-right-ness.

 

Sloper, you're right that Toslink has zero electrical noise. That is a big benefit. BUT Toslink is among the worst possible options for jitter because not only is there no explicit clock signal but the signal edges from which the clock must be derived are not sharp and clear. If we had a scheme where the DAC transmitted a master clock back for the source to follow it'd be a different story, but that's not what we have. My DAC has the best anti-jitter technology I've ever come across and I've been singing the praises of its Toslink input for a long time... but the Curious USB cable has recently made me eat crow. Whatever it does to reduce electrical noise, works. And the clock is under the DAC's control. Game over for Toslink in this system. 

 

I used to use a Metrum Octave NOS DAC, which supports up to 192kHz sample rates and has absolutely zero protection against jitter. It is a brilliant way to demonstrate the effects of upsampling, different filter schemes, and the sound of high jitter vs low jitter transports. I relied on an Audiophilleo USB-to-Coax converter to get me a low noise, low jitter digital signal and it was night and day difference when compared to Toslink direct from my computer or from any common inexpensive streaming device. 

 

The bits were the same. The song was the same. The frequency response was the same. One sound was very enjoyable, the other really not very. 

 

I really wish it weren't so, but digital transports and cables matter a lot. If you care about sound quality, there's nothing for it but to get out and actually listen and compare. Just don't forget to stop and enjoy the music every now and then.  

[andyr]

1 hour ago, kdoot said:

 

I really wish it weren't so, but digital transports and cables matter a lot. If you care about sound quality, there's nothing for it but to get out and actually listen and compare. Just don't forget to stop and enjoy the music every now and then.  

 

 

Listening to the music when you' re listening to digital - what a lot of tosh!   If you want to listen to the music ... you listen to vinyl. 

 

Seriously, though, thank you for the explanation of the problem with Toslink.

 

Andy

 

[Guest m_james]

On 5/3/2017 at 9:10 PM, kdoot said:

The #1 difference between USB and coax is that USB makes it possible for the DAC to be the master clock, via the asynchronous USB Audio protocol. If everything else is done well enough, that one fact means USB will win just about any comparison with SPDIF.

 

The problem with SPDIF is that there's no explicit clock signal. The DAC has to infer the clock from the edges of the data signal – and that makes jitter really hard to control. 

 

There are challenges with USB too, especially with electrical noise from the source. But that can be overcome. 

 

Question - if my DAC only  has a SPDIF input, would it be beneficial to get a usb to spdif cable from my PC rather than using a RCA cable?

[kdoot]

57 minutes ago, m_james said:

 

Question - if my DAC only  has a SPDIF input, would it be beneficial to get a usb to spdif cable from my PC rather than using a RCA cable?

 

Great question. 

 

The answer is almost always yes, provided that the interface you're buying is lower-noise and lower-jitter than the one you already have. Since the digital outputs on most PCs are awful, the improvement can be dramatic.

 

I'm a huge fan of the Audiophilleo for that job, but there are dozens of good options these days. You'll need to do some research.

[Guest Eggcup The Daft]

1 hour ago, andyr said:

 

Listening to the music when you' re listening to digital - what a lot of tosh!   If you want to listen to the music ... you listen to vinyl. 

 

Seriously, though, thank you for the explanation of the problem with Toslink.

 

Andy

 

If you really want to listen to the music... there are these things called musicians. They appear at things called concerts (gigs, recitals, or whatever they call them with the music you like)...

 

[Addicted to music]

7 minutes ago, Eggcup The Daft said:

If you really want to listen to the music... there are these things called musicians. They appear at things called concerts (gigs, recitals, or whatever they call them with the music you like)...

 

 

Who would have though, I learn something new everyday.....

[Jventer]

4 hours ago, m_james said:

 

Question - if my DAC only  has a SPDIF input, would it be beneficial to get a usb to spdif cable from my PC rather than using a RCA cable?

I am not aware of cables that goes from usb to spdif.

You need a converter as well, so usb to converter and spdif from converter. 

Many converters available, e.g. Audiophilleo.

 

[Guest m_james]

8 hours ago, kdoot said:

 

Great question. 

 

The answer is almost always yes, provided that the interface you're buying is lower-noise and lower-jitter than the one you already have. Since the digital outputs on most PCs are awful, the improvement can be dramatic.

 

I'm a huge fan of the Audiophilleo for that job, but there are dozens of good options these days. You'll need to do some research.

 

Geez, $999 for a converter? I think I'd need government funding to carry out that kind of research..

 

Thanks for the info though.

[kdoot]

12 minutes ago, m_james said:

 

Geez, $999 for a converter? I think I'd need government funding to carry out that kind of research..

 

Thanks for the info though.

 

Heh. OK. Price does matter, I understand that. 

 

I went looking for options for you but it does rather seem like the market has dried up as more DACs offer decent internal USB support. Maybe you should try posting in the Wanted-to-buy section of the classifieds and see what pops up?

[Guest Eggcup The Daft]

Matrix X-SPDIF $299 from Noisy Motel. No idea if it's any good. There's a cheaper one that only goes to 96/24 if you don't use higher rates.

The Musical Fidelity V90 was fairly cheap and recommended in the UK for a while, but is discontinued. You might find a used one if you look.

[davewantsmoore]

On 08/05/2017 at 7:11 PM, Eggcup The Daft said:

Does anyone have evidence to prove this suggestion to be wrong?

 

Yes.    You'll need to understand jitter (and digital transmission in general) a lot better before any of it will make any sense.

 

 

Very broadly.

 

Jitter (timing errors) in the digital signal, causes problems in the digital to analogue conversion process .... and the result is distortion in the analogue output.     The distortion produced is (largely) independent of the pattern of 1 and 0.

 

The design of some devices vs others, makes the more resistant to the effects of jitter in the digital data.

 

23 hours ago, sloper said:

Error correction could be employed on a DAC if people were convinced there is a problem, but then were have now introduced software another layer of complexity.

 

Getting the correct 1s and 0s to turn up at the destination is almost never a problem.    So, not only is error correction difficult - but it's not needed.

 

The the right 101100111100011 turns up .... but issue is the time delay in between each is inconsistent (called jitter).    This causes the analogue output of a DAC to be distorted.

[davewantsmoore]

edit: inappropriate commentary.

Edited May 9, 2017 by davewantsmoore