USB audio transmission isn’t bit true

[fas42]

4 minutes ago, beerandmusic said:

was said in jest....figured someone would get a kick out of it....you must admit other technologies seem to have come a lot further....think of where we were without the internet or mass storage not that many years ago....and we can put a man in space, or cause nuclear holocaust....but we can't figure out how to isolate noise from usb audio with anything but subjectivity....amazing.

 

 

 

 

Because audio isn't important ... close enough is good enough for most manufacturers; people still buy the product ... if people's lives depended on it, this would have been sorted decades ago ...

[beerandmusic]

16 minutes ago, fas42 said:

 

Because audio isn't important ... close enough is good enough for most manufacturers; people still buy the product ... if people's lives depended on it, this would have been sorted decades ago ...

I don't believe that...

There may be a lot more computer engineers than audio engineers so a lot more combined effort has been put into it...audio engineering is a niche comparatively speaking and most engineering is a collaborative effort over many years.  Also digital audio engineering is still in it's infancy (again, relatively speaking).

 

 

[fas42]

The infancy thing is true, in the sense that only a small number of people who do the manufacturing understand how fussy one has to be, for digital playback to really shine ... it's not a universal awareness, by a long shot.

 

The really good thing at the moment is that companies who are at the top of the quality tree have got a pretty good handle on things, finally - many of their products are hugely expensive; but it will trickle down ...

[barrows]

There is nothing fundamentally "wrong" with USB audio, period.  Bit errors do not cause subtle degradation in sound quality, period.  With a good USB source and a good USB receiver, very high levels of audio (DSD 512 for example) can be transmitted without any significant errors.  Perhaps, maybe there might be a single error every once in awhile (as in once every few hours): there is no way that such causes any sound quality degradation.

Just like any transport medium, the quality of the source device and the quality of the receiver device matters.

Same is true for Ethernet.

 

There is no inherent problem with USB audio period.

[Audiophile Neuroscience]

1 hour ago, barrows said:

There is nothing fundamentally "wrong" with USB audio, period.  Bit errors do not cause subtle degradation in sound quality, period.  With a good USB source and a good USB receiver, very high levels of audio (DSD 512 for example) can be transmitted without any significant errors.  Perhaps, maybe there might be a single error every once in awhile (as in once every few hours): there is no way that such causes any sound quality degradation.

Just like any transport medium, the quality of the source device and the quality of the receiver device matters.

Same is true for Ethernet.

 

There is no inherent problem with USB audio period.

 

Which is more fundamentally right for audio, USB or Ethernet ( copper or optical).

 

Which is harder to implement to make it more fundamentally right ?

[beerandmusic]

2 hours ago, barrows said:

 

Thanks....worth framing and adding to my  "Music Facts & Theory" notes...

The golden question is, what is the threshold for what defines "what a good USB source is"?  And even if the source is not "good", does the statement "Bit errors do not cause subtle degradation in sound quality"??

This of course is in contrary to what i have read before that in isosynchronous transmissions that large errors will cause dropouts but that small errors will cause subtle changes?  And what does that mean for the "lush" cable and "usb toys"?

 

 

[Superdad]

12 hours ago, Audiophile Neuroscience said:

Which is more fundamentally right for audio, USB or Ethernet ( copper or optical).

Which is harder to implement to make it more fundamentally right ?

 

Neither.  Ultimately it is all getting converted to the same thing in the DAC: I2S (or (DSD).

So the DAC input stages can be: USB>I2S, Ethernet>I2S, S/PDIF>I2S.

 

There are many variables in implementation as well as challenges. Decoding, clocking, power, and isolation can be done in a myriad of ways.

 

The reasons for ubiquity of USB has little to do with hardware and everything to do with software. Computer operating systems are happy to look at USB ports and send audio to them via OS support of the UAC2 audio standard.  But operating systems do not "see" Ethernet attached DACs as "sound cards" and thus other protocols have to get involved. That's where major market fragmentation came in and we end up with choices of DLNA/UPnP, Roon RAAT, Squeezelite, AirPlay, AES67/Ravenna, Signalyst NAA, etc.

 

Any one (or more) of the aforementioned "endpoint" protocols as well as the Ethernet transmission and MAC stack itself can be implemented into microprocessors/microcontrollers (or even FPGA/ASIC if one really wanted to start from scratch and had deep pockets--into the millions). If done well (again, clocking, power networks, isolation, etc.) there is no reason that the results (measured electrical and perceived SQ) can't equal that of USB or S/PDIF.

 

But again, one can not count out software factors--the how to have your Ethernet-input DAC "seen" as a "virtual sound card." Hence the paucity--and implementation disparity--of and between DACs with Ethernet inputs.  Below are photos of some various available pre-designed board modules that are in use by a small range of DAC manufacturers. This obviously does not include those manufacturers who "roll their own" Ethernet input solutions, which is quite doable since lots of popular small chips (including XMOS, PIC32, Atmel, etc.) already include Ethernet MAC processing cores.

 

Mconnect from Conversdigital: DLNA & Roon. Used by Ayre, PS Audio, and others.

 .

e-RED-MOD from Engineered.ch Electronics: DLNA/UPnP only.

ZMAN OEM from Merging Technologies: AES67/Ravenna only.

 

Don't let the size of these boards or the sight of several seemingly large chips scare you! In each case, aside from the main processor--which is loaded with the protocols and intellectual property code of the designers--most all the rest of what you see is flash memory to load those processors with code each time power is applied. Plus a few voltage regulators and maybe a basic clock.

 

 

[beerandmusic]

35 minutes ago, Superdad said:

 

The reasons for ubiquity of USB has little to do with hardware and everything to do with software.

 

 

 

Interesting....I have always believed that both USB and ENET are fundamentally sound, but believed that my preference in past (for critical listening) was due to first generation dacs not properly compensating for noise.  I "believe" they are getting better in that regard, and I certainly hated that DLNA seemed to have a lot more "hiccups" than USB, but have for the most part stuck with it for the past 5 or so years for critical listening, but would really love to get back to USB.  I don't know if ROON handles enet better than DNLA misgivings, but didn't want to go down that path for money reasons.  Anyway, i decided back then, that i would stick with DLNA until DAC engineers better engineered their dacs to compensate for noise on what I believe we both agree is otherwise a "fundamentally sound" technology.

 

I notice the things you mentioned that need to be compensated for (clocking, power, and isolation) are all things currently marketed in the UNISON design...perhaps DAC design is finally getting there?

 

NOTE: i sometimes mix you up and SONORE...i thnk you are REGEN, right?  I would understand if you do not want to see DAC's to get it right and may not be unbiased in answering?

[mocenigo]

On 11/27/2018 at 12:10 AM, yamamoto2002 said:

USB specification says, in the worst case 1 bit in 1000 000 000 000 bit may be corrupted.

 

 

Which means that well trained audiophiles will hear the potentiality of this error occurring as a form of stress, of nervousness in the audio reproduction, causing the image to collapse, and they will immediately declare that audio chain as pure crap...

 

Provided they have noticed that the USB cable costs less than 2K EUR.

[StephenJK]

Why do we keep revisiting this stuff?  Is there nothing else to talk about?  Is this the Audio Asylum, where we we're stuck in Groundhog Day talking about how to adjust antiskate?

 

Except now it's "How can a digital copy possible be true?

 

There has to be a country song about that, there just has to be......

[Superdad]

19 minutes ago, SJK said:

There has to be a country song about that, there just has to be......

 

There ought to be, but then again I don't know a lot of audiophiles who listen to much country music. 

[Where are you Garth?!]

[Audiophile Neuroscience]

6 hours ago, Superdad said:

 

Neither.  Ultimately it is all getting converted to the same thing in the DAC: I2S (or (DSD).

So the DAC input stages can be: USB>I2S, Ethernet>I2S, S/PDIF>I2S.

 

There are many variables in implementation as well as challenges. Decoding, clocking, power, and isolation can be done in a myriad of ways.

 

The reasons for ubiquity of USB has little to do with hardware and everything to do with software. Computer operating systems are happy to look at USB ports and send audio to them via OS support of the UAC2 audio standard.  But operating systems do not "see" Ethernet attached DACs as "sound cards" and thus other protocols have to get involved. That's where major market fragmentation came in and we end up with choices of DLNA/UPnP, Roon RAAT, Squeezelite, AirPlay, AES67/Ravenna, Signalyst NAA, etc.

 

Any one (or more) of the aforementioned "endpoint" protocols as well as the Ethernet transmission and MAC stack itself can be implemented into microprocessors/microcontrollers (or even FPGA/ASIC if one really wanted to start from scratch and had deep pockets--into the millions). If done well (again, clocking, power networks, isolation, etc.) there is no reason that the results (measured electrical and perceived SQ) can't equal that of USB or S/PDIF.

 

But again, one can not count out software factors--the how to have your Ethernet-input DAC "seen" as a "virtual sound card." Hence the paucity--and implementation disparity--of and between DACs with Ethernet inputs.  Below are photos of some various available pre-designed board modules that are in use by a small range of DAC manufacturers. This obviously does not include those manufacturers who "roll their own" Ethernet input solutions, which is quite doable since lots of popular small chips (including XMOS, PIC32, Atmel, etc.) already include Ethernet MAC processing cores.

 

Mconnect from Conversdigital: DLNA & Roon. Used by Ayre, PS Audio, and others.

 .

e-RED-MOD from Engineered.ch Electronics: DLNA/UPnP only.

ZMAN OEM from Merging Technologies: AES67/Ravenna only.

 

Don't let the size of these boards or the sight of several seemingly large chips scare you! In each case, aside from the main processor--which is loaded with the protocols and intellectual property code of the designers--most all the rest of what you see is flash memory to load those processors with code each time power is applied. Plus a few voltage regulators and maybe a basic clock.

 

 

 

Thanks so much for this excellent explanation Alex !

 

Yes I wondered about what happened with things like drivers for Ethernet based audio and the OS seeing the Ethernet connection as a soundcard. It also clarified some of my confusion regarding "endpoints" and their purpose in the scheme of things. Thank you.

 

I hope I am not putting you on the spot but I was curious about your chosen setup that I saw when lurking on one of the more technical threads. Basically, AFAICT, you have chosen a copper Cat6 Ethernet solution to a Nuc/ NAA endpoint (via Er) and then IsoRegen to usb cable. Was there a reason you did not choose USB > USB cable> IsoRegen? Also, having chosen an Ethernet leg, why not optical?

 

I have been researching an 'optical solution' and of course there are many opinions regarding the superior galvanic isolation and high jitter rejection. When reading through The Taiko Extreme threads here and at WBF, listening impressions also varied regarding optical. initially improvements were touted but not always.....as usual a case of try it for yourself YMMV. The criticisms of optical reminded me of my experience of old with optical when I compared ST connector/ glass fiber with AES. Optical always sounded thinner and leaner, less natural.

 

Cheers

David

 

[Teresa]

3 minutes ago, Superdad said:

 

There ought to be, but then again I don't know a lot of audiophiles who listen to much country music. 

[Where are you Garth?!]

 

I like Kenny Rogers, Robert Earl Keen and others. However, most of the country music I like is early traditional country music, such as Sons of the Pioneers. I also like bluegrass music such as Alison Krauss, Nickel Creek, Joe Weed and others. And dawg music from David Grisman, especially those made with the late Grateful Dead guitarist Jerry Garcia.

 

Not sure why more audiophiles haven't discovered country, bluegrass and dawg music since most are very realistic recordings. Those acoustic guitars, mandolins and fiddles sound quite wonderful on a good sound system IMHO.

[barrows]

58 minutes ago, Superdad said:

 

There ought to be, but then again I don't know a lot of audiophiles who listen to much country music. 

[Where are you Garth?!]

Wait a minute Alex?  Really, I mean there's the typical country pop stuff, and then there is country, like Patsy Cline, some of Shelby Lynne's stuff, and that of her sister, Johnny Cash might be considered country, right?  I know there might be a it of a distinction between what we might call "Americana" these days, but a lot of it is really country, and then there is the cross over with bluegrass folks like Allison Krauss, etc...

[Audiophile Neuroscience]

I like both kinds of music...

 

[beerandmusic]

51 minutes ago, barrows said:

Wait a minute Alex?  Really, I mean there's the typical country pop stuff, and then there is country, like Patsy Cline, some of Shelby Lynne's stuff, and that of her sister, Johnny Cash might be considered country, right?  I know there might be a it of a distinction between what we might call "Americana" these days, but a lot of it is really country, and then there is the cross over with bluegrass folks like Allison Krauss, etc...

 

i don't like much country, but there are probably 100 or so exceptions....

 

 

[vmartell22]

21 hours ago, beerandmusic said:

 

i don't like much country, but there are probably 100 or so exceptions....

 

 

 

Love how Trent Reznor humbly declared that the song was never his... once Cash started singing, realized that it was the man in black's all along...

 

v

[ssh]

[OAudio]

On 6/9/2020 at 6:49 AM, beerandmusic said:

 

audio engineering cannot be that difficult...

 

you are sending

A) 5v reference voltage

B) digital music (does everybody agree or NOT, that the receiving dac gets this data accurately)

C) noise

 

In the past, everyone that spoke up would say that the digital music received is so accurate that it may drop one bit every 24hrs or something along those lines.

 

My thinking, is that if the dac does indeed recieve it's data with near perfect accuracy, and you isolate the 5v reference voltage and "reclock" the "perfect bits" with your own engineering circuit, then it should NOT make any difference how quiet ones pc is or not??

 

If that is not logical thinking, I just want to know the reasoning.

 

My thinking is that, if that is not logical thinking, then the only thing that would make sense is that

A) the bits are not received nearly as perfect as people suggest

B) audio engineers still have not figured out a "good design"

C) noise is not only not being isolated properly, but it corrupts the music data before input to the conversion circuitry.

D) noise is not being isolated properly and it affects the conversion circuitry

E) subjective minds inaccurately believe the two systems sound differently.

 

 

 

 

Beerandmusic, this is a very interesting thread.

 

I am relatively new to the forum so its just caught my eye. It's very interesting that asynchronous transfer errors are not resent meaning that it is possible that data can arrive at the DAC uncorrected.

 

I also think that the vanishingly low errors rates that are quoted earlier in the thread may not truly reflect how a real music system's USB interface performs. The low error referenced by the USB standard assumes that transmission conditions meet the USB standards signal eye specifications all of the time.  But, consider that successful transmission of music data relies on maintaining USB transmission eye quality on the 480mhz carrier and with differential D+ D- signal amplitude of ~1v. When you look into the logic detection thresholds they are as low as 0.4v and this at 480Mhz down say a 0.5-1.5m USB cable. The transmitting (PC) and receiving (DAC) systems are driven by entirely different power supplies (differential power noise between these two platforms impacts threshold margin, already say just 0.4v, at the receiver).

 

I use USB between PC and DAC with D+ D- USB isolation at the DAC. I also (intentionally) have the ability to match the clock rates between the USB transmission (PC) and USB receiver (DAC). I think the set up of the USB link is pretty good, certainly sound quality is very good but I agree with your list of possible problems (above). So I have taken the step of designing this in line USB board specifically to look at some of the factors in play. If it works as intended it will allow USB signal, power at transmitter / receiver and shield continuity conditions to varied whilst looking for subjective sound quality change. I don't know if this approach is going to generate useful outcomes but faced with so many possible factors, a practical approach might narrow things down. 

 

The PCB is produced and on its way to me and I have to populate and test, but fingers crossed that the effort produces useful results.

 

OAudio

 

 

[sandyk]

21 minutes ago, OAudio said:

If it works as intended it will allow USB signal, power at transmitter / receiver and shield continuity conditions to varied whilst looking for subjective sound quality change.

 

 Just remember that most ultra low noise voltage regulators have a much lower output impedance at >100kHz due to the types and smaller values of filter capacitors normally used. Ideally , the PSU impedance should be as low and flat right across the whole range to 1MHZ as possible. Try also using a much higher value electro in parallel at their output ,perhaps  even  an Audiophile grade type such as Elna Gold .

USB is also sensitive to the capacitance to mains earth in a LInear PSU with an R-Core sounding better than a Toroidal transformer, and with a split bobbin type sounding even better. Even earthing the supplied screen wire of an R-Core transformer can cause a small audible degradation. I verified this by fitting a toggle switch at the rear of a <4uV noise PSU to quickly switch between both.

[OAudio]

24 minutes ago, sandyk said:

 

 Just remember that most ultra low noise voltage regulators have a much lower output impedance at >100kHz due to the types and smaller values of filter capacitors normally used. Ideally , the PSU impedance should be as low and flat right across the whole range to 1MHZ as possible. Try also using a much higher value electro in parallel at their output ,perhaps  even  an Audiophile grade type such as Elna Gold .

USB is also sensitive to the capacitance to mains earth in a LInear PSU with an R-Core sounding better than a Toroidal transformer, and with a split bobbin type sounding even better. Even earthing the supplied screen wire of an R-Core transformer can cause a small audible degradation. I verified this by fitting a toggle switch at the rear of a <4uV noise PSU to quickly switch between both.

Hi Sandyk,

 

Thanks for the thoughts.

 

My rendering software can't show the second daughter pcb. The second board has a 4uv low noise regulator and option are built in for where it is powered from. I can use this / the PC supply in different combinations for transmitter and receiver bus power.  

 

Transformer ground current is a tricky one,  its influance is going to show up twice if Im not mistaken from the PC supply and the DAC. 

 

Theres quite some thought behind the approach in both these areas but perhaps best to get the board built and tested then go into some more details.

 

 

[sandyk]

18 minutes ago, OAudio said:

Hi Sandyk,

 

Thanks for the thoughts.

 

My rendering software can't show the second daughter pcb. The second board has a 4uv low noise regulator and option are built in for where it is powered from. I can use this / the PC supply in different combinations for transmitter and receiver bus power.  

 

Transformer ground current is a tricky one,  its influance is going to show up twice if Im not mistaken from the PC supply and the DAC. 

 

Theres quite some thought behind the approach in both these areas but perhaps best to get the board built and tested then go into some more details.

 

 

 

Given that many USB devices, including USB memory, have an internal connection between 0 volts and shield, have you also made provision for switching/isolating the shield lead, as this may result in an inadvertent earth loop .

[OAudio]

21 minutes ago, sandyk said:

 

Given that many USB devices, including USB memory, have an internal connection between 0 volts and shield, have you also made provision for switching/isolating the shield lead, as this may result in an inadvertent earth loop .

We are on the same page :-)

The board is set up to be able to be switched between these options and to examine some ideas to manage transmitter / reciver differential supply noise. I have used this approch in other projects but lots and lots of unknowns with USB transmission so practical testing seems the best way to go.

 

 

[mocenigo]

On 6/10/2020 at 12:47 AM, fas42 said:

 

Because audio isn't important ... close enough is good enough for most manufacturers; people still buy the product ... if people's lives depended on it, this would have been sorted decades ago ...

 

Come on, one bit error in 10^12 is waaaaay more than needed. USB has perfect audio transfer provided the sink buffers and reclocks it and the cable has decent shielding (also, the sink should have good galvanic isolation). There is no need for more engineering there.

[OAudio]

2 hours ago, mocenigo said:

 

Come on, one bit error in 10^12 is waaaaay more than needed. USB has perfect audio transfer provided the sink buffers and reclocks it and the cable has decent shielding (also, the sink should have good galvanic isolation). There is no need for more engineering there.

 

Mocenigo hi,

 

Bit error of 10^12 would do the job for sure...."if" you can get to this performance in a real life implementation. Personally I think there is a lot that needs to be taken care of in PC to DAC system set-up approach this.

 

As I mention above the USB transmission eye detection margin is quite small for the receiver (DAC) at ~ 0.4v. Differential noise between the power supplies of the USB link's transmitter (eg the PC, packed with noisy buck convectors, high transient loads and even an SMPS if your unlucky) and the receiver (the DAC probably on a low noise linear supply) really have the potential to impact eye detection margins.

 

Although the USB cable is shielded, there is also generally a GND loop between the PC and DAC via mains safety earth connections. The PC switching supply noise and ground leakage currents can pollute the safety earth and appear at the USB receiver circuit in the DAC (even though there "should" be a good ground reference established by the USB lead's shield).

 

Then there is EMI coupling into the D+ D- pairs as they traverse the motherboard, if motherboard is used, or from a PCIE card if used. EMI can also be coupled into the USB cable. I have much frustrating experience of "good" quality coaxial cable being insufficiently shielded for low jitter HF signalling. Interesting here is that there are so many reports in forums about improvements in sound from multi shielded USB cables, something I use here too.

 

Final area I think is very important is transmission timing both phase noise and clock speed. The differential noise above may or may not be enough to cause data errors due to eye detection errors. Even if errors are not being caused by detection errors, the differential power noise in the transmitter's & receiver's supplies will cause threshold detection jitter in the USB data stream and this does matter to sound quality (although I would agree this is not data error). I mentioned in my earlier post above I have developed the ability to accurately set the relative frequency of the individual USB clock domains governing the transmitter and receiver. I have been working on this stuff for many years, and know that as little as an 0.000005% difference in the speed of the USB transmitter and receives clock domains can be heard. USB timing really matters if you are aiming for truly high end sound quality.

 

The point of the diagrams below is not to highlight that using a 3 or 5m cable could be a bad move (most people just don't go that long for audio :-) ), rather my point is something as simple as the cable length alone can really degrade the eye detection margins. The issues listed above I think have far greater potential to harm eye margin performance than these example cable lengths.  

 

sample USB transmission eye 9 inch cable

 

eye 3m cable

 

eye 5m cable

 

I don't have USB test equipment (way too expensive) but I have been modifying USB interfaces and audio servers at board level for > 14 years. I just can't say beyond doubt that the above issues cause actual errors but I have come across lots of evidence that the areas above really matter for quality. 

 

I happened to be working up that board I posted about above prior to seeing this thread.I though it would be fun to post about this. It is able to do much more than condition power and make or break shield links etc. I felt its worth the time to see if anything useful can be pinned down.

 

Best regards,

 

OAudio.