Is USB straight from a Mac computer to a DAC really that bad?

[beerandmusic]

On 6/1/2020 at 10:02 AM, davide256 said:

 no DAC can fix what the source has damaged at USB output

 

 

 

I have heard this before, but I am not sure I believe it.  Most everyone agrees that the dac gets the digital bits perfectly, it's just that there is noise.  So if the DAC implementation is well, and it can get the bits accurately, why can't it regenerate the bits "that it gets perfectly", and then provide it's own 5v.  In theory, this should be a good implementation, and also what the unsion usb markets?  I would think that every "good" dac should do this by this generation of dacs?

[beerandmusic]

1 hour ago, davide256 said:

time flows backwards for no one. While DAC's matter they can't make a bad source good.... if its error-ed in hand off to USB out, its too late.

Cheaper digital sources remind me very much of cheaper vinyl TT's, compressed dynamics and blurry for background details.

 

You may not be able to make a bad source good but let's think about what we are saying.

A bad source can be a bad recording, but that is not what i am speaking about.

 

Let's start with the bits on the media are perfect as proven by checksum.

 

Exiting the source, you have the perfect bits and call it noise.

Assuming music = perfect bits + noise and nothing else.

 

We know the dac gets the perfect bits.

 

The dac conversion of course will be different based on the dac engineering, but right now, all we are trying to deal with is the input.

We already know they get the bits perfectly.

 

Are we together so far?

 

[beerandmusic]

2 hours ago, sandyk said:

 That is definitely not my experience, although noise can play a part, the Data is normally received without error.

 The PSU itself also determines how the Audio will sound. Use too many Low ESR filter capacitors in the PSU for example, especially the PSU area powering the OS SSD and Music storage SSD, and the sound can become too bright/HF detailed.

 The same even applies to the external PSU powering a USB Regen, which in itself increases HF detail slightly due to the necessarily small value filter capacitors it uses with it's internal voltage regulators. This is even when powered directly by a 12V Li Ion battery and saving to USB memory . Many members also obtain added HF detail when using the ultra low noise LT3045 voltage regulators to power various items, and for the same reason, which is to increase HF detail and soundstage.

 Some even use a couple of these regulators in tandem for this reason.

(Check other areas of the forum such as Music Servers )

 

The power is just another aspect of noise.

For purpose of this conversation, lets just deal with 2 parameters,  data and noise, and the data is received without error...so all we are trying to resolve for is "noise".

[beerandmusic]

2 hours ago, sandyk said:

 That is definitely not my experience, although noise can play a part, the Data is normally received without error.

 The PSU itself also determines how the Audio will sound. Use too many Low ESR filter capacitors in the PSU for example, especially the PSU area powering the OS SSD and Music storage SSD, and the sound can become too bright/HF detailed.

 The same even applies to the external PSU powering a USB Regen, which in itself increases HF detail slightly due to the necessarily small value filter capacitors it uses with it's internal voltage regulators. This is even when powered directly by a 12V Li Ion battery and saving to USB memory . Many members also obtain added HF detail when using the ultra low noise LT3045 voltage regulators to power various items, and for the same reason, which is to increase HF detail and soundstage.

 Some even use a couple of these regulators in tandem for this reason.

(Check other areas of the forum such as Music Servers )

 

Also the more usb toys, the more chance of introducing noise and needing additional "perfect" power supplies.

 

The objective is to get the bits perfectly to the dac (which people agree happens), and then to isolate noise. 

A well designed DAC should be able to do that.

It has the perfect bits, so it is just a matter of introducing the perfect bits for processing and isolating the noise.

 

Edit --- i thought we were in the thread i created about usb isolation...not realizing this is the thread about macs....disregard.

 

[beerandmusic]

3 hours ago, sandyk said:

 That is definitely not my experience, although noise can play a part, the Data is normally received without error.

 The PSU itself also determines how the Audio will sound. Use too many Low ESR filter capacitors in the PSU for example, especially the PSU area powering the OS SSD and Music storage SSD, and the sound can become too bright/HF detailed.

 The same even applies to the external PSU powering a USB Regen, which in itself increases HF detail slightly due to the necessarily small value filter capacitors it uses with it's internal voltage regulators. This is even when powered directly by a 12V Li Ion battery and saving to USB memory . Many members also obtain added HF detail when using the ultra low noise LT3045 voltage regulators to power various items, and for the same reason, which is to increase HF detail and soundstage.

 Some even use a couple of these regulators in tandem for this reason.

(Check other areas of the forum such as Music Servers )

 

so you disagree, so there is no reason to discuss this topic further with you...

we need to have an initial agreement before we can discuss or look further....

 

The music is stored as a "flat" file on a storage device, can we start there?  If not then there is no purpose in going on further... (I am not trying to compare one file to another...that is not my purpose).

 

My purpose is to discuss the logical reason why engineering cannot seem to get one flat music file to the input of the d->a conversion without compromise without the need to daisy chain gadgets and additional power supplies.  This should NOT be rocket science.

 

The unison usb seems to market "at least the theory".

I tried to go here about 5 years ago but fell on deaf ears....that a well designed dac should be able to do these things without relying on "usb toys".

 

 

[beerandmusic]

On 5/29/2020 at 2:08 PM, audiophile911 said:

Is USB straight from a computer to a DAC really a bad or should everyone always strive to isolate the computer's USB output from the audio stream???  I connect my Chord Quest directly to my Mac Mini; which is dedicated to only running ROON Core; with an AudioQuest Diamond USB and I think it sounds great.  But I according to manufactures of network streamers, eliminating the computer (or using an expensive audio optimized PC like an Innuos) will always sound significantly better.  I've also read that this is not necessarily the case and it really depends on how usb is implemented in the source and the DAC?  Specifically, I heard from Rob Watts of Chord explain that Chord DAC's are optimized for USB direct input.  So, I'm trying to decide if I need to try something like a SOTM SMS-200 Ultra or a Sonore UltraRendu but I'm hesitant to go to the expense and hassle of more boxes.   I recently read this update on this $150K system: https://www.soundstagehifi.com/index.php/opinion/1392-after-25-years-is-this-the-worlds-best-audio-system  Specifically:

"Some Facebook readers criticized me for not using an audiophile-grade music server or USB link. I responded that if anyone can show me a music server or USB link that actually sounds better that what I have in terms of resolution, tonality, soundstaging, imaging, whatever -- I’m all ears. But so far, I’ve heard nothing that has proven itself better-sounding or more versatile -- my computer plays any digital music format and file type from streaming services and my local music drive, and my USB link, with its lengthy length, transfers the bits just fine.

The reason I can get away with using a laptop has to do with the next component in the signal chain: the EMM Labs DA2 Reference DAC ($25,000). Designed by Ed Meitner, who’s been creating digital-audio products since the 1970s, the DA2 Reference seems immune to swaps of USB links, as well as differences in source components."

 

Am I missing something??

 

In my thinking it should be possible with a well designed dac that supports DSD, but even if we take out high resolution and dsd and just consider 192K PCM, a well designed dac should could care less about the input, noise, etc...it should be able to compensate for all of those things, especially since the end product includes the analog conversion and "wave shaping".   The "color" may be different based on the dac, but the input should not carry any weight if the dac is designed properly.

 

Whether we are at that stage and generation of DACS now is yet to be decided, but i do believe that a 25K dac definetely should be able to compensate, especially if a $199 schiit with unison usb truly does what it markets to do.

 

[beerandmusic]

18 minutes ago, 6aardvark9 said:

Perhaps all DACs are essentially limited in varying degrees as to what they can do (i.e. reclocking, jitter removal) with incoming digital signals, and therefore any 'pre-work' that can be performed by upstream 'USB gadgets' to improve the signal will lessen to work required by the DAC and thereby may improve the analogue output... 

 

Remember the only thing going to the dac is digital data which is recieved correctly, the reference voltage and "noise".

The purpose of these "gadgets" is to isolate the noise, provide a "cleaner reference voltage" and the digital data.

The dac already receives the data accurately, so the only thing remaining is the reference voltage and the noise.

A well designed dac "should" be able to resolve for those 2 nuissances without the need do daisy chain additional power supplies and noise isolation. 

 

Dac designs continue to improve with this knowlege, and with current generation dacs by reputable designers should compensate.  I believe this is what schiits unison and perhaps some of the higher end dacs do already.   Galvanic isolation for dacs wasn't even a big thing a short couple years ago.

 

I think we are finally getting there....jmo...just taking awhile...

 

[beerandmusic]

1 hour ago, davide256 said:

 DAC's reproduce/flow control what the USB chip sends them. They have no influence over what happens before the USB chip. USB just happens to

be the 2nd best problem prone/ closest to internal computer protocols transport solution compared tto coax and toslink. Ethernet is the least problem prone

digital data transport for audio if you have the rare DAC  and software driver to support it. The reason I say its best is because USB is a hybrid digital/analog

signaling method for very short distances, Ethernet is pure digital domain until the Ethernet enabled DAC begins processing data and can be used regardless

of distance between source and DAC

 

What you miss is that endpoint computer hardware and software integration are key. Computer audio is easily degraded between disk read and

hand off to the DAC output by software and hardware. Apple doesn't have any better a USB port than anyone else... but they control

completely the software and hardware integration which as a byproduct means that most Apple computers do audio acceptably well vs the crap shoot

with an off the shelf windows PC. Likewise NUC's as an Intel controlled product do well. But more is always possible, Apple and Intel both have to make machines

that are general purpose vs solely designed/optimized for audio.

 

Do you agree that the only thing that is sent from the usb to the dac is:

a) digital music (most everyone agrees that this digital data is received accurately)

b) reference voltage

c) noise

[beerandmusic]

45 minutes ago, AudioDoctor said:

Interestingly, my new DAC has a Roon Bridge built in, however I find the sound to be better coming from the USB output of a Sonore Signature Rendu SE via Roon and HQPlayer than using the Roon Bridge directly.

 

Also keep in mind liking the sound better does not always equate to accuracy and there may be some emotion and perception involved as well. 

[beerandmusic]

nvm

[beerandmusic]

16 hours ago, davide256 said:

Do you agree that the fuel pump on your car sends gas? And if so do you disagree that a fuel pump doesn’t behave any differently whether the gas is good or poor? That’s all asynch USB is, a “fuel pump” for the DAC that doesn’t care or know how good or bad it’s source is

comparing good or bad gas would be more like comparing one song to another, and besides doesn't make sense to compare usb to a fuel pump.  usb always transmits the bits accurately, even if there is noise on the line....if you can get the bits perfectly, a good design should be able to compensate for any noise.

[beerandmusic]

4 minutes ago, 6aardvark9 said:

 

It's the timing of delivery, not the fact that the right 'bits' get delivered. Accuracy has to apply to both, that's why master clocks make a difference and why high end dacs/streamers have external word clock inputs

 

I would not argue that...but that could/should be a function of the dac.

If the dac can get the bits accurately, the design of the dac should be able to take that "perfection" up to the conversion circuitry...especially when you have a dac that does it's "own" regneration...

 

this should not be rocket science...and perhaps we are already there, but just subjective opinions are creating their own market.

[beerandmusic]

Just now, 6aardvark9 said:

 

I think in some respect its a limitation of the processing power... this is why Chord developed the MScaler to take a huge chunk of the upscaling outboard from the DAC, so that the DAC (Dave/Hugo(TT)2/Qutest) has less to do, thereby improving the 'sound'. They couldn't put both MScaler and Dave together in one box because of the required current draw (the BluMk2 CD transport draws up to 10A for MScaling, and gets f'ing hot to the touch.

 

this processing power is for "upsampling", which is a totally different subject, which i have my own debate about different than this one.

 

[beerandmusic]

Just now, 6aardvark9 said:

 

I would also add that the performance of Dave/MScaler is very much improved with upstream reclocking… 

again, might not argue that point, but that is different topic.

 

 

[beerandmusic]

5 minutes ago, 6aardvark9 said:

 

I thought it relevant, since the (reclocked/improved) timing of the digital signal makes a difference, even with high(er) end DACs

 

iimproved timing is likely needed for upsampled DSD rates which is not at least what i am looking for.

 

I just want to know if a technology such as unsion that includes isolation and regeneration is sufficient without getting into the debate of high-res which i consider a totally different topic, that i have my own and separate debate about.

[beerandmusic]

1 hour ago, davide256 said:

So basically you are arguing that their is no difference between the output presented to USB from the cheapest computer possible to the best possible. Good luck with that approach,

you won't get past the level of sound quality you hear at a Walmart. Software, hardware and power supply do matter.

 

No, I am not arguing there is no difference, I am questioning the logic and reasoning.  You would think a DAC engineer would be able to provide a "logical" explanation...but even DAC engineers shrug like they do not know why.

 

Keeping in sight these basic principles.

 

The only thing transferred to the dac from the pc are:

1. Digital bits (which everyone agrees the DAC receives with 100% accuracy)

2. A reference voltage

3. Noise

 

4. If you take into consideration that the DAC receives its bits with 100% accuracy and that the PC's reference voltage should have no impact since it is isolated from the DAC's Design.

 

5. If you take into consideration that the DAC has it's own reclocking/regeneration of the bits that it recieves with 100% accuracy, then the PC should not make any difference.

 

A different dac may make a difference, but a different pc should not make a difference if you accept 1-5.  If you don't accept 1-5, then which don't you accept and why?

 

Again, not comparing DAC A to DAC B, comparing using different pc's with same DAC.

 

[beerandmusic]

2 minutes ago, davide256 said:

 

DAC engineers would consider anything before the USB output as "someone else's black box"... they don't get paid to solve someone else's problems.

A software engineer for a computer audio program might be a better place to start... they could give you a sampling of technology reasons why current computer audio

is not "the best of all possible worlds"

 

Any software engineer (or hardware engineer) would state what I have...that the bits are transferred perfectly along with a reference voltage and noise.  I am not referencing potential issues with upsampling. 

[beerandmusic]

7 hours ago, sandyk said:

 

The Data is sent as analogue waveforms , the shape of the waveforms and voltage levels, as well as the timing is dictated by the front end, and will also obviously look different at the DAC end of the cable which in the case of USB may be <1M long,  or not much more than 5 Metres long,(at which length errors may occur) where the input circuitry (Schmidt Trigger etc.) will try to convert it into a correct

1 or 0.
With a less than perfect input waveform, the receiver's Schmidt Trigger etc may toggle at a slightly different part of the waveform, resulting in some timing variations.

This also applies to Coax SPDIF where they may use a Schmidt trigger (e.g. 74HCU04)

 to try and " square up" the waveform.

 

 I get it that you don't like what I am saying, so now feel the need to childishly revisit every recent post I have made as payback, on any subject ,and click Disagree, like in the case of the recently banned Charleslatham who  felt the need to click " laugh" on every post by 2 members.

 

See for example

#528

no, you have childishly clicked disagree with every post i make, without offering any reasoning, even on posts that i never made a statement that you can disagree with.  In one such post I said I am trying to understand the logic of something, and that is all i said in that post, and you clicked disagree with.  There was NOTHING in the post to disagree with, i was asking a question.  You were just being annoying on purpose.  Next time read what i write, and if you do not understand what I am asking, then either don't respond, or ask for clarification, but don't disagree with a "question" i make when it is posed as a question.  You were just trying to be annoying and you know it.   Instead of just squabbling with me, or following me around, just put me on ignore, and I will ignore you as well so we can both go on peacefully. 

[beerandmusic]

7 hours ago, davide256 said:

Actually they would be mystified as to why you are reducing  a layer 7 application level problem to just a  layer 2 link level problem... assuming they were educated  in computer technology and data transmission. Which is definitely not the skill set of a DAC engineer.

There is NO application problem in the transfer of bits where the bits are received with 100% accuracy. Everyone to date has agreed that the bits are received accurately.  And it also has been stated that if bits are dropped that it would be heard as a tic, not as "clarity" in music as you have suggested.

[beerandmusic]

10 minutes ago, sandyk said:

 I gave you an indepth explanation in post #49   but you have shown that you simply aren't interested in  any explanations that are in disagreement with your own P.O.V.  I only click Disagree when I have a problem with the content of a post , which is normally based on hands on experience in that area as an  experienced DIY person who has constructed several DACs and hyas also

done indepth investigations in the USB area , including this old thread.

https://www.computeraudiophile.com/forums/topic/13905-continuing-pursuit-of-power-supply-improvements-and-improved-dac-performance/

 Your 5  "Disagrees'' ) appeared in quick succession showing that you quickly went through all my recent posts in different threads specially , just as another member who has already been banned a couple of times already is presently doing.

I didn't click disagree with you UNTIL AFTER you clicked disagree with a question i had.  It makes no sense to disagree with a question.  I have removed my "disagrees with you".  Please just put me on ignore.

 

[beerandmusic]

7 hours ago, davide256 said:

Actually they would be mystified as to why you are reducing  a layer 7 application level problem to just a  layer 2 link level problem... assuming they were educated  in computer technology and data transmission. Which is definitely not the skill set of a DAC engineer.

 

Or perhaps, let me ask it in a different way.  If you believe that the data is streamed differently than the bit-pefect flat file (lets just assume flac file pcm), how would you know that it is not transmitted without corruption, and what would it sound like, if a few bits were received incorrectly?

[beerandmusic]

8 hours ago, davide256 said:

Actually they would be mystified as to why you are reducing  a layer 7 application level problem to just a  layer 2 link level problem... assuming they were educated  in computer technology and data transmission. Which is definitely not the skill set of a DAC engineer.

 

Very informative regarding what I am speaking about:::

 

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

 

esp...I can confidently state that you’ve never heard a timbral or fidelity change when connecting a digital source and a destination using a USB cable.  (with crc errors, you may get glitch, but unlikely, and you would hear it as a dropout, not as change in sq).

 

[beerandmusic]

41 minutes ago, audiobomber said:

I disagree, and I believe so do most of the people on this website. I'm not sure why you are still here belaboring this point. If that's what you believe, just buy AnyDAC and be happy. 

 

This is an old article but the basic principles hold up today, IMO.

http://www.enjoythemusic.com/magazine/manufacture/0509/

 

I am not going to read the entire article, but if you want to share what are the line(s) that state what the resulting sound will be in the case of loss bits, i will entertain the thought?

 

This discussion came up many times, and i used to be on the other side of the thinking...but it seemed all/most of the audio engineers here agreed that the bits are received correctly and when not, it just caused a glitch.

 

 

 

 

 

[beerandmusic]

4 minutes ago, davide256 said:

you need to stop spouting BS. What you mean is the engineers you want to agree with in the objective camp are spouting this... not the engineers who are actually exploring and selling

products.

 

I don't want to agree with one side of the camp or the other.  I am unbiased, and just want the facts.

Two years ago, i started this a topic with this same thinking...

============================================================

As a matter of fact, here is an exact quote of mine from 2 years ago:::

 

I accept noise of all types affect SQ, but everyone kept suggesting that the DAC gets the music bits perfectly, but that noise affects the dac in processing.  This suggests that the DAC doesn't even receive the music bits with accuracy due to noise....big difference.  Some suggest that dacs can resolve for most noise, that's all dandy, but if the music is already inaccurate before it gets to the DAC input pin(s), there is no way to fix it at that point, if it is already corrupted.

===========================================================

The topic ended with basically no one agreeing with me or IFI....that bits lost are either SO RARE and even if they do happen it would cause a dropout, not a difference in SQ....but if people think differently today, that certainly goes along more with what I find logical.

 

[beerandmusic]

^^^  I will bring the old topic back to life and see if there is still a consensus....

IMHO, the only thing that is logical to me is that noise does indeed cause corruption of data and therefore distortion as far as why a LPS on a unix fanless computer would sound better than a general purpose windows machine?

 

Again, i am not on either side of "any camp"...i just want to understand the logic and reasoning.

 

From 2 years ago, I thought most everyone agreed the bits were "nearly" always perfect and if they weren't you would hear dropouts.